IDP Education has published an annual report for the 2026 financial year. There is some interesting news regarding IELTS. Like:

1. 1.195 million IELTS tests were delivered by IDP in the year, a decline of 8% compared to the previous year.

2. More specifically, IELTS volumes are down 22% in India and down 3% outside of India.

3. These declines come despite opening 13 new test centers in China (up from zero).

4. A bright spot is Australia, where IELTS volumes are up 8%.

5. The overall number of IDP IELTS venues declined from 1,500 to 600 worldwide. I wrote about the closing of IELTS venues in Canada a few months ago. I guess this was part of a larger trend.

6. IDP specifically notes that IELTS volumes have declined in India, Iran and Canada and increased in Kazakhstan, United Arab Emirates and Kenya.

7. IDP plans to roll out what it calls “Computer Mediated Speaking.” I was excited about this for about 45 minutes, but a smart commenter quickly mentioned that it is likely a reference to increased use of video calls for the IELTS speaking section, rather than something more innovative.

8. Cost reductions within the IELTS program have included “moving some countries from direct models to third-party models and exiting some countries.” I know that Japan has moved to a third-party model. I don’t know which countries IDP has exited.

9. The average IELTS fee is now AUD $325.30 (up 2%).

IDP’s shares are down a startling 27% as I write this.

There isn’t much good news here. IELTS volumes have dropped 8% during a period in which PTE volumes have essentially stabilized and smaller competitors seem to be growing.

Meanwhile, this morning’s conference call contained some useful information about IDP’s expansion into China. It describes this expansion as utilizing a third party delivery model with plans for “further staged rollout” over time.

A caller asked about the impact of governments which are starting to look favorably on at-home tests (specifically mentioning the UK Home Office) but the response from IDP contained no useful insights on that file.

A small detail in the call stood out to me. Regarding changes to the IELTS, it was mentioned that IDP will:

“…continue to innovate the product with the launch this year of our IOCP product, which essentially allows test takers to complete the writing component of their test on paper, in a computer based format.”

What this describes is a system wherein the IELTS test taker looks at an essay question on a computer screen and then writes their essay on a piece of paper they have been handed. This is innovation of a sort, but hints at the slow overall pace of change within the IELTS partnership.

IELTS Increases Acceptance!

The various IELTS websites have been updated to reflect the fact that the IELTS test(s) are now accepted at 13,500+ institutions (up from 12,500+).

This pushes IELTS ahead of TOEFL, which is accepted at 13,000+ institutions.

ETS Partners with Passage

ETS has announced a new partnership with Passage that will “will help institutions more effectively identify and connect with best-fit students at scale.” A press release notes that the partnership will link TOEFL and GRE results with Passage’s AI infrastructure to create “new opportunities to use trusted assessment information within AI-enabled workflows, while maintaining ETS’s existing role in assessment delivery and administration.”

Passage was founded in 2023 by Martin Basiri, who previously co-founded the study abroad platform ApplyBoard.

The partnership was covered in Toronto’s National Newspaper. That reporting includes this interesting bit about the early days of Passage:

“But after the Canadian government pulled back on foreign student visa approvals and cut immigration levels in 2024, Mr. Basiri needed a new strategy. He assembled financial backing to buy two of the world’s best-known college and university admissions tests – TOEFL (Test of English as a Foreign Language) and GRE (Graduate Record Examination) – from Educational Testing Service (ETS), but the U.S. non-profit didn’t proceed with the divestiture.”

This shouldn’t come as a surprise, as Martin Basiri was specifically mentioned in the Wall Street Journal article about the potential sale of those tests, but it is fascinating to see how close he actually was to making the deal.

The article goes on to describe the present arrangement, which it describes as a “backup plan.” There is also a bit of back-and-forth about whether Passage is now in direct competition with ApplyBoard.

New TOEFL Bulletin

A updated version of the TOEFL Bulletin is now available. It’s the same as the old one, but references to paper score reports have been removed. Those are no longer available.

New Duolingo Demographics

Now available is an updated version of the Duolingo English Test’s “Demographic and Score Properties of Test Takers.” It covers the year ending June 2026.

There’s a lot in here. I wish other test makers would publish comparable data.

A few things caught my eye:

1. Mean scores (both overall & per section) have dropped about one point each since last year.

2. 68% of test takers are first timers, compared to 70% the year before.

3. The countries with the most repeaters? In this order: Korea, Japan, Taiwan

4. Test takers are 48 percent female, 47 percent male

5. Countries with greatest percentage of female test takers: Philippines, Puerto Rico, USA, Iran, Kazakhstan, Canada, Afghanistan, Poland

6. Spanish is the most common first language (12.71% of test takers, up from 10.87% last year), followed by Mandarin (8.39%, down from 14% last year).

7. Seems like Korean and Portuguese are growing, in terms of overall share.

8. 46.89 of all tests were taken in Asia, down from 55.52 last year

9. The top countries by test taker ID: Mainland China (10.19%), Canada (8.33%), India (7.83%), United States (5.19%), Brazil (4.38%), South Korea (4.03%), France (3.83%), Mexico (3.70%), Japan (2.51%), Pakistan (2.47%)

10. Last year, the top countries were: Mainland China (12.88%), India (12.43%), Canada (5.49%), United States (4.23%), Brazil (3.49%), South Korea (3.39%), Mexico (3.36%), France (2.95%), Nepal (2.71%), Pakistan (2.40%)

—

IELTS usually publishes demographic data around the middle of August, so keep an eye out for that. The TOEFL program usually publishes an updated “Test and Score Data” summary at the end of the month, so watch for that as well. And I think CELPIP owes us an annual report of some kind.

And Other Duolingo Stuff

Also available from the Duolingo English Test team are the inaugural “Duolingo English Test Responsible AI Transparency Report” and an updated version of the “Duolingo English Test Responsible AI Standards.”

Find both right here.

Duolingo for DAAD

Students can now use Duolingo English Test scores when applying for scholarships from the DAAD German Academic Exchange Service.

This will make life easier for students. An increase to the range of available options always does.

Readers interested in the business of English testing must understand the continuing growth of this particular test. I still encounter people who ought to understand it fully, but choose not to.

Want to Support this Newsletter?

So I keep getting financial pledges via substack from people who want to support this newsletter. That’s really awesome, but due to draconian banking regulations here in Korea, it is unlikely that I will ever be able to turn on paid subs and claim those pledges.

If you really want to support my work, feel free to visit my Ko-Fi page, where you can easily make a donation via credit card or PayPal or something else. Two cool people have already made donations.

Research from LanguageCert

Now available is the fourth volume of LANGUAGECERT’s “Certifying quality in assessment & learning: Research and validation at LANGUAGECERT.” It collects research published by LanguageCert between 2024 and 2026. The editors are Leda Lampropoulou and David Coniam. I think it was published this month. Or maybe at the end of last month.

This is an impressive body of research. Most publishers of English tests don’t seem to produce any ongoing research at all – beyond an opening gambit in the form of a white paper, they don’t seem to produce research in support of either their test or in support of language assessment generally. But LanguageCert publishes both on a regular basis.

Language Testing Forum 2026 Registration Opens

Registration is now open for Language Testing Forum 2026. The event will be hosted by Oxford University Press at Worcester College in Oxford in November. This year’s theme will be “Methodological Choices in Language Assessment: Traditions and New Directions.”

I visited Oxford a few months ago. It’s a very pleasant place.

Early-bird rates are available ’till September 20.

CAEL To Require Second Camera

Beginning September 4, the CAEL test will require a second camera, when taken remotely. According to the most recent test data I’ve got, this is how most people take the CAEL nowadays.

We don’t see too much news out of Camp CAEL these days, but I welcome the updates we do get. The CAEL might be one of the last tests of its type – 3.5 hours, 100% “academic” English, long items, some unscored items, and plenty of integrated questions.

Cambridge-Michigan Financials

A new 990 form for Cambridge Michigan Language Assessment, LLC is now available. This is the non-profit organization which owns and administers the Michigan Language Assessment suite of tests (the MET, most notably).

The 990 indicates program service revenues of 9.21 million USD for the year ending July 2025. That’s an increase of 8.67% from the previous year. Revenues have increased every year since 2020, actually.

9.21 million is not a particularly large amount of revenue when talking about this sort of thing. But here we have, once again, a test that appears to be growing (and meaningfully so) at a time when the most prominent tests have reported significant declines. This is a story (and a theme) to follow.

On the Sharing of Test Responses

A report by The PIE’s Amy Baker from the recent LANGUAGECERT summit in Greece describes a debate about whether universities should have access to English test responses.

It is nice to see this being discussed. As noted in the article, this can possibly “help admissions teams make more informed decisions.”

Many test makers already do this sort of thing, to various degrees. Duolingo English Test shares a speaking and writing sample. I know that Password shares a bunch of responses. I think iTEP International shares basically everything. TOEFL probably still shares one speaking and one writing response with score users, but the relevant documentation hasn’t really been updated to match the new TOEFL format.

New Book on Remote Testing.

Springer has published a new book called “Quality Standards for Remote Digital Language Testing.”

Edited by Tony Clark and Emma Bruce, the book “is a collation of evidence-based test provider and institutional accounts of the remote online delivery and administration of tests originally developed for in-centre delivery.”

The editors also “propose a framework… to support remote online testing as an established mode of delivery.”

Photo of the Week

The TOEFL website now lists updated rosters for the TOEFL Board and the TOEFL Committee of Examiners. And for the first time, the page now includes biographies of each board member. That’s neat.

Amy Dambrowitz of the Association of Registrars of the Universities and Colleges of Canada continues serving as chair of the board. Koenraad Van Gorp of Michigan State University remains chair of the Committee of Examiners.

The Board “[ensures] ETS staff are aware of trends in student mobility and the needs of higher education institutions so the TOEFL Program can better serve students and institutions.”

The COE “[provides] insights into trends and developments in the areas of English-language teaching, learning and testing” and “[offers] advice on test content and research related to the TOEFL test.”

Subscribe now

Leadership Change at ETS

There was a change this month to the masthead on the ETS website. Rick Sinkfield no longer serves as Chief Legal Officer of the organization. He has departed for Covista, an operator of for-profit higher educational institutions in the USA and the Caribbean, where he will serve as Vice President of Expansion.

It appears that Sinkfield’s duties at ETS will now be handled by Peter Betaneli, who has been with ETS since 2008.

ETS on Security

There’s a nice article on the ETS website by security boss Wallace Dalrymple Jr called “Future-ready assessment security starts with better questions.”

It notes:

“Good security is asymmetric: nearly invisible to honest candidates and expensive for bad actors. The goal isn’t to create more friction for everyone. It’s to create meaningful friction precisely where it belongs, for the people trying to undermine assessment integrity, while maintaining a fair, accessible, and positive experience for legitimate test takers.”

TOEFL Essentials Research!!!

New in Language Testing Journal is an article by Daniel Isbell, Dustin Crowther, Jieun Kim and Yoonseo Kim which explores the degree to which the speaking tasks in the TOEFL Essentials test reflect the ability of students to speak English in academic settings.

Regular readers know that the TOEFL Essentials test can no longer be taken. But this is important work because the new TOEFL iBT test is pretty similar to the Essentials test. Actually, the new TOEFL speaking section consists of two of the three tasks found in the Essentials test (virtual interview & listen and repeat).

On top of this, the TOEFL Essentials test will be reborn as an option for Canadian immigration any day now (or not).

The article describes a study in which 199 students submitted “lab based” speaking tasks (a graph description and a lecture response) meant to resemble authentic academic tasks. Next, 149 of those students successfully completed the TOEFL Essentials test. Sixty-five of those participants then submitted samples of course-based speaking from their field of study. Both the lab-based and course-based samples were rated and analyzed (check the article for details, as this goes deep).

The authors found that the overall TOEFL Essentials speaking score correlated strongly with the lab-based tasks (r=.85 and r=.86). The overall scores also correlated strongly with the course-based tasks (r=.76).

TOEFL ITP Test and Score Data

You can now read the TOEFL ITP’s “Test and Score Data” for 2025.

I don’t pay much attention to this product, but it is still pretty popular in Japan, Indonesia and some Latin American markets. It is often used as a placement test or an exit test in those locales.

The test is quite similar to the old pre-2005 version of the TOEFL test, though there is an optional speaking module that can be taken. During the pandemic, ETS introduced a product called “TOEFL ITP Plus for China” that was used for university admissions when regular test centers were closed.

An updated test taker handbook was published last year.

On Tracking Eyes and Logging Strokes

Here’s an ETS Research Report by Ching-Ni Hsieh and Renka Ohta concerning a study that tracked the eye movements and keystrokes of 19 people who completed the “Writing for an Academic Discussion” task in the TOEFL. Participants were also asked to recall their experience completing the task.

It seems like most of the students planned their answers on the fly rather than in advance. Revisions mostly focused on lower level edits rather than higher level stuff like coherence and organization. This is not so surprising, since the task is completed in just ten minutes.

On Note-Taking and Performance

There is a fantastic article in TESOL Quarterly this week by Jieun Kim which explores the impact of note-taking methods on performance in English tests.

The article examines the performance of 305 students with varying levels of proficiency, each of whom completed four listening tasks (lectures only) from the former TOEFL format. Of these participants, 102 took notes by hand, 102 typed their notes, and 101 completed the tasks without taking notes.

(Note that each lecture was 4.5 to 6.5 minutes long, and was followed by six questions)

Overall performance was best among the hand-writing group, but only slightly so. The author notes that “no significant difference among the three note-taking groups was found.”

Language Testing Journal on SDS

You’ll find a ton of articles about AI-driven Spoken Dialog Systems right now over at Language Testing Journal. Writing an item about each one will push me past the Substack character limit. So I’ll just send you over to the main index.

Duolingo Reports Quarterly Numbers

Duolingo just published its Q2 results for 2026. Revenue from the Duolingo English Test is $10,109,000. That’s almost exactly identical to the same quarter last year ($10,088,000). The test’s price tag has not changed since then.

Assuming a base price of $70 per test, we might assume that the test was taken 144,414 times in the quarter.

This is a very rough estimate, though. Some people pay more than $70 to get express scoring, while others pay less than $70 by purchasing a bundle of two tests. Some pay nothing at all as they get free tests through the DET Access Program. In addition, Duolingo (like other test creators) is quite integrated with third-party sellers of test vouchers, which likely depresses the average test fee somewhat.

IDP Now Operating 25 IELTS Test Centers in China

IDP Education is now operating 25 (!) IELTS test centers in China. That’s up from 13 the last time I checked. Here’s what I’m seeing:

• Beijing x1

• Hebei x1

• Shanghai x2

• Jiangsu x6

• Zhejiang x2

• Anhui x1

• Fujian x1

• Shandong x2

• Henan x1

• Hubei x1

• Hunan x1

• Guangdong x2

• Hainan x1

• Chongqing x1

• Sichuan x1

• Shaanxi x1

Regular readers know that IDP has spent 2026 building up its IELTS operations in China, starting with five test centers back in January. They are in competition with the British Council, which previously had a monopoly on IELTS testing in China.

For reasons that aren’t clear right now, IDP only offers the IELTS Academic test in China.

IELTS on Paper?

I chuckled when I saw official messaging from IDP Education refer to the new “Writing on Paper” option as “Paper based IELTS. Now better than ever.”

On IELTS and the Certification of Teachers

An excellent new article in EducationHQ by Xiaoxiao Kong explores the fitness for purpose of the English language tests (specifically IELTS) that are used for teacher certification in Australia. The article summarizes the author’s doctoral work.

The conclusion seems to be that while there is some overlap in the English that shows up on the tests and the English that teachers use in their work, there are also significant gaps.

For instance:

“Although the [IELTS Academic speaking section] is conducted face-to-face, the examiner largely controls the interaction: the examiner asks the questions and the test-taker responds. Teachers in the focus groups described parts of this format as resembling a job interview rather than the reciprocal conversations they engage in at work.”

And:

“IELTS Academic listening uses recordings of adult English speakers in social, educational, and academic contexts. Teachers in my study, however, described workplace listening that frequently involves children and young people. Their speech may be incomplete, unclear, and shaped by their age, development, and emotional state.”

The author notes that her work is not a call for standards to be lowered or removed. Rather, she argues that her findings “point to the need for stronger evidence that the assessment being used is fit for its intended purpose.” And that an ideal test “should sample the communication skills that are most important in the profession.”

Our Sponsor: PrepEx.ai

This month’s newsletter is sponsored by the folks at Prepex.AI. Based in Hong Kong, PrepEx provides AI-based prep for a variety of standardized tests. One thing that seems to set this platform apart from others in the same category is a sort of “Expert in the Loop” approach - test and subject matter experts regularly review the platform's practice items and train the AI to ensure accuracy.

Meanwhile, I read last week on LinkedIn that Prepex recently completed a pre-seed around at a two million USD valuation. That seems like a big number.

If anyone wants to sponsor next month’s newsletter, just gimme a shout.

Pearson Publishes Half-Year Numbers

Pearson just published its latest 6-month financial report. It indicates that PTE volumes are down 3% compared to last year. The report says:

“PTE revenue declined, with volumes down 3%, as market conditions became more difficult driven by tight migration policies and geopolitical disruption.”

Note that volumes declined 5% in 2025.

Coming Soon: Official PTE AI Practice Platform

I attended the big PTE online event yesterday. I learned that Pearson will soon launch an “Official PTE AI Practice” product. It will contain 15 full PTE mock tests, 20 sectional tests and 1000 individual questions. It will provide “instant AI scoring and feedback aligned to official PTE scoring rules.” The scoring will be “immediate and accurate” and be accompanied by “actionable feedback.”

This will be a paid product. There will also be a PTE Core version.

Big Oxford Test of English Volume Increase

The Oxford Test of English was mentioned in the latest OUP annual report. It says:

“The Oxford Test of English reached multiple new markets, while the Oxford Test of English Advanced launched in China Mainland, Vietnam, and Thailand. Collectively, these contributed to a 30% increase in the number of test takers across the world.”

Thirty percent!

This is one of the key themes I’ve tried to explore in this space. Beyond the “big four” English tests (IELTS, PTE, TOEFL, DET) there is a long tail of smaller tests, and most of them seem to be growing in popularity, even in challenging market conditions.

When a major test publisher says that volumes are decreasing due to regulatory and policy changes in key markets they are being honest… but it isn’t the whole picture. Competition and choice are also relevant to the discussion.

TOEIC Numbers for FY 2025 (Japan Only)

The IIBC has published TOEIC Program stats for the 2025 fiscal year. They indicate how many times the TOEIC was taken in Japan, and how test takers performed. Check the comments for a link.

Reports from IIBC are always worth keeping an eye on, as Japan seems to account for at least 60% of global TOEIC sittings. On top of that, such statistics are simply not available for other markets.

The report indicates that the TOEIC was taken 1,811,979 in the country. That’s a 2% increase from the previous year. The overall total is still well below the 1,985,051 who took the test in the final pre-pandemic fiscal year, but things are looking positive at the moment.

Most readers probably know that TOEIC is by far the biggest English test in Japan… but might not be aware of exactly how much more popular it is than the high stakes players. Compare the 1.8 million people who took the TOEIC (in Japan alone!) with the 1.1 million who took the PTE worldwide over roughly the same timeframe, or the estimated 750,000 who took the TOEFL. Just keep in mind that the TOEIC costs only about $47 to take in Japan.

Below is a Google Trends image below which shows how interest in the TOEIC really outstrips interest in TOEFL and IELTS (in Japan):

Oxford ELLT Gets Express Option

I saw that the Oxford ELLT test from Oxford International Education Group now offers an express option. For a fee of £80, test takers can get a same-day speaking appointment, and test results within 12 working hours. Lots of tests have an express option (Duolingo has a 12-hour option and TOEFL may still have a 48-hour option) but I haven’t seen such a thing for a test that still uses live speaking interlocutors.

I’m sure there will be challenges related to appointment availability and time zones that don’t quite line up, but it is a worthwhile project to pursue. Test makers are finally starting to realize how many of their customers leave this critical step until the last possible moment.

CELPIP Seems Popular in Canada

For the first time, CELPIP has surpassed IELTS in the Google Trends chart for Canada. This means that (for the moment) there are more people in Canada searching for information about CELPIP than about IELTS. I’ll paste an image of the 5 year trend below.

Canada isn’t a great market for tests right now, but we can see that search interest in IELTS has declined, while search interest in CELPIP has increased.

This data is obviously of limited usefulness, but it is worth noting. I think it speaks to the ability of Prometric to grow their test in a market that IELTS once had mostly to itself. It also comes at a fortuitous time for Prometric, as IRCC changes are set to expand the amount of testing done in Canada sometime next year.

In conversations, I’ve heard students express what they like about the test. Specifically:

1. It is cheaper (approx $375 vs $340)

2. It is completed all in one sitting.

3. It only uses Canadian accents.

4. It is perceived to be easier.

It is worth noting that IELTS is used for both immigration and university admissions in Canada, while CELPIP is only used for immigration (for now, anyway).

On the Marketing of English Tests

I really liked this recent article in The PIE by Kimberley Martin. It describes how certain education brands are promoting themselves in 2026.

There is some stuff in the article about the ways ETS has marketed the TOEFL in recent months. Their efforts include the TOEFL music video, the TOEFL theme song, the TOEFL dance contest, the TOEFL playlist of AI-generated songs, and the TOEFL animated AI spokesperson.

These things are… what they are.

But how does one go even about marketing a standardized test? Test takers view tests as obstacles they must overcome while they are on the road to a fulfilling and meaningful future. How might a company promote one obstacle as being better than some other obstacle? How do you convince someone to pay $400 for YOUR obstacle when they can buy someone else’s obstacle for $150 and still end up in the same place?

I’ve been on the “Business of English Testing” beat for about 15 years now. I’ve seen a lot of really bad marketing, a little bit of decent marketing, and some promising starting moves that ultimately didn’t go anywhere.

It is said in the article that “[t]he new generation is immune to marketing.” And that’s in the context of marketing a university. Marketing a test must be 100 times more difficult.

On Getting Your Test Accepted in Spain

Over on the Linkedin, Dan Shepherd has a nice article about how a test company might get their English test approved by ACLES, the association of university language centers in Spain. They determine which tests are used by Spanish universities.

On The Future of Test Security

I finally watched the ATP panel discussion on test security featuring Basim Baig, Wallace Dalrymple Jr and Claire McCauley. It’s a worthwhile view, if you are into that sort of thing. I’ll link to it below.

Basim made an interesting point near the end of the discussion about how smartphones were once viewed solely as a security threat, but are now used as a tool to increase security (by providing a second camera). Accordingly, he imagines a future world where smart glasses and other wearables become common enough that they can also be used to enhance rather than degrade security. Perhaps smart glasses will never become popular among the general population, but I can envision other kinds of wearables being used to increase test security in the distant future – even for tests taken at test centers.

Photo of the Week

Cambridge Fined by Ofqual for Scoring Mistakes

Cambridge English has been fined £875,000 by Ofqual for incorrectly scoring tens of thousands of IELTS tests taken between 2022 and 2025. As regular readers recall, 62,794 test takers were impacted by this problem. Most were given scores which were lower than they deserved.

Ofqual’s press release (and accompanying documentation) contains some details that are new to me. Including:

1. Exactly 93,865 tests were incorrectly graded. This number is higher than the figure mentioned above, since in 30,649 cases the screwup didn’t result in either a band score change, or an overall score change.

2. 1,108 of the affected test takers “had taken the Secure English Language Test (SELT) offer within IELTS.”

According to Ofqual, the incorrectly graded questions were “gap match” items (I might call them “sentence completion” items). The report describes two problems impacting their scoring. Specifically:

1. “Incorrect ordering of answer keys passed between the test content platform and marking system.”

2. Inconsistent handling of diacritics (accents, umlauts, etc). These are supposed to be ignored when tests are marked… but sometimes they were not.

Interestingly, Cambridge’s initial reaction was to NOT correct all of the incorrect results. Says Ofqual:

“On 26 September 2025, Cambridge English informed Ofqual that, based on its consideration of factors under Condition H6.3, it intended to correct all upward grade changes and apply downward grade changes selectively in scenarios where the adverse effect of not correcting a downward result outweighed the benefit.”

Ofqual “highlighted concerns” about this approach, so Cambridge did end up correcting all of the incorrect results. That’s nice.

What caused the problems in the first place? Well, the report says that “[a]n independent audit commissioned by Cambridge English traced the root causes of the incident back to changes introduced as part of a system modernisation programme which commenced in June 2019.“

These root causes “included (but were not limited to) developments introduced without adequate requirements analysis, documentation, design assurance, resourcing or testing, and ongoing resource needs not being appropriately reviewed or addressed.”

Ofqual’s report calls out Cambridge for “serious deficiencies in the design, implementation, testing and ongoing monitoring of changes made under its modernisation programme to its automated marking systems.”

It also notes that these failures “highlighted systemic weaknesses in Cambridge English’s approach to automated marking systems, where inadequate monitoring and error-detection mechanisms allowed the incident to remain undetected for an extended period, thereby undermining the integrity of high-volume, rule-based assessment processes.”

Compensation was paid to two (2) learners following complaints about missed opportunities.

IELTS Issues Statement

The IELTS partners have issued a statement regarding the big fine they’ll have to pay to Ofqual.

As I read it, a few questions came to mind:

1. The statement notes that “UK visa related tests accounted for 1,108 of affected tests.” I’m left wondering how many misscored tests were taken in the pursuit of Australian visas. Or Canadian visas and study permits. Did anyone look into this?

2. The statement notes that “270 complaints were received and 24 were upheld. Of these, 19 complaints related to a missed opportunity.” The Ofqual documents indicate that two of those complainants were given financial restitution. What did the other 22 get?

3. Were all of the affected test takers specifically notified that it was possible to submit complaints? Were they told how to do so? Were they notified that doing so might result in the payment of compensation for missed opportunities? Would improved communication have led to more complaints?

4. Is it now too late to submit a complaint?

5. The statement notes that “26,246 test takers requested and received a refund” and that 1,145 got a free resit. And everyone else? That isn’t stated. I wonder if it would have been more appropriate to process (or attempt to process) a refund to everyone who sat for a defective test, regardless of whether one was requested or not. Needless to say, no one really got what they paid for.

Anyways.

It is occasionally distressing to me that there are test companies located in comfortable headquarters in America or the UK which extract vast sums of money from young test takers who are predominantly in the Global South without always fulfilling the promises they have made to them.

Reflections on the TOEFL/IELTS Concordance…

There is a new article in Language Testing Journal by Spiros Papageorgiou and Tony Clark called “Reflections on the Practical Implementation of Knoch and Fan’s (2024) Good Practice Principles for Score Concordance Studies.” It recounts the authors’ experience carrying out a concordance study that compared the IELTS and the old TOEFL test (offered from 2023 to 2026).

The article notes that some of the “good practice principles” were easy to implement, while others were not. The authors suggest how to adjust those principles for easier use, and also suggest additional considerations for good score comparison studies.

For one, the authors note that:

“…incentivizing candidates to take a test they had not previously intended to take can be costly and also problematic from a methodological point of view. Test takers who took one test and are invited to take the other might not be motivated to perform well on the second test and might not engage in test familiarization and preparation because that test does not ‘count’.”

And:

“…it is unlikely that participants in a score concordance study constitute a representative sample of the overall population of either test.”

They also note that it was difficult to recruit participants from the higher and lower ends of the score scales.

There was also the question of whether the TOEFL and IELTS tests were similar enough to be compared in the first place. The authors employed some third-party assessors to look over the tests, and concluded that they were similar enough to compare.

That’s fine, but as we enter the next era of testing, this requirement might represent more of a challenge. Already we can see quite a difference between the IELTS and the 2026 iteration of the TOEFL. This might render it impossible to carry out a study comparing those two tests using the current “good practice” principles. And what happens when IELTS itself metamorphizes into its own next-gen form?

Smartly, in the article and the concordance report itself, the authors urge score users “not to rely solely on published score equivalencies but instead weigh evidence from additional sources when making decisions about language proficiency score requirements.”

Concordance studies are complicated business. To me, it sometimes seems like concordance studies are treated like the word of god by schools and government agencies.

Actually, it isn’t even the studies themselves that are seen as divine (I’m not sure they are actually looked at). Rather it is the resulting score conversion tables that are adored. But as we can see, they shouldn’t be the only piece of data used when setting score requirements.

…And a Response

Also published was a response to the aforementioned article.

In it, John Read argues that while score concordance is useful, more could be done when it comes to picking which tests are suitable for Australian immigration. He describes how, in the 1980s, the Australian government commissioned a test called “access” that was specifically designed for migration to Australia. For reasons related to cost, this test was abandoned in 1998 in favor of a requirement that migrants take the IELTS. Read notes that while the IELTS, at that time, addressed some of the needs of migrants coming to Australia, “there was no systematic process of… ‘change retrofit,’ whereby the construct definition and specifications of the test would be fundamentally redefined to be consistent with this broader use of the test for immigration purposes.”

Over time, as we all know, more tests were deemed acceptable for Australian immigration. The article notes that most of the tests currently deemed acceptable were originally designed for assessing academic English skills, and that only two – the OET and CELPIP tests – were designed for some other immigration purpose.

Despite this, Read notes:

“There is little guidance on the immigration website as to which English tests might be appropriate for which type of visa, apart from a note that the OET is designed for health professionals. The assumption appears to be that, once a concordance study has established equivalences with the IELTS-defined proficiency scale, the tests are largely interchangeable.”

This means people are taking tests of academic English for the purpose of non-academic migration to Australia.

I understand, for example, that the PTE Academic test is extremely popular among young people heading down under on working holiday visas. While the PTE is a very fine test, perhaps those folks could be nudged toward a more appropriate test (if such a test even exists).

The article also discusses how migrants are incentivised to chase higher scores in order to get more immigration points, even though “none of the proficiency tests is primarily designed to assess ability reliably at that high level and in particular… to target the kind of sophisticated communicative abilities that are essential in contemporary skilled work environments.”

On the Linking of new TOEFL and old TOEFL scores

There is a document called “A banded score scale for the TOEFL iBT test: Rationale, development, and score linking” from ETS that describes how old TOEFL scores were mapped to new TOEFL scores.

Note that the data used to carry out this mapping is from fieldwork that was done in support of the development of the TOEFL Essentials Test. That test is mostly the same as the new TOEFL iBT, but there are a few small differences. Further study is required, I guess.

I’ll always be curious when the decision was made to flip the switch and replace TOEFL iBT with TOEFL Essentials. I don’t believe that was always the plan… but eventually, like most test watchers, I realized it was an inevitability. Alas, I suppose I’ll never know exactly how the decision came to pass.

ETS Buys ACT

The Educational Testing Service (ETS) has purchased ACT. This comes as something of a surprise. It is a big deal.

Most readers know that ACT’s flagship product is a standardized test used for college admissions in the United States. It is a direct competitor to the SAT. Traditionally, it has been strong in the midwestern and southern states, while the SAT has been strongest on the coasts.

Between 1959 and 2024 the ACT was developed and administered by the non-profit ACT, Inc. In 2024 it was purchased by a private equity firm called Nexus Capital Management for a net gain of about $147 million USD. According to some reports, the non-profit retained a minority stake in the test following that sale.

It must be mentioned that ETS has some experience running tests like this one – it developed and administered the SAT between 1947 and 2024.

According to Higher Ed Dive, ETS will operate the new ACT as a for-profit subsidiary. The same report indicates that Encoura was not included in the sale.

More details as they become available. I’ll do my best to follow the money trail, though such details are generally slow to emerge.

Paper TOEFL Score Reports to End

After July 26, 2026 paper TOEFL score reports will no longer be available.

For many years paper score reports were mailed to all test takers at no charge. A few months ago free mailing was eliminated and only those willing to pay a $35 fee could get one. But soon paper score reports will no longer be sent to anyone at all.

UKCISA Recommendations

The UK Council for International Student Affairs (UKCISA) has published a new report called “International students’ experiences of meeting English language requirements for UK higher education.”

The report encourages us “to ensure that debates about English language requirements do not lose sight of the students who experience them.”

This seems like something which should be obvious to everyone. But as time passes there seems to be more and more distance between the people running test companies and the people taking tests. In 2026 test firms have a really hard time meeting the students where they are… both literally and figuratively.

The report includes five recommendations for UK universities. The following are paraphrases. Forgive me if I have misunderstood anything:

1. Universities should clearly communicate test requirements so that students can better understand the full range of usable tests.

2. Universities should be flexible and supportive of students “who have different financial, geographical, technological and logistical circumstances.”

3. Test providers should minimize the effects of format dependency and cultural unfamiliarity.

4. Universities should understand the limitations of what English tests say about student preparedness, and provide necessary support for new arrivals.

5. Universities should be more clear and consistent when it comes to recognizing prior English-medium instruction.

These are valuable recommendations. A few thoughts on each come to mind:

1. Universities are so so so bad at communicating test requirements. My old university still lists the MELAB test as an acceptable option. One can easily find schools that still list the TOEFL CBT, which was discontinued 19 years ago, as an accepted test.

2. Needless to say, access is important. This is obvious. But at the same time, we have a legacy testing firm that is currently “resetting the conversation on accessibility” by implying that at-home testing is a mere “convenience” rather than a thing which increases accessibility.

3. This one doesn’t get enough discussion. Some legacy test providers are making up for lost revenues by leaning heavily into the test preparation business. As a result, test takers with money have access to nearly unlimited test familiarization materials. And access to official scoring AI, which further increases their familiarity with the test. Does this give them an unfair advantage?

4. Testing companies have started making some pretty grandiose claims about the value of their tests in order to distinguish themselves from their competition. They should probably reign themselves in.

5. When I did a sweep of the twenty UK universities with the most international students, I encountered one that accepted a high school diploma from Ontario or Manitoba in place of an English test score… but not a diploma from the other eight provinces or three territories of Canada. How strange.

IELTS Writing on Paper Coming to London

According to this blog post from MTS, IELTS with Writing on Paper will be available in the UK starting July 13. It will first be available at MTS’s London (Chiswick) test centre.

I think the option is already available in Nepal and Canada.

On the Survival of the British Council…

I spent two hours watching the whole public accounts committee hearing on the financial sustainability of the British Council (and so should you).

There wasn’t much said about English tests, but a couple of lines by Nicholas Dyer of the FCDO caught my attention. He said:

“The British Council entered the pandemic with a business model that was high fixed cost delivering in-person teaching and exams. It’s now got to move to a model – and is moving to a model – where it provides more of this digitally and online and the in-country and the in-person teaching and exams is switching to where people are.”

And later:

“Ultimately, do we want British Council to survive? Absolutely, we do. But in this world we live in, they’ve got to do it in a different way.”

Here we have, from the FCDO, a suggestion that the British Council deliver more of its exams online and that they move closer “to where people are.”

There wasn’t much new information to be found in the discussion, but I did learn that “trapped cash” (monies that the British Council can’t move out of certain volatile countries due to currency controls) will be used in those countries for foreign office business in exchange for a break on loan interest payments. Apparently there is more than 50 million GBP trapped in such conditions.

…And an Audit of the British Council

The National Audit Office has published its “Investigation into the financial sustainability of the British Council.” That’s what the committee hearing was in response to, actually.

There is bad news all the way through.

A few points found in the separate 11-page summary are worth highlighting here. They are:

1. The British Council has lost £184 million since the pandemic. Its free reserves hit negative £111 million in 2024-25.

2. A five-year business plan was implemented starting in 2020. The BC was able to just about reach the cost reductions outlined in that plan through layoffs, outsourcing and closure. They did not meet the revenue increases called for in that plan.

3. The loan now stands at £197 million. No repayments have been made since April 2024.

4. £42 million in interest and fees have been paid since 2020-21. A further £53 million will be paid up to 2029-30.

5. The loan will be extended until September 2027 as negotiations continue on a repayment plan.

6. The BC is proceeding with a turnaround plan which is “expected to deliver £306 million in net benefits by 2029-30.” This will include closures in 11 countries. More on this below.

7. An operating surplus is not expected until 2029-30.

8. A loan write-off seems out of the question.

9. But perhaps a resolution (if you can call it that) is at hand: “At present, FCDO and the British Council are yet to reach agreement on a revised loan arrangement though the expectation is that this will happen soon.” It will likely involve a 15-year repayment plan.

The Guardian has some details on that agreement and the turnaround plan. They are grim. A further 15% of staff would be eliminated worldwide. That’s 1,180 staff in total. And that comes on top of the 2,110 jobs already eliminated. As noted above, operations would cease in 11 countries. They would be “pared back” in a further 15 countries.

IELTS Loses Market share

That audit contains a couple of numbers which may be useful to everyone studying the business of language testing and its long-term trajectory.

The authors of the audit note that before the pandemic, the IELTS partners had about 64% of the global market for English tests. Today, though, they have just a 52% share.

The problem for legacy testers is not that volumes are declining because of reduced numbers of international students. Those volumes will come back, eventually.

Instead, the problem is that they are losing market share to new tests. That market share will, for the most part, never come back.

Here’s the relevant passage from the audit:

“[The British Council] is a member of the consortium which manages the International English Language Testing System (IELTS), a major standardised English test used by visa applicants to the UK and other countries. In 2025-26, the estimated global market share of the British Council and its IELTS partners was 52%. Pre-pandemic, its management of IELTS meant that, with its partners on IELTS, the British Council had 64% of the global market for English tests…”

IDP Education Begins Stock Buyback

IDP Education has begun a stock buy-back program worth about $50 million dollars (Australian). This was reported in a trading update, along with the news that IDP’s cost reduction strategy for 2026 is exceeding expectations ($30 million, against a target of $25 million).

IDP recently outsourced on-shore (Australian) visa support to Bravo Migration, as reported in The PIE.

IELTS Score Reports to be get Ofqual Logo

According to a notice published on the IELTS website, IELTS score reports (TRFs) will soon contain the Ofqual logo and the Ofqual qualification title. These will be added to every TRF generated on July 17 and later.

It isn’t necessary, but test takers with an existing TRF can fill out a form and get a new one at no charge.

I guess this is part of the “Action Plan” that Cambridge English agreed to in the wake of the penalty imposed them by Ofqual a few days ago.

New Duolingo English Test Technical Manual

The 2026 version of the Duolingo English Test technical manual is now available. I’ll link to that below. Remember to scroll down to page 40 for some updated demographic details for the test taker population.

It appears that Spanish is now the most common first language among test takers.

On the Proctoring of the Duolingo English Test

Here’s a cute video that describes how the Duolingo English Test is proctored.

I took the Envoy Test from IDP Education

I recently took the Envoy test from IDP Education. I like it a lot. This is a test mostly used for lower and middle stakes purposes, I guess. Especially within academic settings for things like placement and progress tracking. It is not sold directly to students. You can find some initial reflections on the blog.

LanguageCert Approved in Vietnam

The Ministry of Education and Training of Vietnam has approved the LANGUAGECERT Academic Test and the LanguageCert International ESOL Tests for use in university admissions and exits.

University students in that country are now able to use these tests to meet graduation requirements. For this use case, any test which is recognized by the MOET must be accepted.

More importantly, universities are now able to accept the LanguageCert tests for admissions purposes. This use case requires that individual universities officially announce their acceptance of the test… but I am sure the LanguageCert team is up for the challenge.

This is a pretty big deal. As I’ve written many times before, Vietnam is one of the few places on Earth where test volumes (and interest) are actually increasing. English tests are a huge part of domestic university admissions, as good scores grant students extra “points” when applying to competitive schools. On top of that, most university students are required to prove a certain level of fluency in order to graduate. There has also been some talk of making English a second official language in the country, which may increase testing volumes further.

As a result of all this, testing companies have flooded into the country… all trying to get a piece of the pie.

I spoke to a teacher in Vietnam who was quite impressed by LanguageCert’s accomplishment. Said he: “LangCert took 1 year to do what took Pearson about 5 years.”

Oxford Test of English Approved in Vietnam

The Oxford Test of English has also been approved by the Ministry of Education and Training in Vietnam. It can now be used by university students to meet graduation requirements (any test approved by the MOET must be accepted). And, more importantly, universities can use it for admissions (if they want to).

It seems appropriate to copy and paste what I wrote a few days ago when another test announced its acceptance:

“This is a pretty big deal. As I’ve written many times before, Vietnam is one of the few places on Earth where test volumes (and interest) are actually increasing. English tests are a huge part of domestic university admissions, as good scores grant students extra ‘points’ when applying to competitive schools. On top of that, most university students are required to prove a certain level of fluency in order to graduate. There has also been some talk of making English a second official language in the country, which may increase testing volumes further.”

IRCC Increases Scrutiny of English Test Results

The eagle-eyed journalists over at CIC News noticed recently that IRCC has updated the instructions it gives to staff, re: language test results.

One new requirement is that immigration officers must double-check photos. The new guidance notes:

“If a photograph of the test taker is available in the verification portal, processing offices must cross-reference this image with the photograph submitted with the application to confirm that the test results belong to the applicant. Any discrepancies must be properly documented in GCMS. ”

(GCMS = Global Case Management System)

As regular readers know, the use of proxy test takers is an ongoing problem at test centers. Recent reports in the Hong Kong press have shown us that test takers can just show up at certain test centers and be approached by potential proxies.

It is good that IRCC is attempting to tackle this issue. That said, depending on IRCC staff to simply eyeball photos may not be a good idea. Research has suggested that humans aren’t particularly good at photo matching… even in situations without deliberate fraud. How often will IRCC staff be able to spot proxy test takers who have put in the work to look like the candidate?

This is something that score users may want to think about. They may also consider talking to test providers about it as well.

Here’s what I would ask:

1. Are you doing more than just eyeballing the test taker’s ID when they check in to determine that they are who they say they are?

2. Are you using any kind of software (after the fact) to confirm the test taker’s ID?

3. Are you taking fingerprints or palm prints at the test center in order to give score users something to work with if they have suspicions of fraud?

4. Are you tracking which test centers are more likely to be infiltrated by proxy test takers?

5. Are you tracking keystroke biometrics to detect the appearance of repeat test takers appearing for tests under different names?

Oxford ELLT Gets Updates

I read that the Oxford ELLT test will change (and probably already will have changed by the time you read this). Details are here. Below is a quick summary:

1. The reading section will contain three readings (up from 2), which will range in length from 200 to 600 words. It will now include paragraph sequencing and gap-fill questions (instead of just multiple-choice questions).

2. The listening section will now contain some note completion questions (instead of just multiple-choice questions).

3. The writing section will now contain a short written summary of 170 to 200 words. The existing essay task will be retained, but it is now described as being 190 to 250 words (down from 190 to 260 words). The length of this section will increase by five minutes.

4. The speaking section will no longer include a 3-5 minute presentation (and accompanying 15 minutes of prep time). It will be replaced with what is described as a “student-led monologue” which includes 45 seconds of prep time.

Note that both the Digital and Global versions will change.

This test is one to keep an eye on, as it has achieved surprisingly widespread acceptance in the UK. And at a few schools in Canada.

Photo of the Week

Here’s a well-loved copy of “English Grammar in Use” which I spotted in Lombok.

The Health Resources and Services Administration has published new English test score requirements for foreign healthcare workers seeking to work in the USA. A few points are worth highlighting:

1. The score requirements have increased in most cases. Sometimes by quite a lot.

2. The requirements for those submitting PTE scores have increased significantly, especially the speaking requirement. This mirrors changes implemented by the Australian DHA last year.

3. TOEIC scores are no longer accepted, apparently.

4. Occupational Therapists can now make use of a wider range of tests.

5. The published requirements no longer distinguish between IELTS academic and general scores. Registered nurses (and others at the B.S level and above) were previously required to submit academic scores, while other health care workers (below the B.S level) could submit either.

You can read more on this from TruMerit, who has spent some amount of time pushing for changes like these.

This is big news when it comes to the business of English testing. I’ve already heard from individuals and groups who plan to switch their focus to a new test.

HOELT Debated in Parliament

The remote HOELT was debated in Parliament last week. Several honourable members had a lot to say on the topic.

Blake Stephenson MP (Conservative, Mid Bedfordshire) doesn’t like it. He notes that…

“Remote tests are extremely vulnerable to organised criminal gangs and cheats who, as I have seen at first hand, can easily overcome safeguards with technological workarounds, some of which use cheap equipment readily available on Amazon.”

I am curious to learn more about his “first hand” experience watching people cheat on tests.

On top of that, Mr. Stephenson is also of the opinion that universities should no longer be able to self-assess English fluency:

“Universities should be stripped of their power to assess English language skills for incoming students, which allows them to bypass the official secure English language testing system. Those with a commercial incentive to bring in as many students as possible should not be allowed to mark their own homework. That has created an unacceptable loophole, one that trashes rules designed to ensure that the most basic of requirements is satisfied—that those who come to the UK can speak our language.”

Labour MP Daniel Zeichner (Cambridge) has questions. He wishes to know

“…whether the Home Office has consulted the National Cyber Security Centre on the threat model for fully remote English language testing, including AI-enabled cheating, impersonation, organised fraud and cross-border cyber-risks?

He asked if the minister

“…could explain why the Government are moving towards remote-by-default testing when other high-stakes assessment bodies are moving in the opposite direction.”

There is quite a lot more from both members, so do check out the transcript.

Cambridge on Remote Testing

There is a new editorial in the PIE by Cambridge’s Francesca Woodward about remotely delivered English tests. One senses a bit of a tonal shift from our friends at Cambridge University Press & Assessment. The message now seems to be that remote testing is fit for high stakes purposes like immigration – so long as it meets the right conditions. Which is something I think every provider of remote tests would agree with.

Thankfully, Cambridge has shared those conditions with us. According to their team, remote testing is fit for purpose when:

1. It tests what matters.

2. It has rigorous security.

3. It has standardized security.

4. It includes human involvement.

5. It is fair and inclusive.

6. It is comparable to other delivery modes.

Again, I think all of the current providers would agree with this. The folks at Duolingo English Test, perhaps, could publish the exact same editorial, without changing a single word.

The tone here is certainly a bit different from what Cambridge presented to parliament back in January. And to some of the comments that were leaked to the BBC back in March.

Today in the PIE we have:

“Remote testing certainly has its uses. Indeed, Cambridge offers some forms of remote assessment. But when the stakes are high – for instance, when testing will be used for immigration – standards of test integrity must be equally high. It is critical, then, that the assessment meets the right conditions to ensure it is valid, secure and fair.”

Compare that to January, when we read (in CUP’s submission to Parliament):

“It is essential that testing for immigration purposes delivers both quality of assessment and the highest security standards. This means a fully-remotely proctored approach cannot be appropriate when the security and integrity of the immigration system is at stake.”

More Testing in Canada?

Immigration News Canada has a nice update on likely changes to Canada’s International Mobility Program.

As regular readers might recall, it was announced that the IRCC might start requiring certain applicants to this program to submit a language test score (currently only applicants for the post-graduate work program stream and the francophone stream submit scores).

By way of an update, consultations are now completed. A first draft of the rules will probably be published by the summer. That will be followed by thirty days of public comment. The changes could be implemented in early 2027. Or later (or not at all).

This will be of interest to those who follow the business of English tests. Mainly because it could be really good for business.

There were about 153,000 new arrivals under the International Mobility program in 2025. There were more than 434,000 extensions issued to existing permit holders between January and September of that year. That’s a lot of potential test takers.

IRCC hasn’t indicated which IMP streams might require an English test. Nor have they indicated whether or not extensions will require an English test. The impact of the changes may be smaller than some have speculated. But you can see how this could very well create tens of thousands of new customers (out of thin air) for the testing companies.

The IRCC currently accepts English test scores from IELTS, CELPIP and Pearson. ETS, meanwhile, has been trying to crack this nut for some time.

ETS Gets a new CEO…

The Educational Testing Service (ETS) has a new Chief Financial Officer. That position is now filled by Ruma Mukerji, who joined ETS in 2024. Prior to that, they held positions at Stanley Black & Decker and PepsiCo.

The outgoing CFO is Jeffrey Melnick, who joined ETS in 2022.

…and Board Changes

The composition of the ETS Board of Trustees changed in May.

Dr. Earl Lewis (University of Michigan) no longer serves as Chair. That position is now filled by Ms. LaVerne Srinivasan (formerly of the Carnegie Corporation of New York).

Mr. Youngsik Chi (Elsevier, Inc) and Dr. Mildred Garcia (California State University) no longer serve as co-chairs. Those positions are now filled by Mr. Kevin M. Guthrie (ITHAKA) and Dr. Barbara M. Jenkins (Chiefs for Change).

Dr. Lewis, Mr. Chi and Dr. Garcia will remain as ordinary trustees.

Exiting the board are Dr. Darienne Hudson (United Way for Southeastern Michigan), Mr. Jeffrey A. Sine (the Raine Group) and Dr. Beverly Daniel Tatum (Spelman College). Joining the board are Dr. Michael Feuer (The George Washington University) and Dr. Christina Grant (Harvard University).

The change at the top is worth dwelling on for a moment. Given his introduction to Nicholas Lemann’s excellent “Higher Admissions: The Rise, Decline, and Return of Standardized Testing” it would be reasonable to refer to Dr. Lewis as something of a standardized testing skeptic. That skepticism could be a valuable resource at a time when ETS seems to be enthusiastically promoting a wide expansion of student assessment in America through its “Skills for the Future” program (in partnership with the Carnegie Foundation).

TOEFL Pod on “Take an Interview”

I like this new(ish) podcast conversation (embed below) between John Clark and Spiros Papageorgiou. There’s some interesting talk around 19:00 about the “Take an Interview” task on the new TOEFL. Spiros talks a bit about some of the shortcomings of traditional interview tasks involving a human interlocutor. In tasks like this, the “human factor,” it is said, can impact how test takers perform. Actually, some rather strong words are spoken on this topic. It all makes for an interesting discussion, especially at a time when IELTS itself is publishing articles about the benefits of including AI scoring and AI chatbots in high stakes assessments.

You know, I spend a lot of time hanging out in the online spaces frequented by test takers (social media networks other than THIS one). I see frequent complaints about human interlocutors – test takers talk about how they were rude, how they interrupted too much or simply how they seemed annoyed by the whole thing. The usual response from test prep professionals is to pooh pooh the complaints, and insist that the interlocutors are highly trained professionals who wouldn’t possibly risk their jobs by being less than perfect. But this response misses the mark to some extent – even if the interlocutor performs according to spec something like an ill-timed frown or heavy sigh may well impact how the test taker responds to their questions. When a test is administered in this way 3.5 million times a year, standardization seems something like a pipe dream.

As I write this, a memory from the old days comes back to me. Many years ago test takers in China would pile into a certain test center in Xi’an because they all thought the nice old British lady who did the interviews there was a real sweetheart. Perhaps they were on to something.

TOEIC Report on Test Takers

The 2025 “Report on Test Takers Worldwide” for the TOEIC Listening and Reading test is now available.

The report suggests that 3,024,969 took the test in 2025. That’s down from 3,229,884 in 2024.

Historical reports suggest that the TOEIC L&R is a test that is steadily declining in popularity. I think the test peaked at about 5,226,536 total test takers back in 2014.

The most specific numbers we’ve got are from Japan, which is probably the test’s biggest market. There were about 1.8 million test takers there in 2025. That’s down from a high of about 2.6 million in 2015.

Specific numbers are no longer available for Korea (probably the second biggest market), but anecdotal evidence suggests a similar decline in volume. Market share has been lost to other tests in the low- and medium-stakes categories that can be used for similar purposes (especially the selection of public officials). G-TELP, for instance, has become quite popular in recent years.

Scored TOEFL Speaking Responses

The TOEFL website now includes a few scored speaking responses. I think they were added on the weekend.

We can now hear “take an interview” responses (to the same question) with grades of 3, 4 and 5 (out of 5).

Same for the “listen and repeat” task – we can hear three three different repetitions of the same sentence.

How ETS Detects Similar Speaking Responses

A new research report from ETS describes how “similar speech responses” are flagged in the TOEFL. “Similar” means responses that are overly similar to those of another test taker, or overly similar to a given question prompt. I’ll post a link to the report in the comments.

Basically, test taker responses are transcribed using speech recognition software (AWS transcribe) and run through three similarity checks. Responses that exceed a certain threshold in all three checks are flagged for human review. The particulars of each check are described, but remain beyond my comprehension.

If my understanding is correct, in the study at hand 29% of flagged responses in the independent speaking task were subject to cancellation after human review, as were 48% of the flagged integrated task responses.

As you may have guessed from the terms above, the article summarizes a study carried out during the period of old TOEFL iBT. Actually, it was carried out during the period of the old-old TOEFL (even before the 2023 revisions).

This is relevant because answer similarity, I think, was probably more of an issue in the old TOEFL than it is in the new TOEFL.

The central conceit of the speaking section of that test was that it required the use of integrated skills and asked test takers to combine details from multiple information sources. Accordingly, test takers around the world were trained to use the same “scaffolding” to structure most of their responses (“According to the reading… the lecturer illustrates this concept with a couple of examples… First, he notes…”). This situation may have been exacerbated by a lack of variety in the structure and flow of those sources (recall the announcement and matching conversation in Q2, for example). On top of that, test takers would inevitably paraphrase the sources in similar ways.

The old independent speaking task was a bit better, but false positives are easy to imagine. For instance, there are only so many ways to say that you prefer to work on projects in groups than alone. If you ask 50,000 people this question you are bound to hear some eerily similar responses even when everyone is on the level.

The new test is probably a bit better. There are no integrated questions and no information summarization questions, so scaffolding is less of an issue. And items in the “take an interview” task seem to have more variety and nuance than items in the old “independent speaking” items. On top of that, the reduction in prep time (0 seconds vs 45 seconds) on the new item turns it into more of a stream of consciousness thing than an overly rehearsed thing.

The design of the new test suggests other issues, of course. I can’t help but think that ETS will need to hire raters to flag every test taker response as on-topic or off-topic, as I think Pearson now does on the PTE. But that’s something for a later discussion.

Pipplet Shut Down

Pipplet shut down in December. I didn’t notice until now because the website makes it look like everything is currently business as usual.

Pipplet offered English assessment services for businesses, mostly in the form of 30 minute tests taken remotely. They had a general English test called “Pipplet FLEX” and assorted tests targeted at businesses in specific industries. They also assessed other languages.

Notably, the company was acquired by ETS in 2021. That was one of the first investments made by ETS Capital, ETS’s investment arm. The shutdown is described on the Pipplet website as being “part of a broader strategic realignment within ETS to streamline and focus its services.”

Baptise Derongs, the founder of Pipplet, wrote about the shutdown over here.

IELTS and “Automarking”

There’s a new article from the IELTS partners about AI scoring in language assessment called “What automarking means for language test validity and integrity.” It addresses some potential strengths and weaknesses of that sort of thing.

This may be the most pro-AI thing I’ve seen out of IELTS in some time.

Interestingly, much of the article is about systems that can be used to overcome the limitations of AI scoring. The author notes that “…best practices mitigate risk, but don’t eliminate it. So, for the time being, a hybrid approach is essential to ensure scoring efficiency and reliability – and safeguard test validity.”

It is also mentioned that “[t]he priority remains to use a hybrid model combining automation with a high degree of human oversight and control – so scores accurately reflect candidates’ true abilities.”

Notably, IELTS does not employ a hybrid model, as it does not utilize AI at all. But when this article is read alongside last month’s article (“Exploring the potential of conversational AI in speaking assessment”) one can’t help but think they might be seriously looking into the possibility of doing so.

No “Writing on Paper” for China

I reached out to the British Council and the NEEA regarding the availability of the IELTS “Writing on Paper” option in China. Both indicated that it will not be offered in that market.

Meanwhile, I talked to my colleague M, who helps individuals prepare for the IELTS in a large Chinese city. M is employed by one of the big test prep firms, so has a pretty steady churn of clients. About 30% of his students prefer the paper-delivered test. He says that the computer-delivered test is popular among university students, working professionals and other adults… but high schoolers are dependent on the paper option as they don’t really know how to type. In his words: “I’ve had high schoolers come to my class and use their index fingers to poke at each key… they are only 16 years old, but they type like my granny.”

A number of his students have changed their schedules and will take the IELTS earlier than originally planned to avoid being inconvenienced by the switch to the computer-delivered format in September.

IDP Education Shares Collapse

I know I promised to stop writing about stock prices, but IDP Education’s ticker price plunged 20% a few days ago after its shares were downgraded by Macquarie Group.

They reached a low of $2.00 AUD following that plunge. That’s down from a high of $38.88.

Among other things, Macquarie cited “soft demand signals for English education as evidenced by lower Google searches for the IELTS test across India and China.”

On top of that, I’d invite readers to check out some recent “Google Trends” reports for IELTS in Canada. And then check out the same reports for CELPIP in Canada. While interest in both tests has declined, they are now neck and neck. That’s a rare thing in a country that IELTS has long dominated. Perhaps it can serve as a lesson that a testing company must always meet their test takers where they are. If it doesn’t do that… someone else will. Ultimately, there are consequences for testing firms that lose track of how test takers feel and what they want (and expect) from the tests they take.

British Council to Sell HQ in Madrid

The British Council will sell its headquarters in Madrid, according to The Guardian. Staff have taken to the streets in protest.

This comes about a year after the BC announced the sale of its flagship school in Madrid for about 50 million GBP.

The beleaguered IELTS co-owner continues to struggle with the repayment of a large loan taken out during the pandemic period.

The Guardian’s report indicates that a building in Barcelona may also be sold.

British Council Staff Strike

British Council staff in Italy participated in a one-day strike on May 21 and did so again on June 4. This comes in response to a formal notification that the BC will eliminate 108 (of 130) positions in the country.

I’m not sure how the layoffs will impact delivery of the IELTS test in Italy. A report in the Guardian notes that “[i]t is understood its exams division would continue under partners,” while the union’s strike notice refers to “l’esternalizzazione delle attività di certificazione della sua conoscenza.” Specific details, though, are not yet available.

Memories of New Jersey

The British Council isn’t the only test company looking to offload some real estate. As I’ve reported in this space, the Educational Testing Service is seeking to rezone part of its New Jersey headquarters to residential… with plans for about 800 homes.

A former neighbor of ETS spotted some of my writing on this topic and sent along a recollection:

“I grew up next to their driveway on Rosedale Road. Mr. Chauncey was alive and in charge in those days. He was very kind … the front door of his house, (now a guest residence), was always unlocked and he allowed the neighborhood kids to come take hockey sticks and pucks from his front hallway closet! The lone security guard in the original building would let us enter and wander to a water fountain on hot evenings when we would ride our bikes around the complex. Yes, a Norman Rockwell childhood!”

Mr Chauncey is Henry Chauncey, founding president of ETS. He ran the organization from 1947 to 1970.

Duolingo Lobbied IRCC

The Globe and Mail reports that Duolingo lobbied IRCC to have the Duolingo English Test accepted for immigration applications, but that the firm “was ultimately unsuccessful in its attempts.”

According to the report, back in 2025 Duolingo attempted to get its test accepted for the Economic Mobility Pathways Pilot (EMPP), which granted permanent residency to certain highly-skilled refugees. There were also discussions about using the DET “for federal immigration programs more broadly.”

These overtures were rebuffed.

The Globe’s report describes the case of one Hamdan Albuhaisi. A resident of Gaza, Albuhaisi applied for Canadian residency via the EMPP program shortly after 11 of his family members were killed in an Israeli airstrike. As the last traditional test centers in Gaza closed in 2023, his application included DET scores. That application was rejected by IRCC.

The EMPP program has since been terminated.

Duolingo on Security

This post on security over on the official Duolingo blog seems like a response to a list of questions I posted on Linkedin. It’s nice to know that someone is reading my stuff.

On Fairness and Proctoring

There is a nice article by Jieun Kim in TESOL Quarterly which explores how the language spoken by remote proctors (of English tests) impacts test takers. Researchers created a remotely proctored test using reading and listening questions from the IELTS test and administered it 66 times either with an English-speaking proctor or a proctor who spoke the test taker’s first language (Korean). Scripted disruptions were included in each administration.

The article notes that “proctoring language (Korean vs. English) did not have a statistically significant effect on test scores across reading, listening, or overall scores.”

It also notes that “on average, participants took longer to report the technical difficulties when proctored in English compared to Korean” and “participants in the English-proctored condition demonstrated greater difficulty in communication, as indicated by the number of conversational turns required to resolve the technical issue.”

A few test takers just gave up communicating with the proctors. That made me chuckle.

Not surprisingly, lower-proficiency users required more turns to resolve problems. And most test takers (from both groups of test takers) expressed a preference for Korean-speaking proctors.

In interviews, a majority of test takers expressed a preference for test-center administrations over remote administrations for reasons of fairness, security and privacy.

A few thoughts come to mind:

1. It is worth remembering that Korea has really (really really) nice test centers. Anecdotally, I’d say that Korea has the best test centers in the world. There is also nice geographic coverage for major tests. Test takers from other countries might express different preferences when it comes to test delivery.

2. When I registered for the remote LANGUAGECERT test in December of 2024 I was able to choose which language my proctor would speak (from eight different options). Sadly, it seems that choice is no longer available.

3. I’ve written before about live proctors being on the way out. I predict that more tests will eschew this mode of administration in favor of “record and review” options.

4. Test takers are right to be concerned about their privacy. Every now and then a remote proctor will use a test taker’s personal information to engage in cyberstalking via social media accounts. Beyond that, reports abound of room scans that have descended into terrifyingly creepy invasions of privacy.

I Took VTest Academic

I took the VTest Academic from Burlington English. For some notes about my experience, you can consult the blog. I also took the VTest General last month.

ATP Panel on Test Security

On June 24, the Association of Test Publishers will host a panel discussion called “Security Across Exam Modalities: Are We Having the Right Conversation?”

It will focus on “the ongoing debate between remote proctoring and traditional test center delivery.” Participants will be Basim Baig from Duolingo, Claire McCauley from Cambridge University Press & Assessment and Wallace Dalrymple Jr from ETS.

The email I got says…

“[r]ecent moves by major organizations to withdraw from remote proctoring as a modality option for test takers is creating a lot of discussion in the assessment industry regarding the effectiveness and security of different testing modalities. This session moves beyond a simple comparison of modalities and instead examines a more fundamental question: how should assessment programs design and apply layered security controls to protect integrity in an increasingly complex threat environment?”

Register here. If you live in a peculiar time zone like me and can’t attend, don’t worry – a recording will be sent to all registrants.

New Book Coming from Routledge

This winter, Routledge will publish “The Routledge Handbook of Digital Language Assessment.” Members of the Duolingo English Test team will contribute to the book.

Photo of the Week

I got this cool postcard showing part of ETS headquarters. I think it is from the 1960s.

IDP Education has opened eight new IELTS test centers in China, bringing the total number to 13.

I’m now seeing the following locations:

• Beijing x1

• Shanghai x1

• Jiangsu x3

• Zhejiang x2

• Shandong x2

• Hubei x1

• Hunan x1

• Guangdong x1

• Chengdu x1

Test dates appear to be available for as early as next week, but I didn’t click all the way through to the final purchase.

Interestingly, the price of taking the IELTS with IDP seems to be 100 RMB ($14) less than what the British Council charges.

Paper-Delivered IELTS to End Soon in China

The British Council has announced that the paper-delivered IELTS will come to an end in China on September 1 of this year. After that date, only the computer-based IELTS will be offered.

IELTS for UKVI will go entirely computer-based in that country even earlier. That switchover will happen on July 1.

The termination of paper-based testing worldwide was announced a few months ago, but this is the first time a specific date has been given for the key Chinese market.

Curiously, the announcement for China doesn’t mention the “writing on paper” option (take the test on computer, write your essay on a sheet of paper) at all. IELTS has announced that such a testing mode will be available in some markets… at some point.

Paper-based testing is more popular in China than you might think. Indeed, it is actually pretty popular all over the world – while zoomers possess many well-documented talents and skills, they aren’t really known for being proficient typists.

More Deets on IELTS “Writing on Paper”

Actually, the IELTS partners have published a description of the upcoming “Writing on Paper” delivery method. I’ll link to it in the comments, but in the meantime here are a few facts:

1. Writing on Paper will be available for both IELTS Academic and IELTS General Training tests.

2. Writing on Paper will not be available for IELTS for UKVI tests.

3. Writing on Paper has not yet been approved for Australian visas.

4. The IELTS Test Results form will not indicate that the writing section was done on paper.

5. One Skill Retake is available as normal, but when the writing section is retaken it must be done so via the original delivery method.

6. Timing is the same as in the wholly computer delivered IELTS.

7. This format will be available “in select countries only from mid-2026.”

CUP on Remote Testing

A new paper from Cambridge University Press & Assessment says that:

“…the use of scores from remotely delivered assessments for high‑stakes decisions should be considered with particular care and should not prematurely replace established in‑centre delivery modes. Ultimately, where an evaluation of the advantages and limitations of remote delivery indicates that its risks cannot be sufficiently mitigated for a given highstakes context, in-centre testing should remain the preferred mode of delivery. Remote delivery should only be adopted once its security and operational challenges can be addressed to a level that ensures it is fit for the intended purpose.”

This is a refreshingly reasonable and nuanced take on the topic from CUP.

(Frustratingly, though, it did lead Times Higher Education to trumpet “Remote English language testing ‘not yet fit for purpose’” in a headline that seems to miss the point of the research.)

That said, I’m left wondering how Cambridge approaches this topic in conversations with IELTS score users. The fully remote IELTS Online has been used since 2022 for high stakes purposes (especially in Canada and the USA). Including the issuance of study permits and visas. How does Cambridge respond when they are asked if the test is fit for those purposes?

IDP Education Shares are Down

As I write this, shares of IDP Education are trading at $2.71 a piece. That’s a 93% climbdown from IDP’s all-time high of $38.88, achieved in 2022.

It also appears to be lower than the closing price on IDP’s very first day of trading (November 27, 2015), which Google tells me was $3.40.

I don’t think we’ve completely exited the “monopoly era” of English language testing for high stakes purposes, but we’re getting there. In any case, we can check off another box.

In any case, I promise I will stop reporting on share prices now. I mention this new low only because it means IDP now has a market cap of about 546 million USD. Compare that to the 500 million USD purchase price that ETS was reportedly seeking for the TOEFL and GRE (per reporting by the Wall Street Journal).

IELTS on AI in Speaking Assessment

New from IELTS is a short article which explores “the potential of conversational AI in speaking assessment.”

Nothing really earth-shattering here, as it mostly summarizes longer stuff they’ve put out over the past little while. I liked this part at the end, though:

“…traditional rule-based SDSs, as opposed to their AI-powered counterparts, therefore seem to be a more practical choice in the short term, given their greater consistency, scalability, and control over test delivery.”

It is mildly interesting that IELTS can only bring itself to commit to traditional human interlocutors in “the short term.” One of these days, I think, they’ll announce a big change. Don’t say that Old Man Goodine didn’t warn you.

A silly thought experiment might be to estimate how much it costs to maintain the current method of speaking assessment. It certainly isn’t cheap – the global average is probably close to $12 USD in examiner wages per test. On top of that are the costs of payroll taxes, training, recruitment, management and physical infrastructure. The proposed AI model would still involve ratings by humans, but offshoring that process could bring the cost down to just a few bucks per test. Or less. The partners would love it – I can’t help but think such a move would be just as lucrative for IDP Education as, say, opening a bunch of new test centers in China.

On Resetting Accessibility

Cambridge notes that:

“Despite the assertion that at-home assessment democratises access to testing, in practice it often fails to provide the reliable infrastructure that test takers need – from adequate equipment and stable internet connection to consistent test conditions and appropriate security.”

Of course, we can say the same thing about test centers.

I’ve been to many test centers over the years. So, so, so many test centers. Their quality has always been inconsistent, often in ways that intersect with class and privilege. It isn’t uncommon for a test provider to have an ultra-lux option in a high-traffic downtown location, and a shabbier option in a poorer part of town.

One test center might have shiny new equipment and white noise machines, while another might have sticky keyboards and traffic sounds from outside.

I’m talking about the same test, the same test provider… even the same city… but inconsistent quality.

I think this leads to what they call construct-irrelevant variance.

And, needless to say, we can easily find even bigger differences in test center quality when we cross international borders.

To me, score users who believe in accessibility will accept both remote tests and test center options. This empowers test takers to choose the option that is best for their unique circumstances. Whether a particular test maker provides both options is unimportant when test takers have a range of options to pick from.

Movement on TOEFL Essentials in Canada?

Back in August of ‘25, it was announced that the TOEFL Essentials test was “recognized” by IRCC. Most observers assumed that recognition would be followed by acceptance of the test for migration to Canada. And we all celebrated. Alas… seven months have passed and the test has yet to move past the recognition phase. It is still not accepted by IRCC for any purpose.

However, I have spotted some movement on this front. According to a recent LinkedIn post, ETS is currently seeking to hire a “Business Development and Immigration Director” for Canada.

Says ETS’s new job listing:

“We are seeking a Director of Immigration Operations to own the execution and delivery of ETS’s immigration positioning in Canada. This role is accountable for ensuring TOEFL and TOEFL Essentials are effectively positioned, operationalized, and sustained across relevant immigration pathways, while enabling channel and strategic partners to support adoption within agreed commercial frameworks.”

I’m left wondering if ETS might pitch the standard TOEFL iBT as a test for immigration to Canada. As regular readers know, it is pretty much the same test as TOEFL Essentials.

In any case, testing for immigration to Canada has become more competitive in recent years. IELTS enjoyed something of a monopoly for many years, but that has been eroded by the folks at CELPIP, which is a surprisingly popular test among people already in the country. Pearson’s PTE Core is growing as well.

TOEFL Specs

Now available is a “Test Blueprint and Specifications Document” for the new TOEFL test.

It provides more information about the number of items on the test and how they are designed.

Our Sponsor: Prepex.AI

This issue of the newsletter is sponsored by Prepex, which is a TOEFL test preparation platform. It includes AI scoring of both the speaking and writing sections of the test. I think it’s pretty good, and have contributed some content to it over the past few months. If you are helping someone prepare for the new TOEFL, do check it out.

You Can be the Next Sponsor

If you want to sponsor a future issue, do reach out. Your message will find its way to the 400 coolest people in the English assessment space, all of whom subscribe to the newsletter. And you’ll also reach another 400 readers who don’t subscribe (and are much less cool).

New TOEFL Podcast

There is a new podcast from ETS on Youtube called “TOEFL Sessions.” Hosted by John Clark, there are currently a couple of episodes featuring Spiros Papageorgiou (discussing TOEFL‘s listening section) and Larry Davis (discussing TOEFL’s writing section).

I really enjoyed hearing Larry talk about some of the thinking that went into the design of TOEFL’s “Writing for Academic Discussion” task.

Here’s an embed of the first episode:

TOEFL: Validity by Design?

New on the official TOEFL blog is an article called “Inside the TOEFL iBT Updates: Validity by Design.” It explores some of the thinking that went into the recent TOEFL relaunch. Like some other recent missives from ETS it seeks to justify the inclusion of some shorter speaking and writing tasks. It’s all useful stuff if you want to know more about the new TOEFL, so check the comments on this item for a link.

It’s mildly frustrating that this content is buried on a blog full of posts about Tina’s low-fi study beats and “8 Tips to Prepare for Life at University.” As I’ve said here before, test makers need a channel where they can share a steady drip of meaningful articles that explore the thinking behind their products. IELTS does this with its “news and insights” page and Duolingo does this with the “Duolingo Test Center” blog. They do a good job of keeping the meaningful content away from the fluff and the SEOified junk.

The audience for such a channel isn’t huge, but it does include a great number of people that test makers need to influence. And, needless to say, LinkedIn isn’t the place to reach such folks.

No More Free TOEFL Paper Score Reports?

It appears that TOEFL no longer provides free paper score reports to test takers. I only see the option to pay $35 for express shipping… even after changing my address to a bunch of different countries. Worse yet, I think it used to cost $25 for express shipping.

That’s too bad. But on the plus side, it appears that ETS no longer charges test takers a hefty fee to update their address on the TOEFL website.

ETS Annual Audit Now Available

The 2025 audit of the Educational Testing Service is now available from the Federal Audit Clearinghouse. It covers the year ending September 30, 2025. If you can’t find a copy via the search engine just lemme know.

A few notes (all amounts USD):

1. Operating revenues are 1.02 billion. That’s a drop of 12.1%. I don’t think operating revenues have been this low since 2010.

2. In any case, I’m more interested in program service revenues (that is, revenues from tests), but we’ll have to wait ‘till the 990 form is published in the summer.

3. ETS finished the year with an operating loss of 60.7 million dollars. That’s less than 2024, when its operating loss was 146.8 million dollars.

4. Total assets are 1.67 billion. That’s down 6.7%.

5. Net assets are 1.34 billion. That’s down 4.3%.

6. One interesting thing about ETS’s assets is that they include quite a lot of goodwill (valued at 261 million) and intangible assets (valued at 194 million). That’s a lot more than before. These things seem to be a result of the purchase of PSI Services. If we go back to just 2023, goodwill was valued at 26 million and intangible assets at just 6 million. I don’t know if this means anything.

7. The organization’s restructuring is starting to show up in the audit. Operating expenses for the year were 1.08 billion, a drop of 17.6%. This is mostly the result of less money being spent on wages, test centers, and subcontracted services.

8. 13% of ETS’s revenues came from the new College Board contract. The old one accounted for 30% of revenues.

TOEFL (old) Compared to IELTS

ETS has published “Beyond Score Correlations: A Content Comparison of IELTS Academic and TOEFL iBT® Tests in the Context of a Score Concordance Study.”

The report compares the content of the IELTS (Academic) test and the old TOEFL that was discontinued in January. It is a good read for folks who are interested in this sort of thing, but perhaps not so useful for score users in 2026.

ETS/TED/KHAN Launch Institute

ETS, TED and Khan Academy have launched the KhanTED Institute. Details are sparse at this time, but the ultimate goal seems to be the creation of a degree-granting institution.

Once accredited, the institute will issue undergraduate degrees which it describes as competency-based. Says the website:

“Move forward by mastering skills, not by just logging class credits. Emphasis is on developing, applying and demonstrating competency. This means motivated learners can complete the program faster, and everyone graduates with verified competencies.”

The institute will begin taking applications from students sometime in 2027. A full degree will cost under $10,000.

It is an interesting initiative. I suppose the trick will be for the partners to distinguish themselves from organizations that are already successful in the competency-based learning space (like, say, Western Governors University).

PSI on Security

Speaking of podcasts, I very much enjoyed last month’s episode of “Tried and Tested” with Isabelle Gonthier. Gonthier spoke to Wallace Dalrymple and Paul Gollash about the security that supports the new TOEFL. As I’ve mentioned here before, there is a huge amount of interest in the new test, and I’m glad that ETS is being a bit more proactive when it comes to providing new information.

A snippet of conversation around the 14:00 mark really caught my attention. Specifically:

“Would you ever get rid of a control? Yes, when a control is no longer providing the measured outcome for the trust in the integrity for our customers… well then we retire the control and replace it with a new control […] maybe it’s just a control that provides multiple layers of value.”

The episode doesn’t go into detail about controls that have been (or might be) retired, but I can’t help thinking about the debate around asynchronous vs synchronous proctoring (by which I mean human proctoring after the test vs human proctoring during the test).

I was honestly a little surprised when I learned that the new TOEFL would retain live human proctors, as I fully expected the test to switch to an asynchronous system.

Readers focused on the UK scene might not realize how popular “async” proctoring is in the USA. This is the method used by the Duolingo English Test, of course, but it is also the method Pearson chose for the new “PTE Express” test. And it is also used by other USA-focused test-makers like iTEP International and English3.

Those of us who have taken a lot of at-home tests realize that live proctors are useful… but perhaps of limited value. They often seem to be tasked with watching quite a few tests at the same time and reports of test takers waving their hands around for 5+ minutes to get a proctor’s attention don’t inspire confidence in their ability to spot cheating attempts on the fly.

Compare this to a system where each test administration is recorded and reviewed by a human after the test has been completed. Better yet, a human who watches only one video at a time.

A security scheme that makes use of both systems is ideal, of course. But I worry about tests that use live human proctoring as an excuse to not have 100% of all administrations reviewed by human eyes after the fact.

I know there is a strong case to be made for proctors as an onboarding aid, but to me that seems like more of a customer experience thing than a test security thing.

Anyway. I will be so bold as to predict that TOEFL goes async by the end of 2026.

Enjoy an embed:

TOEIC Link Technical Manual

There is now a technical manual for the new “TOEIC Link” test from ETS. Some might find this to be interesting reading because the Link test is quite similar to the current version of the TOEFL test.

For instance, the TOEIC Link’s writing section contains the following tasks: “build a sentence,” “write a brief e-mail,” “write an extended email” and “write about your perspective.” The “perspective” task seems to be a re-skinned “write for an academic discussion” task.

There are other similarities, including a “listen and repeat” task in the new TOEIC speaking section and a two-stage adaptive design.

The TOEIC Link test remains a peculiar product. It hasn’t replaced the TOEIC that we all know and love – that product is still available in most of the world. The TOEIC Link, on the other hand, isn’t so easy to take. It can’t be taken in Korea and Japan, which collectively make up about 80% of the current worldwide market for TOEIC tests. Those countries are still exclusively using the classic TOEIC. Nor can it be taken in France, another key market for the classic TOEIC.

I guess it is available as an institutional product in certain countries in the Middle East, but it all seems like small potatoes compared to how the classic TOEIC is currently used.

Duolingo on Security

Last month I attended an online “Ask Me Anything” about security and the Duolingo English Test featuring Basim Baig and John Woodcock. I’m glad I found the time to drop in, as I learned quite a lot. Like:

1. 100% of DET administrations are reviewed by a human proctor after the fact. Ninety percent of tests are reviewed by one proctor, about 10% are reviewed by two proctors, and a small number are reviewed by more than two proctors. The proctors review one test at a time. The review includes watching a video recording of the test session. This info came in response to my own question.

2. Without the need to livestream a video of the test to a human proctor during the administration, the test becomes somewhat more resilient to short connectivity issues.

3. Some sort of firewall is used during the administration of the DET to prevent the test browser from communicating with anyone other than the Duolingo server.

4. Keystroke biometrics are used to identify the presence of proxy test takers who have gotten involved in multiple tests.

Point four is interesting and I would love to hear more about it in the days ahead. Keystroke biometrics have been used to prevent cheating for quite some time, but their effectiveness isn’t fully understood. Indeed, I don’t even know exactly how reliable they are as a personal identifier (compared to, say, finger or palm prints). Perhaps this is something that could also be implemented in the delivery of tests taken wholly at test centers as a way to further reduce the scourge of proxy test taking.

I’m glad that test makers are a bit more willing to talk about this stuff in a public forum. Obviously this builds trust in their products, but it also can inspire competitors to up their game. I remember sitting in on a wonderful presentation about test security at Duolingo HQ in Pittsburgh some years ago. Everyone there was NDAed, but at the time I thought that a sort of reverse-NDA that punished people for NOT spreading the word would be more useful.

One day I’ll compile a list of questions that score users might ask their favorite test makers to suss out whether they take test security seriously. Some of this stuff will be on the list!

Duolingo Wants HOELT

Reported by The Telegraph:

“Duolingo is lining up to run migrant English language tests for the Home Office, The Telegraph can reveal.”

Duolingo Q1 Numbers

Duolingo has released Q1 results for 2026. Revenue from the Duolingo English Test was 11,317,000 USD. If we divide that by a $70 test fee, we can guess that the test was taken about 162,000 times in the quarter.

That represents about a 9% volume decline compared to Q1 of 2025.

Just keep in mind, as always, that these are crude estimates. Some people pay extra to get faster results, some people pay less by purchasing a bundle of tests and some tests are sold to resellers in big batches. And many students pay nothing at all, as they get vouchers through Duolingo’s Access program.

In recent months we’ve learned that the largest tests continue to lose volume. IDP Education reported a 7% drop in IELTS volumes for the half year ending December 31, 2025. And a few days ago Pearson reported that PTE volumes “declined slightly” in Q1 of 2026 (see below).

Pearson Q1 Numbers

Pearson released its Q1 trading update. Regarding the PTE, it says “Pearson Test of English (PTE) declined slightly due to a continued tough market backdrop.”

Oxford Test of English on “Talking ELT”

The latest issue of the Talking ELT Podcast from Oxford University Press is pretty great. In it, Anthony Green and Sara Pierson discuss the Oxford Test of English and Oxford’s recent white paper on English language assessment. I’ll include an embed at the end of this item.

A few bits caught my attention:

1. Around 15:20 there is a discussion about the ability of students to combine information from multiple sources. The white paper suggests that score users value this skill quite a lot.

2. There is a discussion around 18:20 about the “authenticity gap” between what can be tested in a short English test and what students do on campus (and beyond). This has come up quite frequently in recent conversations. I talked about it in London, even.

3. At 23:50, there is some conversation about how score users “don’t value long form essays quite as much as they used to.” And that “the essay isn’t the only game in town” right now. Indeed, we’ve seen some test makers move away from this sort of task.

4. There is some interesting discussion at 36:00 about how test providers should think about the whole experience as comfortable as possible for test takers.

A few thoughts come to mind:

1. It is interesting to hear about how much score users value the ability to combine information from various sources. For twenty years (2005 to 2025) questions that tested this skill were one of the main selling points of the TOEFL iBT. During that period they formed the backbone of the TOEFL’s speaking and writing section. In January of this year they were all removed from the test. One day I hope to learn more about the factors that contributed to this decision. One one hand, people really liked them, and they weren’t particularly time consuming. But on the other hand, they were somewhat imperfect items due to prep strategy and familiarity, and test takers found them challenging.

2. The ability to combine sources is often associated with integrated tasks, but it doesn’t necessarily have to be. The TOEIC, for instance, has included some attractive “triple passages” in its reading section for the past decade. This task might require the test taker to read a business pamphlet, an email to the business, and a sales invoice before answering multiple choice questions that require comprehension of all three. I think these tasks have been carried over into the next-gen TOEIC Link product.

3. Traditional test makers have struggled, now and then, with creating a comfortable testing experience - including during the post-test period. This can be partially explained by the fact that from a business perspective, it didn’t used to matter. It does now, though, since test takers have more options to choose from. One throughline of my writing in this space is how the traditional players continue to struggle in this area.

Anyway. I’m going to take the Oxford test as soon as I reach a place where they operate test centers.

Here’s an embed of the podcast:

MET Annual Report

Back in February (I think) Michigan Language Assessment published the 2025 MET Test Report.

Most interesting to me are the reasons why people take the test. According to the report, at least 26.68% of test takers took it as part of having their nursing credentials recognized. That’s the number one category of test takers. It’s also up 11% since 2024. Regular readers might recall that I’ve written before about MLA’s success in marketing the test to health care professionals.

Just 0.38% of test takers took the test to meet Australian visa requirements. The test wasn’t accepted until August, but that’s still a small number which reminds us how important it is to avoid overstating the impact of regulatory changes.

The report concludes by noting that “[i]n 2025 the number of MET test takers increased.” That’s nice to hear, considering that most of the really large tests lost popularity in 2025.

CAEL. CAEL!

Our friends at Prometric have uploaded a long walkthrough of the CAEL test to YouTube. That’s really great, as we don’t hear a ton about the CAEL test nowadays. I’ll include an embed at the end of this item.

I haven’t actually taken the CAEL under real conditions (but let me know if you’ve got a voucher), though I did complete a couple modules of teacher training back in 2021. It’s an interesting test, and I don’t think there is anything else quite like it on the market.

This could be called a fairly traditional test – it contains a number of very long reading and listening passages (similar to the old TOEFL iBT) and takes a whopping 3.5 hours to complete. I suppose that makes it one of the last of the marathon tests.

It is also traditional in the sense that it heavily (really heavily) emphasizes integrated skills – each of those long reading and listening passages is accompanied by questions that are answered via extended written and spoken responses. It feels silly calling integrated questions an aspect of “traditional” tests… but you know what I mean.

I believe test taker responses are scored wholly by human raters.

Pricing is somewhat traditional as well. The CAEL is cheaper than the TOEFL and IELTS, but only slightly. And test takers can only send five score reports for free.

On the other hand, the test is modern in that it is mostly taken remotely. An annual report from 2024 notes that 80% of administrations were taken online that year. I suspect that the 2025 report (check for it next month) will include a higher number, as in-person test dates seem to be in short supply, at least according to the registration system on the CAEL website.

So there ya go. Some CAEL content in 2026. The test is widely accepted by schools across Canada.

(A history lesson: CAEL was developed at Carleton University in the 1980s by Janna Fox and others, was sold to Paragon – a subsidiary of the University of British Columbia – in 2015, and Paragon was purchased by Prometric in 2021)

Would you like an embed?

CELPIP for Uni?

I wrote that item about Prometric’s CAEL test a few days ago. A few hours later it was pointed out that Prometric is currently seeking to hire an academic relations manager for the CELPIP test. The successful applicant will “develop and deliver a test recognition program for the CELPIP across higher education institutions in Canada and internationally and work to increase CELPIP volumes for admissions purposes.”

Traditionally, CAEL and CELPIP have served distinct purposes – CAEL has been for admission to higher ed, and CELPIP has been for migration.

Obviously, though, Prometric is hoping that CELPIP will someday be used for both purposes. Indeed, a few weeks ago it was announced that Lambton College had begun accepting the test.

CELPIP is, of course, a test of general English without much “academic” content. But perhaps that doesn’t matter as much as it did in the past.

Will CAEL have much of a future if CELPIP gains widespread acceptance at Canadian universities? Maybe. But…

I’ll share a link to the job posting in the comments below.

(Elderly readers might recall that for many years there actually was an academic version of the CELPIP test that was used for admissions. It was discontinued after Paragon purchased the CAEL in 2015.)

On Validity and Assessment Literacy

New in Language Testing is an article by Vahid Aryadoust. It proposes two measures to enhance transparency of a test. Specifically:

1. Validity reports for individual test forms.

2. Clear explanations of the test’s limitations.

Number one seems challenging in a world where AI, adaptive testing and security concerns have changed what we mean when we talk about a test form.

Number two seems obvious but useful at a time where tests are increasingly adopted for off-label uses. We’ve all seen instances of test makers trying to recapture lost volumes by proposing their test for some purpose it hasn’t heretofore served (and for which some other test may be much more suitable). Meanwhile, stiff competition has caused test makers to play up the strengths of their products while downplaying their limitations.

Questions to Ask About Remote Proctoring

Over on the LinkedIn, I published a list of questions that score users might ask when approached about accepting remotely proctored tests. I sort of like how it turned out. If you want the list but don’t use LinkedIn (smart idea) just sent me a note.

Photo of the Week

This week’s photo is the TOEIC section at the Kyobo Books store in One Grove Mall near Magok Station in Seoul.

Thoughts on PIE Live Europe 26

I had a lovely time at PIE Live last month. I do recommend the PIE events to anyone seeking to expand their knowledge of the international education sector and to connect with some fairly interesting folks. And if you are just interested in tests? Good lord, do the test companies have a presence. There will be an event in Australia in July… or you can wait ‘till April of 2027 for the next London gathering.

A few stray thoughts:

1. I gave a talk about the state of testing in 2026, and I think it went pretty well. There was a big crowd, I said everything I wanted to say, and got a decent amount of participation from the audience. I signed up for this in part to get out of my comfort zone, so I’m happy to have done a good job. Regular readers might argue that talking about English tests is precisely my comfort zone, but you know what I mean…

2. I can’t mention all the excellent presentations, but I will highlight “Kathmandu or Kathmandon’t? Is Nepal a viable recruitment market?” That’s a market I should know more about. I’m always happy to learn about the scene in specific countries. There were a few interesting comments and questions about proof of language proficiency.

3. Walking around and around the exhibition space I was able to chat with kind folks from just about every testing company. And a handful of off-site meetings took me all over London (and beyond). I learned an incredible amount about these assessments, and collected enough literature to keep me learning for some time to come.

4. I learned a bunch about the new ISE Digital Test from Trinity College London (a blind spot because of its newness), about the new PTE Express, about the new TOEFL… about all the tests!

5. Notwithstanding recent changes to the TOEFL, I’m starting to get the impression that integrated questions are on the rise.

6. I went to Oxford to gain a bit of knowledge about the Oxford Test of English. That’s been a real blind spot for me since it is only taken on-site and they don’t have any test centers in Korea. And, as a somewhat new-ish member of the “long tail” of upstart tests, it hasn’t been covered in the press as much as other products. But it is backed by an impressive depth of talent that few testing companies could ever hope to match. That may set it apart from the pack. We should all keep an eye on it.

7. Oxford is a pretty town. The realization that I would not be able to remain there for the remainder of my days filled me with some wistfulness.

8. I actually recommended a couple of test centers in Korea that the Oxford test might make use of. But a later meeting with some folks who are actually in that business showed me that there is a lot more to the industry than I am aware of.

9. If you go to the British Library you can peer at the handwritten manuscript for Mrs. Dalloway. How about that?

IELTS Partners Quit HOELT

The IELTS partners have now confirmed via a post on their website that they have withdrawn from the HOELT tender.

The post indicates that…

“[d]ecisions as significant as granting visas to live in the UK require the highest trust in assessment outcomes. Evidence from research and regulatory practice shows that fully remote testing presents challenges in meeting these standards consistently – especially in the highest stakes environments where security is paramount.”

I’m interested to learn more about this decision, as the above statement is somewhat at variance with the fact that the IELTS consortium is currently one of the largest providers of fully remote tests for high-stakes purposes, including for admissions to universities in the UK and elsewhere. As regular readers know, students who take the IELTS (academic) have the option to forgo their local test center and instead take the test from the comfort of their own homes.

It is wise of the IELTS consortium to step away from the tender if they feel incapable of creating a sufficiently secure testing environment. That said, there may be other organizations which are more capable.

When I talk to clients about this sort of thing, I often explain it by referring to “generations” of remote testing technology. Generation 1 was characterized by tests that were delivered in regular Internet browsers with pretty basic security (like proprietary plugins and pop-up warnings when the test taker’s cursor strayed too close to the side of the window). Webcam access let a proctor on the other side of the world peer at a dozen test takers simultaneously as they took the tests. Generation 2 saw the introduction of custom-made secure browsers. Key features of succeeding generations included things like secondary cameras, asynchronous proctoring, rudimentary deepfake detection… and more. The lines between generations are sometimes blurred, but I think you get my point.

I’m not sure how to number the current generation (5? 6?) and I don’t know exactly what’s coming up, but I’m excited about the possibilities – like how psychometrics and technology can intersect to produce new kinds of test items that serve to mitigate the impact of test taker malpractice.

It is important to note that not every provider reaches the same tier at the same time. In 2026, there are still some tests (including ones that are widely accepted) that run on gen-1 technology. That’s the best the companies behind them can do, and score users often lack the assessment literacy necessary to tell the difference.

I fear that I’ve meandered far off-topic. But the point I am trying to make here is that testing firms don’t all have the same level of expertise and technology at their disposal when it comes to remote testing. The IELTS partners might not be able to securely deliver a a remote test. But it would be wrong to automatically assume that the requirements are, therefore, impossible for anyone to meet.

Me on Security

I was quoted in this PIE News article about the IELTS consortium pulling out of the HOELT tender. I said:

“If the IELTS consortium has determined that they are unable to deliver a suitably secure remote test, their decision to withdraw from the process makes a lot of sense. I haven’t read the letter they sent to the Home Office in full, but it almost seems like they have suggested that no one is capable of delivering a secure [remote] test. I’m not sure I agree with that assessment, as every potential bidder has their own strengths, expertise and technology.”

It could be true that remote testing can never be fully secure. It wouldn’t surprise me if the whole tender gets derailed as a result of the IELTS withdrawal. But it may be more complicated than some coverage is suggesting.

Consider that the IELTS itself can be taken fully remotely. The remote IELTS is used and trusted by universities around the world and has been used to admit tens of thousands of students into higher education. What strikes me when I look at that product, though, is how it lacks some of the security features currently used by other testing firms. For instance, as far as I can tell, it still does not require a secondary camera (but please correct me if this information is out of date). A secondary camera is something that’s been a standard part of remote tests from LANGUAGECERT, ETS and Duolingo for quite some time. In case you aren’t familiar with it, the secondary camera (a phone or tablet) points at the test taker’s keyboard and mouse for the duration of the test. Among other things, it can help determine if the person taking the test is actually answering the questions or if someone else did that via remote access. Basically, you can compare the keys pushed by the test taker with the keystrokes actually recorded by the test software.

While I’m sure that the remote IELTS is completely secure, I can’t shake the feeling that other testing firms have spent much more time working on this issue. They may be capable of stuff that the IELTS partners haven’t yet thought of.

IELTS Paper Elimination Not a Big Deal?

I was also quoted in this great article by Abhishek Nair about the coming end of the paper-delivered IELTS. I think it is really important that we keep an eye on how widely available the “writing on paper” option becomes. Wide availability could certainly allay many of the concerns people have with computer-based testing. If widely available, the change may not be such a big deal.

IELTS and Vietnam

My elaborate system of news alerts keeps sending me articles about how authorities in Vietnam may reduce the impact of English test scores on admissions to Vietnamese universities. Vietnam is somewhat unique in the world, as such scores are used by domestic students seeking entrance to domestic universities. Apparently, a good test score can give applicants a real edge in this competitive process. But some think that it provides too much of an edge and that, perhaps, other factors should count for more.

IELTS seems to have been the main financial beneficiary of this situation, as they do boffo business in Vietnam. But all of the other testmakers have been drawn to the country, hoping to encourage the use of their tests (where recognized). Or to gain recognition in the first place.

Vietnamese readers might comment on the potential impact that regulatory changes might have on the sales of certain tests in the country. I don’t know enough to comment on that.

(Late Edit: IELTS did commission some research on this phenomenon.)

But those interested in this sort of thing might look at domestic admissions in Korea from the late 90s to the early 2010s. In that country, universities once offered a “special admissions track” for students with advanced second language proficiency. Basically, this meant students with a good language test score. The track was attractive because it helped students escape the dreaded Korean CSAT test (to some extent). Perhaps you’ve read about that particular madness once or twice.

In practice, advanced second language proficiency was usually proven by reporting a high TOEFL score. Accordingly, ETS made a lot of money selling TOEFL tests in Korea. According to some press reports, Korea was actually the number one market for TOEFL when the iBT version launched back in 2005/2006.

It was all a pretty big deal. During this period, I worked for a hagwon chain (Avalon English) which specialized in teaching TOEFL prep to middle school students. I’ll resist the urge to write about negative washback of tests at this time. But you get the point. And hey… nowadays I also get sent a ton of articles out of Vietnam about the amazing IELTS scores earned by very young children.

Anyway, the authorities in Korea eventually eliminated this special admissions track (for the most part). They even introduced regulations which prevented students from even mentioning TOEFL scores in their application essays. Very quickly, the number of TOEFL tests sold in the country dropped significantly. Korea is now probably the number three market for the TOEFL, behind China and India. Or maybe the number two market, if what I hear from India is accurate.

I’ve lost track of my point, but I think it is worth keeping an eye on domestic admissions in Vietnam for hints about the volumes of certain tests in the medium term.

IELTS Research Re: Listening

The IELTS partners have published an excellent article by Nicola Latimer, Chihiro Inoue, Samantha Chan and Daniel MK Lam about how learners process lectures that involve both auditory and textual content… and how that sort of stuff could be included in assessments.

The authors note in their introduction that:

“Currently, the listening component in most major international language tests used for university admissions (e.g. IELTS, TOEFL iBT, PTE Academic, CAE) involves the presentation of audio material as the primary/sole source of information for comprehension.”

And in the conclusion:

“We argue that tests of lecture listening should present audio and slide text simultaneously as input, as real-life academic lectures would. The input would then need to cover a range of degrees of integration between slide text and the lecturer’s speech.”

On stage at PIE Live 26, I spoke about how English tests should iterate and evolve to keep up with research and with the test taker journey. I spoke even more about it on the convention floor. I mentioned how test companies themselves often produce research that seems to wither on the vine, never managing to influence the assessments themselves.

One hopes that the research at hand eventually has an influence on the IELTS test.

That said, I’m reminded of an earlier IELTS report about the usefulness of integrated writing assessment which ended with:

“The question is not whether an integrated writing task can be implemented in an official IELTS test, but rather, the question is when and how it can be best implemented.”

Those words were written several years ago. We don’t seem to be any closer to seeing an integrated writing task in the IELTS.

British Council Issues 24/25 Annual Report

The British Council has finally published its 2024/25 annual report. It covers the 12 months ending March 2025.

I’m off to the airport in a moment and can’t really dig in too deeply. But a few highlights follow:

1. There was a net loss of 51 million pounds.

2. The government has committed to extend The Loan to September 2027.

3. The BC delivered 2.3 million IELTS tests during the year. That’s exactly the same number as in the previous year.

4. IELTS Online (the remote version) has been expanded to new markets. It is now available in 25 EU countries, and 28 countries outside of the EU.

5. Says the report: “Last year some 25 per cent of our growth in the EU IELTS delivery was attributed to the new online product.”

6. Interesting fact: “As part of the IELTS Secure English Language Test Consortium working for the Home Office, we delivered nearly 90 per cent of all SELT tests globally.”

7. And also: “There was lower demand in some IELTS markets across Sub-Saharan Africa and South Asia due to changes in International Student Mobility policies, which has led to students opting to study through a UK Transnational Education partnership and new destinations where an IELTS test is not mandatory. This has been partly offset by continued growth in China IELTS.”

TOEFL Gets Fresh New Website

The official TOEFL website has a fresh new look. The test’s new slogan – “The World’s English Test” – is featured prominently, as is “Tina,” the test’s new virtual spokesperson.

And a Music Video

Here’s the new TOEFL music video. It’s alright.

In Defence of Listen and Repeat

I wrote the following screed over on LinkedIn in response to some recent attacks on the new TOEFL’s “listen and repeat” task.

It is important to realize that English tests consist of various items that combine to form a useful whole. These items serve various and differing functions, so it is critically important that we consider the entirety of a test and not pick apart individual aspects that, in isolation, don’t seem to serve a desired purpose. Even if we value academic-ness in items we must realize that not everything on a test need resemble academic work.

Some items are more highly correlated with scores on classroom activities than other items. Some items tap into academic skills more than others. That’s normal.

Consider, for example, the listening section of the IELTS. The first two parts deal with everyday situations like training for a job or going on a ski trip (source: my copy of IELTS 13). They involve fewer skills linked to academic work than parts 3 and 4, which deal with campus situations like discussing nutrition labels with classmates or listening to a lecture about the history of coffee. Should parts 1 and 2 be criticized? Of course not… if we consider the listening section as a whole they fit in quite nicely.

To be frank, the spontaneous “take an interview” task on the TOEFL works better because it is preceded by an elicited imitation task. As I’ve said before, the interview task can theoretically be gamed through the use of memorized content. But since the imitation task cannot be gamed in this way, it possibly offsets undeserved high scores in the interview. Do you see what I mean? The “take an interview” task serves score users better because of the item which comes before it. Without the presence of the “listen and repeat” task, it would be somewhat less useful. The new TOEFL speaking section is the sum of its parts – but we only realize that when we view it as a whole.

Same goes for the new writing section, by the way. It is anchored by two constructed response items (“write an email” and “write for an academic discussion”). These are great items. The “write an email” task is extra fun because it feels a lot more spontaneous than what was on the old TOEFL. But like the writing tasks on the old TOEFL (and on the current IELTS) these tasks can be gamed through the use of memorized templates. Fortunately, they are preceded by a short-ish “build a sentence” task that, while it doesn’t resemble academic work, does require a strong command of English grammar and cannot be gamed. Once again, the low-level student who sneaks his way through the constructed response items using test taking strategies will see his score dragged down by poor performance here.

Some might be bothered. They might prefer that tests not include items that feel a bit strange. But test makers are increasingly cognizant of the impact of test-taking strategies and are using all the tools at their disposal to ensure that their tests remain relevant and useful.

By the way, the head of TOEFL wrote, in response, that…

Recently an industry analyst explained the rationale behind some of the design choices in the updated TOEFL with remarkable accuracy. Getting the “why” right without access to the full picture is not easy. So, credit where it is due.

But, alas, I wasn’t actually credited for my keen insights. I remain, now and forever, an anonymous industry analyst.

IELTS USA v. TOEFL

Over on LinkedIn, I published an article about a pamphlet distributed by IELTS USA at TESOL. You can also find it at the blog if you don’t like LinkedIn.

Part of ETS Campus May be Rezoned

A portion of the ETS campus in New Jersey may be rezoned and converted to housing. According to reports, ETS has approached local authorities about the possibility of redeveloping about 75 acres, mostly north of the “ring road” that goes through the property, to make way for 800 houses and commercial space featuring a walkable “town center” layout. Approximately 20% of the homes would be set aside as “affordable” as defined by the state of New Jersey’s Fourth Round Housing Plan.

The 75-acre figure amounts to about 20% of ETS’s bucolic 355-acre campus, which is located just outside of Princeton.

A standing-room-only crowd of local residents objected to the development at a March 16 meeting of the local planning board.

The published version of the housing plan isn’t exactly clear, so while out for a run I listened to the whole frigging public hearing about the redevelopment. As I understand it, all 800 homes may be built in the space between ETS Drive and Research Road. The space currently contains a bunch of older buildings (but not as many as before). As I understand it, Conant Hall, Anrig Hall and “B Building” came down a few years ago, Wood Hall was emptied of most staff just recently, and McBride Hall is currently being rented out to Princeton University as a temporary replacement for a certain library. The unfortunately named Brigham Library doesn’t show up on Google Earth anymore, so I guess it stopped existing at some point in the past decade. Maybe it was the “B Building” referred to above.

South of ETS drive one finds all of the more modern ETS buildings, plus the ETS hotel and conference center.

Apparently, no development is currently planned for the property fronting Rosedale Road and Carter Road.

This all comes as ETS continues to grapple with a tough financial condition that has necessitated layoffs and led it to consider the sale of flagship testing properties like the TOEFL and GRE to overseas private equity.

Regular readers might recall that the former ACT, Inc (now the altruistic but poorly-named IntermediaryEd) recently sold the entirety of its Iowa City campus to developers.

Notes on NABP’s TOEFL Changes

The Pharmacy Times has published a two-part interview with Jasmina Bjegovic which explores the recommendations of the National Association of Boards of Pharmacy task force on the continued use of the TOEFL as a requirement in the certification of foreign trained pharmacists.

There are a couple of interesting things here:

1. The NABP will implement the task force’s recommendations. Perhaps this isn’t a surprise, but it wasn’t necessarily a given. This is confirmed by a note on the NABP website. So at some point in the near future, pharmacists will have the option of taking the OET instead of the TOEFL… and will be able to make use of the TOEFL’s “MyBest Scores” system.

2. In 2020, the NABP introduced a requirement that foreign trained pharmacists complete the TOEFL before taking the FPGEC (that is, the certification exam itself). I remember when that happened, but didn’t think too hard about it. Apparently, that change reduced the number of FPGEC test candidates by 50%… but the FPGEC pass rate increased from 74% to 91%.

Anyway. My question is whether the task force – which met in November – explored the changes to the TOEFL which were set to be implemented three months later. As the changes transformed the TOEFL into something wholly new, they could have impacted the test’s usefulness for this purpose.

Testing Ain’t Easy

Here’s me defending (for the most part) the new TOEFL on the “Testing Ain’t Easy” Podcast.

Oxford Test of English and Rethinking Assessment

I watched this long webinar from Times Higher Education and Oxford University Press about “rethinking language assessment for a changing world.” It includes a wonderful exchange at around 23:00 about the use of “proxy tasks” to predict success in academic programs. Around that point, Hannah Jones talks about the need for clear and accessible peer-reviewed research into the predictive validity of these sorts of tasks.

To me, that’s a really fair point, as we sort of take it for granted that certain tasks are more predictive than others. But are they really? I’ve written here about how tasks on certain tests sort of LOOK better for various reasons (perhaps because they are quite long)… but a closer examination suggests that surface-level observations aren’t particularly useful.

Tony Green, meanwhile, talked about how all English tests don’t have very much predictive validity at all and that “looking to predictive validity isn’t a very productive route to go down.” He mentioned that other measures, like A-Levels, have even less predictive power than these tests.

Again, this makes a lot of sense because success in university depends on a whole lot of things which are not measured on an English test. And, of course, these things are often acquired and honed sometime after our admission. I, for one, earned my highest GPA only in my final year of studies. Needless to say, it took me some time to get the hang of university-level work and, for that matter, the academic register.

I also liked this bit from Tony:

“…a 250-word essay is not actually what people have to write when they get to university. And I think we could do a good deal more in helping [them] to understand that it’s not what they need to do when they get to university. And there’s a big gap between those things.”

That’s a welcome perspective from someone working on a test. I say this because in recent years some traditional test makers have, perhaps, been overselling their products in response to competition from newer providers.

So what does this all mean when it comes to the development and use of tests? That’s the “rethinking” part and you should probably watch the webinar to learn more about it.

International Students in Korea and the Tests they Take

I read today that there are now 238,905 international students in Korea (plus 75,033 people studying the Korean language). That’s kind of interesting.

The top sending countries are China and Vietnam (76k and 63k students, respectively).

(raw data is here)

Some students are on a Korean-language track and might submit a TOPIK score during the admissions process. Others are on an English-language track and might submit some kind of English test score during the process.

I bet that most of the Vietnamese students on the English track submit an IELTS score, while the Chinese students make use of either TOEFL or IELTS scores, depending on their preference (and I think they are more likely to be on the Korean-language track).

Most of the schools here seem to favor TOEFL and IELTS (and, god bless ‘em, TEPS). There may be an opportunity for smaller tests to make inroads in Korea if they engage with the universities a little more.

If anyone from Duolingo is reading this, you could add Sungkyunkwan University to the tracker on your site. People say that it’s a pretty good school. Many years ago I lived near their Suwon campus – I never went inside, but would use it as a starting point for runs along the Suwon Dulle Gil.

von Ahn in Nat Geo

Yesterday morning I packed a thermos of coffee and ambled off to the Doksan Public Library to read the new National Geographic (as I often do). And aha! I spotted this blurb about Luis von Ahn and the Duolingo English Test in the magazine’s feature on “33 Visionary Changemakers.”

I’ll check next month for the Korean translation.

Take ISE Digital for Free

Here’s a chance to take the new ISE Digital Test from Trinity College London for free. You’ll need to take the IELTS-Academic as well (or have recently done so). The fun thing is that you’ll get your IELTS fee refunded by Trinity! I guess this means that a new concordance study is coming soon.

Photo of the Week

I Took the New TOEFL

I Took the New TOEFL for the first time! I scribbled these impressions while everything was still fresh in my mind. If this is all a bit too “inside baseball,” just skip to about the middle of this issue (but if you want even more inside baseball, you can swing by the blog to see the notes I took during the registration process).

Here goes:

1. I was somewhat surprised that my fingerprints and palmprints were not taken. I guess ETS doesn’t do that, but it was done at all the PTE tests I took last year.

2. ETS has distributed custom noise-cancelling headphones to test centers around the world. They are excellent and will improve the in-person testing experience for a lot of people.

3. I think the pre-test microphone calibration test has been adjusted, widening the acceptable range in the “audio visualizer.” That will likely reduce yelling at test centers, which is a common complaint. Regardless, I still recommend removing the visualizer entirely. I don’t think any other tests include one.

4. I think the warning about templates given at the beginning of the test has been updated. It now says that memorized reasons and examples will receive “lower scores.”

5. Overall, the test was broadly similar to the practice tests that ETS has published. That said, there are a few noteworthy differences. I’m not sure if I’m permitted to speak about them at length, but I suspect they will become well-known as more people take the test and more official materials are published.

6. This is a roughly 90-minute test, including instructions. The description on the website is accurate. I suppose most test takers will probably use about 85 minutes, as they can’t go back and use leftover time in the listening section.

7. Speaking of timing, one insufficiently celebrated improvement is the presence of intuitive timers in the listening section. Each listening question now has its own little timer that counts down to zero. This compares favorably to the old TOEFL’s listening section, which had weird and unintuitive timers.

8. The voices used throughout the test (instructions and items) were sometimes… not great. In a few cases, the AI voices used in shorter listening items had a noticeably odd cadence. On top of that, some of the conversations in that section were between speakers seemingly using microphones of wildly different quality, which was a bit jarring. Test takers might hear a woman with a crystal clear voice (probably AI) talking to a man who sounds like he recorded his side of the conversation on his old Nokia flip phone. Thankfully, the academic lectures all sounded great. They might have been recorded by real actors.

9. One of the academic readings was super hard and contained a classic “stare at the screen for five minutes before picking an answer” question that the old TOEFL was known for. One of the others was about a topic I am pretty familiar with so I could breeze through it.

10. It is worth noting how much engagement students have with the new reading passages, despite them being shorter than the ones on the old TOEFL. In the old test, every reading question (except for one) was prefaced with something like “in paragraph 3…” and it was not necessary to go outside of that paragraph to find the answer. So while the article might have been 700 words, the test taker only engaged with ~140 words for each question (except for the last one). That preface has been removed from the new test, so test takers engage with all ~200 words for every question.

11. That said, with shorter articles the risk increases that pre-knowledge of the subject can be used to answer the questions without meaningful engagement with the article. This occurred a few times on my test, but never really happened when I took the old TOEFL. Perhaps this occurs because the item writers have fewer details to work with when crafting questions. I don’t suppose it happens frequently enough to impact score validity, but it is worth thinking about.

12. Reduced dependence on multiple choice questions in favor of more interactive tasks that require more holistic(ish) engagement could address this. Other tests have already gone in this direction.

13. The “build a sentence” UI is better than in the practice tests. Clicking and dragging is much smoother.

14. The spontaneous conversation required in the “take an interview” task is pretty challenging. As expected, the questions were all about a single everyday topic… but I got one that doesn’t come up too often in my everyday life. I actually chuckled a bit to myself when it was introduced at the beginning of the task. I fumbled around trying to quickly come up with some fake thoughts and opinions on it. This is something students will have to work on as they prepare for the test.

15. Students should be warned about the beeps in the speaking section. For the “listen and repeat” item they must speak after an audible beep is played. Two minutes later they’ll start the “take an interview” item, where there is no beep. I missed a few seconds waiting for a beep before I noticed that my timer was ticking down.

16. I know there was much hullabaloo about the elimination of “niche topics like Greek mythology” during the TOEFL revision, but honestly it seems like the academic parts of the new test include a similar range of topics as the old one. Some were pretty darn “niche.”

    Thanks for reading Some of the Above: The Business of English Testing! Subscribe for free to receive new posts and support my work.

Hanging out at Fulbright Korea

By the way, I took the test at the headquarters of Fulbright Korea, which I think is home to the second largest TOEFL test center in the world. Apparently, it can accommodate 245 simultaneous test takers. You can read my “test center report” over on the blog.

How TOEFL Scores Convert

So, to learn how ETS converts between new and old TOEFL scores, you can consult the TOEFL website. You will see, for instance, that a score of 5.0 on the new TOEFL equals 95+ on the old TOEFL. But that’s not exactly how the scores actually get converted on the score reports given to students. I’ve looked at dozens of TOEFL score reports since the new test launched in January. Here’s how they convert new 1-6 scores to the old 1-120 scale:

• 6.0 = 118 of 120

• 5.5 = 110 of 120

• 5.0 = 100 of 120

• 4.5 = 90 of 120

• 4.0 = 80 of 120

• 3.5 = 65 of 120

(I haven’t seen anything lower than 3.5)

And this is it. No other numbers are possible. Everyone who earns 5.0 on the new TOEFL gets 100 on the old scale. Everyone who gets 5.5 on the new TOEFL gets 110. No one ever gets 101, 102, 103, 104, 105, etc.

And no matter how you earn a particular new scale score, it gets converted to the same old scale score.

Just remember that due to rounding, the following is true:

• Section scores of 4.5 + 4.5 + 5.0 + 5.0 = an overall new score of 5.0 = an old score of 100

• Section scores of 4.5 + 5.0 + 5.0 + 5.0 = an overall new score of 5.0 = an old score of 100

• Section scores of 5.0 + 5.0 + 5.0 + 5.0 = an overall new score of 5.0 = an old score of 100

• Section scores of 5.5 + 5.0 + 5.0 + 5.0 = an overall new score of 5.0 = an old score of 100

In every case, the overall score of 5.0 gets converted to an old score of 110 overall, even though the English ability of the test taker is slightly different in each case.

This is only mildly interesting. It becomes slightly more interesting when we look at how schools might use the 1-6 cut scores that have been recommended by ETS.

An institution that previously required a score of 100 points might use the aforementioned score conversion charts and set a new score requirement of 5.0.

It is likely that a greater range of students will be able to meet that new score requirement than could meet the old one. Indeed, the conversion charts suggest that anyone who previously scored between 95 and 106 points would be capable of meeting this target.

But that’s not all. Look at the first score line I provided above (4.5, 4.5, 5.0 and 5.0). The section score conversion charts provided by ETS suggest that a student who would score only 90 points on the old test could achieve that first score line. Thus, the school which previously maintained a firm 100-point cut score is now accepting students that would have previously scored 90 points and above.

Maybe this is still just mildly interesting.

But is it meaningful? Perhaps. Though a few caveats come to mind:

1. ETS hasn’t published any research in support of the score conversion tables. Accordingly, it could be wrongheaded of me to use them in such a precise way.

2. My friends at Cambridge might point out that it is a bad idea to compare scores from the new TOEFL and the old TOEFL at all since they are very different tests. For this reason, it could be totally wrong to assume that a student who scored 100 on the old test would get any particular score at all on the new test.

Anyway…

Language requirements can be tightened up by maintaining both overall and section cut scores. But for the most part, American schools don’t do this. I googled the first ten schools that popped into my mind, and only Brigham Young University had section cut scores listed online.

In any case, given that TOEFL is still widely used, schools might be advised to carry out some longitudinal studies of students who submit TOEFL scores (and adjust reqs accordingly).

Interestingly (and not just in a mild way) this could greatly reduce the amount of re-testing done by TOEFL test takers. This is quite fun, given ongoing speculation about how financial motivations partially inspired the TOEFL revamp.

More on TOEFL’s Listen and Repeat

There is an excellent guest post over at the official TOEFL blog by Daniel Isbell which explores correlations between tasks on the new TOEFL test and ratings of non-test tasks.

Among other things, it says:

“Our research strongly suggests that students who earn high scores on tasks like Listen & Repeat and Take an Interview are able to communicate effectively outside of a test context. This finding held for both lab-based, simulated academic tasks (Graph Description and Lecture Response) and for authentic presentations related to students’ courses.

While the Take an Interview task, which features more spontaneous and creative language production from test takers, tended to have the strongest correlations with non-test tasks, the Listen & Repeat task had similarly strong correlations in most cases. Notably, it seemed to effectively capture the fluency and accuracy related aspects of test-takers’ speaking ability that are at play in all speaking contexts.”

This is relevant as the speaking items on the test (especially the “listen and repeat” task) have attracted attention in recent weeks.

I would love to see more of this sort of posting on the TOEFL site. While the TOEFL blog has launched and relaunched many times over the past two decades, it has never been particularly useful to anyone at all.

In contrast, the team at IELTS Official have set up a wonderful “News and Insights” page that provides a steady drip of useful articles that are grounded in research, but still actionable enough to be of use to score users. Same for the content found on the “Test Center” blog that supports the Duolingo English Test. These sorts of spaces highlight how the tests live in the real world and are tended to by real people with real expertise. The audience that consumes this stuff is tiny, but influential.

This task has generated some spirited discussion over on LinkedIn. Follow along there, if you like.

At-Home TOEFL Drops Windows 10 Support

The at-home TOEFL now requires Windows 11. It can no longer be taken on computers running Windows 10. The presence of that operating system now causes the at-home TOEFL’s equipment check to fail. This change was made sometime after January 13.

In the short term, I suppose this could impact the popularity of the test, as there are still quite a few Windows 10 users. That operating system is still widely used in some key South Asia markets, where the at-home TOEFL is pretty popular. It is also widely used in South Korea, which remains a key market for the test.

The Duolingo English Test and the PTE Express test can still be taken on Windows 10. Same for the IELTS Online test.

(You can also use a Mac, of course)

I, for one, will be sticking with Windows 10 until extended support runs out.

TOEFL Head: Rumors of Sale “Untrue”

TOEFL head Omar Chihane spoke recently to The Indian Express about a wide range of TOEFL-related topics. He said that rumors of TOEFL being sold are “completely untrue.”

IELTS on TOEFL

The IELTS partners have published a research report comparing the new TOEFL to the old TOEFL and current IELTS. The report says that:

“Given the new TOEFL iBT’s substantially reduced text length and complexity, its narrowing of task types, the potentially increased susceptibility to ‘coaching’ and its new scoring system, among other considerations, this report concludes that the revised TOEFL iBT represents a substantial construct shift, undermining assumptions of score equivalence with earlier TOEFL versions. The report warns about the risk for score misinterpretation that can arise from using legacy concordance tables.”

Furthermore, the authors note that:

“…these changes warrant careful caution with respect to test comparability and the interpretability of scores, which raise important questions about the continued suitability of the test for the academic domain, a high-stakes context.”

Strong words.

Paper-Delivered IELTS to End

The paper-delivered IELTS will be eliminated worldwide in mid-2026. The IELTS Official team has announced that “…after careful review, from mid-2026, we will no longer offer IELTS as a paper-based test. All IELTS tests will be delivered on computer. Exact timelines will vary by market.”

This will not come as a surprise for regular readers of this space. Announcements about the elimination of the paper option in individual markets have trickled in one at a time over the past year.

Why are they doing this? Well, according to the announcement this is due to “higher satisfaction” among people who take the test on computer. But savvy test waters will already be aware that this will significantly improve the overall security of the IELTS test.

As a concession to people who like writing on paper, a special “writing on paper” option will be offered in certain markets. Says IELTS:

“[w]e are aware that some test takers like handwriting answers, so we are introducing a new option. In selected markets, we will introduce ‘Writing on Paper’. This update will allow test takers to personalise their test experience by handwriting their answers to the ‘Writing’ component on paper if they choose.”

Which markets will get the “writing on paper” option? That isn’t stated. But India will probably get it. Most readers already know that paper-delivered testing is still pretty popular there. I suppose China might get it as well. Many readers probably aren’t aware of how popular paper-based testing still is in that country.

IDP Education Reports Financial Results

IDP Education Ltd has shared results for the first half of the fiscal year. IELTS volumes are down 7% for the half year. That includes a 27% drop in India. The good news is that IELTS volumes are up 1% outside of India.

IDP administered a total of 638,000 tests in the half year.

Regarding the entry of IDP IELTS into the Chinese market, according to the investor’s call, IDP is “working with a well-established third party testing provider.” The provider’s name wasn’t given, but IDP’s entrance was described as a “3P model” which is a comment worth examining.

It was also mentioned on that call that IDP’s opening of five test centers in the Yangtze River Delta area (Shanghai and its environs) gives them access to 60% of international students from China. This may be true, but capacity could be a concern in the short term – note that the British Council runs 13 test centers in Shanghai alone.

Nothing meaningful about the HOELT tender was mentioned.

IELTS drops in India following IDP’s purchase of the British Council’s operations there are really interesting. Here they are:

• FY 2023: 9% drop

• FY 2024: 42% drop

• FY 2025: 50% drop

• H1 2026: 27% drop

Quite a thing.

Possible Layoffs at the British Council

Politico has reported that the British Council may lay off 400 employees across Europe and the UK. That’s about 15% of their staff in the region.

This comes, like other recent traumas at the organization, as a result of the £250 million loan the British Council received from the UK’s Foreign, Commonwealth and Development Office during the pandemic.

According to testimony by British Council head Scott McDonald early last year, the remaining £197 million that is owed gets rolled over into a new loan every year. This has staved off insolvency, but it means that in addition to £14 million of annual interest, the BC has to pay several million pounds of fees each year.

According to Politico:

“[t]he cuts were triggered as the body approached a new three-year funding settlement with the Foreign, Commonwealth and Development Office, which one person familiar with negotiations said would demand savings of around £50 million a year. The FCDO declined to comment.”

Politico’s report also indicates that cuts might be most severe in Italy where “the entire teaching division would close and the exams arm would be outsourced.” Perhaps that will affect the delivery of IELTS in the country.

One imagines that operations outside of Europe will also be impacted by cuts, but they are not mentioned at this time.

IELTS Price Hiked in India

On April 1, the cost of taking the IELTS test in India will increase by ₹1000, bringing the total cost to ₹19000.

This comes after an increase of ₹1000 in March of 2025, an increase of ₹750 in January of 2024, and an increase of ₹700 in April of 2023.

IDP Education is in a tight spot here. I was set to make a dry comment here about the wisdom of continually raising prices at a time when competing testmakers are drawing away their customers by offering more attractive products in terms of content and customer experience.

But perhaps price hikes are the only way to keep their shareholders happy in the face of a declining Indian Rupee. A year ago, that ₹18000 price tag converted to $342 AUD. Now it converts to $270 AUD.

An additional challenge is the per-test royalty that IDP kicks back to Cambridge Press and Assessment, which I assume is fixed in GBP.

Duolingo Posts Annual Report

Duolingo has published its annual report for 2025. Revenue from the Duolingo English Test was $42,006,000 for the year. Given that the test costs $70, we might estimate that it was taken about 600,000 times in the year.

But that’s a very rough estimate as many people pay less than $70 by using a test voucher or buying multiple attempts at the same time. Others pay more than $70 by requesting express scoring. And, also, in January the test only cost $65.

In any case, this count is down a big 17% from my estimate of 723,000 tests in 2024. Duolingo bucked the industry trend in 2024 (with flat volumes) and 2023 (with a big increase) but 2025 wasn’t so kind to the firm. This will generally be ascribed to the unattractiveness of the USA as a study destination at the current time, but I want (as always) to emphasize that the market is more competitive with every year. Duolingo doesn’t have the “contemporary-affordable” quadrant all to itself anymore and is facing increased competition from a range of affordable at-home tests taken by students heading to the key markets of Canada and the UK.

That’s the last of the financial reports for the time being. I suppose we might see the British Council’s annual report next month. ETS’s annual audit could appear any day between now and June 30.

Duolingo on TOEFL

There is a new post on the Duolingo English Test blog about TOEFL/DET concordance in light of recent changes to the TOEFL score scale (and test content). It is kind of interesting.

It could be more interesting to see research in support of the recommended conversion between old and new TOEFL scores.

Financial Times on Duolingo

There is an article in the Financial Times which rehashes some of the discussions that folks in the UK have been having over the past few weeks. The opening sentence is “[a] growing number of UK universities are using a test developed by language-learning app Duolingo to judge whether foreign students speak good enough English to enrol on their courses.”

The article includes some sample items from the Duolingo English Test and the IELTS Test. Specifically, a sample of Duolingo’s “Is this a real word” task and a multiple choice academic reading task from the IELTS. I’m not sure this is fair. Perhaps it would have been better to compare the IELTS academic reading task to the Duolingo interactive reading task. I pasted a sample of that item type on the blog, if you are interested.

The Times on Duolingo

And here’s another mildly frustrating article on English tests out of the UK. It appeared in The Times.

In their discussion of universities accepting scores from the Duolingo English Test, the author quotes a post from the Duolingo English Test subreddit. The quote is presented in the article as a complaint about how the AI scoring used in the Duolingo English Test penalizes accents. But, in fact, the Reddit post was about scoring in a third party practice app… not the Duolingo English Test.

The article also invites users to “Test [their] English with a Duolingo quiz.” The quiz appears to have been built using heavily modified (and simplified) practice questions from the Duolingo English Test website. This could give readers an incorrect impression of the test’s content.

Also: the quiz is incorrectly scored.

Pearson Shares Annual Results

Pearson has published its 2025 annual results. They indicate that PTE volumes were down 5% compared to the year before.

By my math, that puts PTE’s total volume at about 1.04 million for the year.

Far Fewer Students Heading to Canada

A stark decline in the number of international students heading to Canada is one reason why test volumes are decreasing across the board. The Toronto Star reports that 115, 470 students entered the country last year. That’s 61% less than the year before. I’ll post a link in the comments.

In recent years, a major part of Canada’s attractiveness as a study destination has been its immigration proposition: in just a handful of years an individual can progress from a short course of study to a post-graduation work permit and then permanent residence. But keep in mind that due to regulations that came into force back in November of 2024, it is usually necessary to take an English test at all three steps in this journey. As you can imagine, this has made Canada pretty important to testing firms.

Students heading to Canada have traditionally favored the IELTS. But Canadian study visa applications don’t come with a list of mandated English tests, so students are free to choose from a wide variety of options. My alma mater (a middling school on the east coast) accepts scores from eight different tests, for instance. Anecdotally, it seems like Pearson’s PTE and the Duolingo English Test have gained popularity among Canada-bound students in recent years. A recent IDP Education financial report highlighted on-shore testing in Canada as a bright spot in challenging times.

One fun wrinkle is that while university applications require scores from the IELTS Academic test (or some other academic-ish English test), the PGWP requires scores from the IELTS General test (or the CELPIP General test or the PTE Core). So even if your course of study was short enough that your original test score is still valid following your graduation, you will have to retest. Savvy timing could possibly enable people to use the same scores for PGWP and PR, though.

On the Popularity of At-Home Tests

Considering the ongoing debate about at-home testing, I’ve been asked a few times to estimate the total number of high stakes at-home tests being administered for the sake of higher-ed admissions and immigration. This requires a whole lot of guesswork, but it is fun to speculate.

First up, I estimate that the Duolingo English Test is administered about 650,000 times per year. This estimate is based on revenue figures included in Duolingo’s annual reports. Those are all at-home administrations.

Next, I estimate that the TOEFL is taken about 750,000 times per year. This is based on comments from Amit Sevak that the test was taken “almost a million times” circa 2021/22 and my assumption that volumes have declined since then (as they have across the whole industry). Anecdotally, it seems like the at-home TOEFL is extremely popular in some key South Asian markets, where I estimate that more than 40% of TOEFL tests are taken at home. But in the giant Chinese market, the at-home TOEFL appears to be quite unpopular. In fact, test takers there are required to use a workaround via a voucher from Hong Kong just to access it in the first place. The key Korean and Japanese markets seem somewhere in between. Consequently, I estimate that about 25% of all TOEFL tests are now taken at home, or about 188,000 tests per year.

Next, we have the IELTS Official. The most recent numbers from IDP and the BC indicate that they do about 3.5 million IELTS tests per year. Assuming the usual 75/25 split remains accurate, about 2.7 million of those are probably IELTS Academic tests (the only version available online). While the IELTS partners don’t seem very enthusiastic about at-home testing, it is still offered in 86 countries and in the occupied Palestinian territories. The list of where it is available is dominated by EU countries, but it also includes a few spots where access to test centers might be limited. It must be mentioned that the at-home test is NOT available in any of the mega markets that we are all familiar with (India, China, Vietnam and Nigeria). Given all of the above, I estimate that the at-home IELTS accounts for only 5% of all administrations, or about 160,000 tests per year.

Finally, we have the long tail of moderately popular tests that do at least a few at-home administrations (LANGUAGECERT, Michigan, Oxford ELLT, OET, etc) and those which are done exclusively online (Kaplan, Password Plus, PTE Express, etc). I’m not familiar enough with their operations to guess the totals, but surely they all add up.

Based on these estimates, it looks like the big three tests do a combined total of about 998,000 at-home tests per year. If we include the smaller tests in our total, we can assume that well over a million at-home tests are being taken each year for admissions and immigration purposes.

A caveat to keep in mind, of course, is that some people take these tests for professional purposes, certification purposes and for their own amusement.

Photo of the Week

ETS Seeks Buyer for TOEFL and GRE

The Wall Street Journal is reporting that ETS has held talks to sell both the TOEFL and GRE tests. The non-profit is reportedly seeking a purchase price of about 500 million USD, or a for-profit partner who might invest (but not purchase outright) the tests.

According to the WSJ’s report, ETS “has narrowed the talks to several firms that could potentially buy the exams or act as a strategic investor to expand ETS’s reach into the Middle East and India, said people familiar with the matter.”

As for potential buyers, “[i]nterested parties include the Singapore-based investment firm Hillhouse, the private-equity firms Nexus Capital and Veritas Capital and the education entrepreneur Martin Basiri, some of the people said.”

ETS CEO Amit Sevak has declined to comment.

Regular readers know that Nexus Capital recently purchased the ACT test from its non-profit owner for about 240 million dollars. Veritas Capital, meanwhile, purchased the assets of non-profit testing giant NWEA for about 890 million dollars. Martin Basiri, of course, is the co-founder of the fairly successful Canadian firm ApplyBoard (which everyone reading this is quite familiar with) and more recently an admissions-related company called Passage (which we are all much less familiar with).

Most astute test watchers are well aware that we’ve been on the path to such a move for a very long time. The dominance of both tests has been eroded since the pandemic years, while ETS itself has experienced financial woes that have necessitated several rounds of layoffs and buyouts. Several much smaller testing products have already been eliminated from the organization’s portfolio.

I’ll have more to say on this topic in the days ahead, I’m sure.

The New TOEFL is Live

We are in the era of the new TOEFL. A few notes:

1. As expected, the old TOEFL will live on as “TOEFL iBT Australia” for those headed down under. I understand that scores can only be sent to institutions in that country. It could take some time for the new version to be approved by the Australian DHA.

2. TOEFL TestReady lives on as a means of selling prep only for “TOEFL iBT Australia.” As announced earlier, it is no longer used to supply prep for the new TOEFL.

3. Accordingly, there is far less free prep available than before. Someone can correct me if I’m wrong, but as I understand it there are no longer any free graded tests available for the new version. There are plenty of paid options, though.

4. ETS is no longer advertising an updated “Official Guide to the TOEFL.” Instead, they are advertising an “Official Guide to the TOEFL iBT Test: Pocket edition” (coming soon). This product is described as a “time-limited digital release.” I’m not sure what that means. Amazon has a full official guide listed for a mid-August release. No sign of updated “Official Tests” books, though.

I’ll take the new test when I get over my jetlag. Probably at a test center.

IDP IELTS Returns to China!

IDP IELTS is back in China! Students in that country may now register to take an IELTS exam administered by IDP. At this time, registration is only open for IELTS Academic tests. And they are only available in the Shanghai, Jiangsu and Zhejiang regions.

I think there are five test centers overall. If you want the addresses, just gimme a shout.

As regular readers know, IDP Education Ltd briefly launched IELTS testing in China back in late 2024, but pulled the plug due to regulatory hurdles after just a couple of months. Previous to that, the British Council held a monopoly on IELTS administrations in the country.

This is great news for IDP and its shareholders. And also great news for test takers who have been locked into a single service provider for several decades. It’s bad news for the already beleaguered British Council.

IELTS on Washback

I read a new article from the IELTS team about how a great English test can have a positive impact on the academic lives of learners. It says that “[i]f that test promotes useful learning habits, students arrive well equipped for lectures, seminars and coursework.” And “[t]ests with positive washback ensure that applicants arrive with more than a score; they have the academic skills and confidence to thrive.”

Of IELTS in particular, the article notes that preparing for the IELTS means “better student outcomes, less pressure on support services and stronger reputational benefits.”

But it is important to keep things in perspective. We’re talking here about two-hour English tests. These things are great tools for assessing English fluency, but that’s the extent of their usefulness. A middling student doesn’t become a great student because they spent a little while preparing for an English test. A young person who has spent a few weeks summarizing charts like the ones that will appear on an English test will have gained something… but success at a post-secondary institution is about so, so, so much more than the smattering of questions that appear on even the very best English test.

Whether a student is well-equipped to participate in academic lectures and whether they have academic skills and whether they have confidence (etc) depends a whole lot on their whole lifetime of schooling to date and only a tiny bit on the English test they took.

Institutions must realize the specific usefulness of English tests (this is a form of gaining assessment literacy). Test makers should resist the urge to exaggerate what their tests can do (this is a form of promoting assessment literacy).

More on washback in the days ahead… as soon as I finish listening to the new “Talking ELT” podcast series on the topic!

IELTS on Regulations

In conversation with Times Higher Education, Cambridge University Press & Assessment Press and Assessment chief Pamela Baxter notes that “[t]he increased use of poorly regulated tests does a disservice to students using them for university admission.”

This can sometimes be true. I’ve certainly seen some wacky decisions about which tests are accepted. But it is worth noting (again and again and again) that some of the biggest challenges faced by higher-ed providers in recent months have been linked to the use of heavily regulated tests.

I’m referring, of course, to the IELTS test where nearly 63 thousand test takers received incorrect scores over the course of about two years. And to Pearson’s PTE test, where widespread cheating went unnoticed long enough for almost ten thousand test takers to receive fraudulent test scores. In both cases, higher-ed providers were left to deal with the reality of having accepted students who possibly lacked the English skills necessary to thrive in their academic lives.

I love the regulators like brothers, I really do. I even love the test makers. Having some regulations is much better than having no regulations. But ultimately score users need to remember that regulatory bodies can only do so much and only have so much value. Responsibility ultimately lies with the institutions themselves, and they must always take statements from test makers with a heaping grain of salt.

Paper-Delivered IELTS Eliminated in Libya

The paper-delivered IELTS is no longer available in Libya. You can add that country to the list.

I think the list is now: Cambodia, Bahrain, Bangladesh, Brunei, Iran, Jordan, Korea (South), Libya, Malaysia, Mauritius, Nigeria, Pakistan, Sri Lanka, Taiwan, Uzbekistan, and Vietnam.

Want to Support This Newsletter?

So I keep getting financial pledges via substack from people who want to support this newsletter. That’s really awesome, but due to draconian banking regulations here in Korea, it is unlikely that I will ever be able to turn on paid subs and claim those pledges.

If you really want to support my work, feel free to visit my Ko-Fi page, where you can easily make a donation via credit card or PayPal or something else.

New UK Language Requirements Take Effect

Those new English language requirements for the UK have been in effect since January 8. Specifically, Skilled Worker, Scale-Up and High Potential Individual visas require fluency at the B2 level (up from the B1 level).

Telegraph: Cheating Could Happen

The Telegraph has weighed in on the HOELT tender with an article that could raise some eyebrows.

As regular readers know, the HOELT will be a fully remote test. I’ve covered that in this space at great length. However, according to the Telegraph’s bold headline, “[m]igrants can ‘cheat’ Labour’s online English language tests.”

The Telegraph article refers to “leading test providers” who warn that “the move to an online-only approach risks undermining the quality and security of the testing system.”

Who are these leading test providers? That is not indicated, though the article does refer to earlier comments by Cambridge University Press and Assessment which were made before the publication of the tender.

PTE: Flat for the Year

I read in Pearson’s trading update that “Pearson Test of English (PTE) sales were flat” for the year. A result that is truly “flat” would actually be a pretty good result, considering that test volumes at all the big firms are down down down for obvious reasons.

The update also refers to “expected declines in PTE following volume strength ahead of test enhancements in Q3.” I guess that means people piled into the old PTE in order to avoid the challenge of preparing for new item types in the revised test.

A full report will be available at the end of this month, I think.

Pearson English Express Test Gets New Name

The “Pearson English Express Test” is now known as “PTE Express.”

Valuable PTE Scoring Details

I really like this new “Scoring Information for Teachers and Partners” document from Pearson. It contains a ton of wonderful details about how the PTE test is scored. Included are detailed tables explaining how much weight each item has for each section, and for the overall score. This includes integrated items, which affect more than one section score.

The document also explains why the overall PTE score is not an average of the four section scores (which is weird, but I think I’m starting to get it).

There is also a good explanation of how score reviews work.

I would love to see more of this from test makers. It would be especially useful to learn how scores are determined on the new TOEFL, and how the test adapts to test takers as they work through it.

Some DET Figures

A survey by CollegePulse of 855 students in the USA was shared on the Duolingo English Test blog in December. I was waiting for some clarifications before sharing them, but those don’t seem forthcoming so I’ll pass along a few notes now:

1. According to CollegePulse, thirteen percent of the students in the group who graduated in 2025 took the Duolingo English Test. Meanwhile, 21% of students set to graduate in 2028 took the test. That looks good for Duolingo, as it suggests that their test continues to grow among America-bound students. And, of course, it is worth mentioning that the number was pretty much zero back in the class of 2020.

2. Fifty-two percent of students in the group (all graduating years) took the TOEFL test. That’s actually pretty good for TOEFL. I imagine the number was closer to 80% among the graduating class of 2020… but TOEFL seems to have held on to more than half of its old market share. Can’t complain too much about that. I would love to see numbers broken down by graduating year, though.

3. Thirty-one percent of the group (all graduating years) took an IELTS test. Just three percent took a PTE test. That speaks to why Pearson has launched an alternative test just for the American market.

Be sure to take all this with a grain of salt, as the methodology of the survey hasn’t been provided.

Duolingo Responds to Criticisms

ELT Journal has published Duolingo’s response to last year’s article by Bruce, et al. about new online English tests and their use in UK admissions. Regular readers will surely recall the article, as I have referred to it here several times. It has also been mentioned in the higher-ed press many times. In short, the study suggested that the use of these tests has resulted in the admission of students with poor English abilities.

The authors of the response suggest that the study contains “several substantial flaws that potentially invalidate [its] conclusions.”

To break it down, their main concerns are:

1. The study draws conclusions based solely on the opinions of university staff. But it does not include performance data like student marks and progression rates to contextualize these perceptions.

2. EAP professionals make up the largest respondent subgroup, “yet they typically work only with students requiring language support, potentially excluding perspectives on higher-proficiency matriculants.”

3. While criticisms were leveled against the Duolingo English Test and other new tests, such tests were used by only a handful of the schools represented in the study. This suggests that the students with poor English skills referred to by respondents were actually admitted with IELTS scores.

4. I’ll just leave a quote for this one: “Low-quality research, including studies in which the methods and results do not support the conclusions, is potentially detrimental to both future research efforts and society.”

5. Several of the authors of the study are affiliated with the owners of the test (IELTS) which was most praised in the study. It is suggested that this was not adequately disclosed.

NABP Publishes Taskforce Results

The results of the National Association of Boards of Pharmacy task force on TOEFL are now available. It appears that, sometime in the near future, the long-standing monopoly enjoyed by ETS on the testing of foreign-educated pharmacists may come to an end. The task force has recommended giving such pharmacists the option of submitting OET scores.

The recommendations of the task force are as follows:

1. NABP should give candidates the option to pass the OET as an alternative to the TOEFL iBT as a prerequisite for being eligible to take the FPGEE.

2. NABP should accept the MyBest score on each section of the TOEFL iBT within a rolling twoyear period.

3. TOEFL iBT and OET standard-setting should be conducted at least every six years to assess and account for changes in the test format, blueprint (content outline), or number of questions.

4. NABP should encourage states not to cite specific English language proficiency tests in regulations. Instead, they should approve programs that meet board-approved standards for language competency.

–

Point number four isn’t as interesting as it sounds – rather than allowing for a greater variety of tests, it is just meant to allow states to more easily accommodate changes decided upon by the NABP.

Point number three is interesting, though, as the task force report doesn’t really mention anything about the format change that occurred a few months later.

The task force explored the possibility of accepting IELTS and Duolingo English Test scores, but wasn’t comfortable with either due to various factors.

There are some implications for the TOEFL preparation industry – as I’ve mentioned here a few times, pharmacists account for a significant portion of the overall online spend on TOEFL prep. I’ve known many pharmacists who have spent thousands and thousands of dollars on TOEFL prep (and who have taken the test itself dozens of times).

For more on the fascinating history of TOEFL and the NABP, check out this post on the blog.

Australia’s DHA on the New TOEFL

The Australian DHA’s website on English language requirements for visas has been updated to include the following note:

“To ensure your TOEFL iBT result is valid for Australian visa purposes, you must select ‘Taking TOEFL for Australia’ when registering for your test. Results not registered under this pathway may not be accepted for Australian visa purposes.”

I like the author’s use of “may.” For this sort of bureaucratic requirement, it is important to always be vague and keep people guessing.

This update refers to “TOEFL iBT Australia,” which is the same test as the old TOEFL that was mostly retired last month. Right now I see two test dates per month in some big markets (Seoul and Paris, for example) but no dates in some smaller markets (Cairo and Tel Aviv, for example).

ASU Now Accepts At-Home TOEFL

I saw that Arizona State University now accepts scores from the at-home TOEFL (only for results obtained after January 21). I can’t quickly think of any major American schools that reject at-home TOEFL scores now. ASU was probably the final hold out.

The school hosts quite a few international students. I think they are still in the top five in terms of overall numbers.

Second Lingan Student Headed to Jail for Cheating on TOEFL

A student in Hong Kong has been handed a three-month prison sentence for hiring a body double to take the TOEFL test (at a test center) on their behalf.

According to the South China Morning Post, the test taker scheduled a test at a center in Cambodia, but “[o]n the day of the test, the student claimed to be unwell, then hired a fraudster outside the examination hall to take the exam in her place.”

Reporting in the Standard confirms this detail, noting that “[s]he claimed she fell ill on the test day and was approached by an unknown person who offered to take the test for US$300.”

The deception was detected only when the student’s university later “questioned the veracity of the result.”

Jeepers. Some readers might still be under the impression that test-center cheating is a sophisticated and tricky thing. But perhaps it is time for a rethink, as apparently would-be proxy test takers can be found milling about outside test centers looking for clients. And they work cheap.

Regular readers know that test takers who want to cheat often travel across borders to test centers where they think their schemes will be easier to pull off. In this way, a fraudulent test result can be combined with a trip to the beach, or with the consumption of mixed drinks served at very low prices.

This sort of “exam tourism” is a phenomenon most often associated with paper-based testing, but as you can see computer-delivered testing is vulnerable as well.

Second Student?

Yeah, the individual described above is not the only person headed to jail. Jia Peng, 24, has been sentenced to three months imprisonment for cheating on the TOEFL.

According to court records, Jia also hired a convincing lookalike to take the TOEFL. In his case, the crime went down at a test center in Malaysia. It appears, as in the earlier case, that the scheme was uncovered sometime after a score report was submitted. Interestingly, Jia also attended Lingan University.

I spoke last month with a source who has experience being locked up in Hong Kong. They described conditions in the city’s jails as “pretty spartan,” with inmates being kept two to a cell. The same source described the food served to prisoners as “not bad.”

A point I’ve been trying to make here is that some test makers might insist that their on-site tests are far more secure than at-home tests offered by their competitors… all while not being fully aware of exactly how many people are cheating on those on-site administrations.

ETS Pays Penalty to Test Taker

According to a recent report, the Educational Testing Service (ETS) has been ordered to pay Rs 76,000 to a would-be test taker in Hyderabad. The test taker, M. Kiran Sumedha Gupta, registered for the test using her Aadhaar Card, but was denied the opportunity to take the test when she failed to produce a passport on test day.

Per the Siasat Daily:

“[The Hyderabad District Consumer Commission-3], after examining the case, noted that ETS’s own registration system, which accepts Aadhaar as a valid ID, created confusion for the candidate. It held that charging a high fee without giving clear, consistent guidelines and then refusing to help even after a formal complaint amounted to a deficiency in service. The Commission also observed that this denial caused the student to lose academic opportunities and suffer mental agony.”

I’ve long argued that test companies should always strive to provide “clear, consistent guidelines” (a bold position, I know). ETS’s defence in this case is that guidelines regarding ID are listed in the TOEFL Bulletin. That’s certainly true, but requiring test takers to pore over 45 pages of small print to find certain critical information might not be the clearest possible approach.

I suppose it is worth reminding readers that ETS stopped accepting the Aadhaar Card on test day around April of 2023. But for about twelve months following that policy change, the official TOEFL website for India continued to state that it was accepted. I believe that error was corrected when I drew attention to it in May of 2024 (which means it wasn’t a factor in this particular case, as it involved a July registration).

New TOEFL Practice Tests

There are three new practice tests available in PDF format for the new TOEFL. You can find them in the ETS website for teachers (not the one for learners).

New TOEFL Official Guide (sort of)

The new “Official Guide to the TOEFL iBT Test: Pocket ed” is now for sale. It costs $19 bucks and includes one practice test. Described as “time limited,” I suppose it is a stop-gap product that will be offered until the complete official guide is available in May. You can find it on the official TOEFL site, in the test prep section. I’ll buy a copy next week.

One thing to keep in mind is that the guide can only be downloaded one time after purchase. This could create problems for individuals with unstable internet. The audio and practice test files for the current (now outdated) Official Guide can be downloaded four times, but I’ve heard complaints from users who experienced connectivity issues while attempting to download those very large files.

Pretty New MET Website

I saw that Michigan Language Assessment has launched new websites for their tests, including the flagship Michigan English Test. It looks clear and modern, for the most part. I like it!

Even in 2026 it is important to have a proper website. Many English tests (even widely used ones) have pretty bad websites, which can doom them in an increasingly competitive market.

Picture of the Week

From across the street, Mrs. Goodine and I spied the British Council building in Amman. Students can sit for the IELTS there. There is also an Amideast office somewhere in the city where they can take the TOEFL, but we didn’t see it.

Pearson Fined by Ofqual

Pearson has been fined £2 million by Ofqual for what is being described as “serious breaches in 3 separate cases between 2019 and 2023 which collectively affected tens of thousands of students.”

This includes a £750,000 fine related to the PTE Academic Online test.

Regarding the PTE-related fine, Ofqual’s announcement says that “[i]n 2023, malpractice involved other people sitting the secure test on the student’s behalf, avoiding the remote invigilation safeguards Pearson had put in place. Although Pearson identified the incident and revoked 9910 results affected, it admitted it should have identified the malpractice sooner and reported it to Ofqual earlier than it did.” Elsewhere, Ofqual notes that “[a]n additional 2,906 results were either revoked or withheld identified by Pearson’s business-as-usual monitoring processes during this period”

It is further noted that “[o]n 1 December 2025, Pearson entered into a voluntary settlement agreement with Ofqual to resolve this investigation. Pearson made full admissions in relation to Ofqual’s allegations of non-compliance and has agreed to pay a penalty of £750,000”

In addition to reading the announcement, be sure to check out the “Notice of Monetary Penalty” document, which contains detailed descriptions of the problems that occurred and a detailed timeline of events.

As reported on the blog, the PTE Academic online product was quietly discontinued around November or December of 2024.

I know I sound like a broken record, but one wonders how this will impact the ongoing HOELT tender.

The other two fines relate to the GCSE English language qualification and the new GCSE English language 2.0 qualification (a £750,000 fine) and the GCE A level in Chinese (a £505,000 fine). These are tests I am not particularly familiar with.

A Few Stray Thoughts

I published a few stray thoughts regarding the Pearson fine on the blog:

1. Obviously, this situation demonstrates the importance of keeping an eye on student outcomes when it comes to linguistic fluency (however you define that) and pairing them with the language test scores each student submitted during the admissions process. Institutions which do this might be able to notice the existence of organized cheating rings even before test makers, who generally don’t have access to data regarding student outcomes. Indeed, it seems that this is how Pearson learned about the problem (at least in part).

2. It could be time and resource intensive for higher-ed institutions to do the above. But at least they can do it. It is much harder for organizations outside the world of academia to track such outcomes. I’m not sure how, for example, a board that licenses physical therapists might go about it.

3. Assessment literacy now includes security literacy. It is incumbent upon score users to understand the security practices and the technology used in every step of the testing process (including for at-home testing). When higher-ed staff talk to test makers they need to be able to ask the right questions. Sadly, there isn’t much useful documentation of this aspect of the testing process. So good luck with that, everyone.

4. I’ve always been a big fan of in-house tests for all new arrivals, whether they be used as placement tests, for determining whether extra language classes are necessary… or some other purpose. At their most basic level, they can be used as a way to identify which third-party tests are working best for an institution… at least in terms of determining linguistic fluency.

5. I know it’s a pain in the ass, but if you want to make the best decisions regarding which tests to accept, you really need to actually take the tests yourself. And demonstration versions don’t count.

6. If the timeline in the Ofqual announcement is any indication, I guess we can expect some kind of report on the recent IELTS thing in mid-2028.

IELTS Error Hits Right Wing Press…

Just when you thought this IELTS screw-up would blow over, the right wing press has decided to sink its teeth into the matter.

Says the Telegraph:

“Thousands of migrants may have been given visas despite failing mandatory English language tests following a blunder over marking, The Telegraph has learnt.

Up to 80,000 people sitting a language test run by the British Council were given the wrong results, meaning many of them were given pass marks even though they had failed.

Separately, evidence of cheating has been discovered in China, Bangladesh and Vietnam, where criminals sell leaked test papers to migrants so that they know the answers in advance.

It means students, NHS workers and other migrants with a poor grasp of English have been given study or work visas to which they were not entitled.”

And also:

“Because it took so long to discover the problem, many people who were wrongly told they had passed would have been able to obtain visas and come to Britain legally.”

Unmentioned is the fact that a majority of the people who got incorrect scores likely used them to head to Canada, Australia and other places. Or that many IELTS tests are used for domestic purposes. And the fact that the British Council runs the test in partnership with a few other parties.

…Prompting More Information to Be Shared

The official FAQ page for the IELTS difficulties has been updated. It now includes answers to at least a few of the questions which I have posed in the past few weeks.

For instance, we now know that a total of 63,216 tests were affected. That’s at the high end of my estimated range.

And more specifically, we know that:

• 93% of corrected tests (58,867) had an upwards correction to an individual test component

• 7% of corrected tests (4,344) had a downwards correction to an individual test component

• Five corrected tests had one downwards test component correction and one upwards test component correction.

It is also noted in the FAQ that no tests taken before August 2023 were affected.

More details at the link.

IELTS and Cut Scores

There is an excellent new article on the IELTS website which notes how “[r]esearch shows that the average minimum entry score is an IELTS band 6.6 across countries, below our expert’s recommended levels.” The article says that “[w]hen minimum English language scores are set too low, the consequences can be far-reaching.”

We sometimes hear university staff complaining about how new international students don’t have the linguistic competency needed to excel in their new environment. But what do they expect when their institution requires only an IELTS score of 6.5? Or even a score of 7.5?

The people setting cut scores at universities aren’t wholly to blame when those scores are poorly set. It’s a tough thing to do, and they may not have enough information to work with. The aforementioned article links to the IELTS writing band descriptors. But while those are detailed, their usefulness is somewhat limited. For instance, here’s what they say about the grammar in a band seven essay:

“A variety of complex structures is used with some flexibility and accuracy. Grammar and punctuation are generally well controlled, and error-free sentences are frequent. A few errors in grammar may persist, but these do not impede communication.”

It’s a start, I guess.

If readers of the article dig two links deeper, they’ll find some sample essays with band scores. But there are only a dozen of them. And, curiously, they are all from the paper edition of the test, and barely legible.

As scrutiny of tests increases, it is more imperative than ever that test makers provide access to whatever it is that test takers produce during the test. Technology has caught up with the needs of decision makers, and it isn’t terribly hard to let university staff read all of the IELTS essays produced by all of their applicants. Obviously they probably won’t use this information to make admissions decisions, but it will give them crucial information about how to set the best possible cut scores. They’ll finally know, perhaps, what an applicant who has submitted a score of 6.5 is capable of.

A handful of ancient samples from the test maker can only go so far. Scores can sometimes drift, inter-rater agreement isn’t always fantastic, the scoring criterions interact in curious ways that only become apparent when you’ve got a bunch of responses to look at… and there is always the niggling fear that the samples have been graded by item writers or test developers rather than actual raters. So… the more the better. And instead of talking about providing “more” essays, we should probably be talking about providing all the essays.

And we haven’t even talked about the speaking section.

It is worth mentioning that the TOEFL and Duolingo tests have provided sample responses along with every score report for ages. Other tests have as well.

Want to Support This Newsletter?

So I keep getting financial pledges via substack from people who want to support this newsletter. That’s really awesome, but due to draconian banking regulations here in Korea, it is unlikely that I will ever be able to turn on paid subs and claim those pledges.

If you really want to support my work, feel free to visit my Ko-Fi page, where you can easily make a donation via credit card or PayPal or something else.

Confirmed: IELTS Suspended in Iran

While I was typing up the last newsletter, I got news that the IELTS had just been suspended in Iran. That has now been confirmed. According to friends in Tehran, IDP Education will stop offering the IELTS in that country at the end of January. This means that it will no longer be possible to take the IELTS in Iran at all, as the British Council pulled out of Iran many years ago (2009, I think).

As a result of a special license issued to ETS from the Office of Foreign Assets Control, the TOEFL remains an option for Iranian test takers.

IELTS on Paper to Be Eliminated in More Markets

You can add Bangladesh, Taiwan, Jordan and Bahrain to the list of countries where the paper-delivered IELTS will soon be eliminated. Those announcements all came in before the end of the year.

Beauty Queens and IELTS Stars

New reporting from Vietnam suggests that the nation’s most accomplished models are also some of the nation’s best IELTS test takers.

In recent years my news alerts have been full of curious articles from Vietnam which suggest that fondness for the IELTS test in that country is continuing to increase. To be honest, it seems to have surpassed even the levels of adoration felt for the TOEFL test here in Korea twenty years ago.

Question for readers: is there any good reporting on the origins and historical development of this phenomenon?

IELTS Discount in Korea!

I read in The Asia Business Daily that people who take the IELTS at certain British Council test centers in Korea can get a 21,000 KRW discount on their registration fees.

The discount is available at three test centers in Seoul. Included is the one just a couple blocks from my apartment so if I can find some free time before my holiday I’ll take the test. Since the IELTS hasn’t really changed in some time I haven’t felt a strong desire to register in recent years… but I can’t pass up a deal.

Note that your test date must be by January 31 if you want to get the deal. If anyone reading this tests in my neighborhood, do let me know. I’ll buy you a drink.

NABP Holds Task Force, re: TOEFL…

The NABP (National Association of Boards of Pharmacy) held a taskforce to discuss their continued use of the TOEFL iBT as a requirement for foreign-trained pharmacists seeking to practice in the USA. I’ve attached a photo from their social media here just to show that this seems to have been a pretty big deal. Clearly it was not a formality or a rubber stamp sort of thing.

Notably, the task force was called for before ETS announced changes to the test. One wonders if that complicated the proceedings.

I’ll be keeping an eye on this, of course. I asked NABP when we can expect to see the task force’s report, and was told it would be made public “in several months.”

…And Also Sets new Score Requirements

Though the results of the task force are not yet available, the NABP has published new TOEFL score requirements. Here’s what pharmacists will have to score on the new test:

Reading: 4, Listening: 5, Writing: 4.5, Speaking: 5

New TOEFL Practice Tests Available

There are three new practice tests available in PDF format for the new TOEFL. You can find them in the ETS website for teachers (not the one for learners).

Sample TOEFL Score Report Available

Users of TOEFL Access can now download a sample score report (PDF) from that service’s help page.

The sample report seems to confirm that score reports will include the new 1-6 score scale as well as a legacy 1-120 overall score… but no legacy 1-30 scores for each individual skill. I suppose those could be included elsewhere, but it doesn’t really LOOK like that is the case.

The challenge here is that some key stakeholders (including a majority of Russell Group schools) haven’t yet updated their cut scores for individual skills. American and Canadian schools don’t generally institute section score requirements, so they have the luxury of dilly dallying for two years thanks to the inclusion of an overall legacy score. Schools in the UK, which usually require specific section scores, do not.

TTC Program Launches

The TOEFL Teacher Certification Program has launched (sort of). As I wrote a few days ago, this is a paid certification program newly offered by the TOEFL program. According to the website, it includes a self-paced course and a summative assessment. Participants are also required to take the TOEFL test before being awarded a certificate. Those who complete the program will have their names listed on the official TOEFL website.

Currently the program is only available in China (“global rollout coming soon”), though when I scanned the provided QR code with my WeChat app I didn’t get any information about how to get started.

It is unclear how much it costs to participate.

More Upcoming TOEFL Prep

I attended ETS’s webinar about prep materials for the new TOEFL iBT. A few things were mentioned:

1. It appears that TOEFL TestReady will be discontinued. Says ETS: “There are no plans yet to continue the TOEFL TestReady.” This after “a careful study of test-taker feedback.” That’s unfortunate, as TestReady is always my number one recommendation when someone at the beginning of their test taking journey asks me for advice. It’s also the number one recommendation of just about everyone on social media. The strength of this platform is that it provides free sample tests and free section tests that are all scored. It also provides a random “free question of the day” that is also scored. As far as I can tell, none of the free materials currently available for the new tests are scored.

2. There are no plans to update the free TOEFL course on edX. This is also unfortunate. While I know that edX is way past its heyday, the course was my number one recommendation for people wanting both free practice questions and free instructional content.

3. As noted above, additional free practice tests for the new TOEFL have now been added to the teacher section of the TOEFL website, but they are unscored, and provided in PDF format only.

4. As noted above, ETS is now selling “TOEFL Teacher Certification.” This will be a paid product offered to “teachers who have demonstrated strong competency in teaching TOEFL aligned with ETS standards and procedures.” Currently, certificates are offered to ETS workshop participants at no cost. It is not clear if this practice will continue.

5. “Flex Practice” questions for the new TOEFL are now available to test takers who pay extra for a “genius” package during registration. These are questions focused on specific item types.

6. I guess ACRES and SERVAL are the replacements for SpeechRater and e-rater.

7. Much of the presentation was about the ability of partner institutions to integrate TOEFL practice materials and APIs into their own platforms. Traditionally this sort of thing has been a big deal in China, but has not had much of an impact outside of that country. It really does seem like ETS is seeking to replicate the New Oriental success story in new markets. We’ll see how it goes.

–

Regular readers will likely have noted a push to monetize the TOEFL in new and varied ways. This is not terribly surprising, of course. Overall test volumes will increase whenever the big receiving markets become more welcoming of international students… but the big slice of market share that has been lost to Duolingo and Pearson will never come back. Never, ever. Thus, there is a need to explore new revenue sources. The trick is to do so without being too blatant about it.

New TOEFL Official Guide Coming Soon

There is now an Amazon listing for a new “Official Guide to the TOEFL.” It includes a publication date of May 29. Curiously, it promises only two practice tests (down from four in the current edition). One is left wondering if they will be the same two tests currently available on the ETS website.

A shorter book for a newer test isn’t without precedent, of course. Elderly readers will recall that the first official guide for the TOEFL iBT contained only a single practice test.

Note that ETS has elsewhere indicated that the guide would be available in January and (later on) in February.

New TOEFL Book from Siwon School

While visiting the Kyobo Books branch near Seoul City Hall, I snapped this photo of a new TOEFL book from 시원스쿨 (Siwonschool). The book matches the new TOEFL format, and as you can see it includes the ETS logo and a note that the book is from an official ETS partner.

As I indicated above, it seems like ETS is working hard to build up a new network of supportive teachers and tutors. You can spot ETS branding on a wealth of third party products nowadays.

I was going to purchase a copy of the book for my own personal library, but it looks like two of the three practice tests in the book are the same as (or very similar to) the free tests published on the ETS website. That’s an interesting choice.

University of Manchester to Accept new TOEFL Case by Case

At some point in the last month or two, the website of The University of Manchester was updated with the following note regarding the new TOEFL:

“TOEFL iBT scores from tests taken before January 21, 2026 must meet current 0-120 requirements; scores from after this date will be reviewed case-by-case using ETS equivalency tables.”

ETS Pays Penalty to Test Taker

The Educational Testing Service (ETS) has been ordered to pay Rs 76,000 to a would-be test taker in Hyderabad. The test taker, M. Kiran Sumedha Gupta, registered for the test using her Aadhaar Card, but was denied the opportunity to take the test when she failed to produce a passport on test day.

According to the Siasat Daily:

“[The Hyderabad District Consumer Commission-3], after examining the case, noted that ETS’s own registration system, which accepts Aadhaar as a valid ID, created confusion for the candidate. It held that charging a high fee without giving clear, consistent guidelines and then refusing to help even after a formal complaint amounted to a deficiency in service. The Commission also observed that this denial caused the student to lose academic opportunities and suffer mental agony.”

I’ve long argued that test companies should always strive to provide “clear, consistent guidelines” (a bold position, I know). ETS’s defence in this case is that guidelines regarding ID are listed in the TOEFL Bulletin. That’s certainly true, but requiring test takers to pore over 45 pages of small print to find certain critical information might not be the clearest possible approach.

UBC Accepts Duolingo English Test

The University of British Columbia now fully accepts the Duolingo English Test for undergraduate admissions. Previously, the school’s website still mentioned that it was acceptable only in cases where no other test could be taken.

I think Duolingo has now achieved near-universal acceptance among undergraduate programs in Canada.

New Duolingo English Test White Paper

The Quantum Hub and Duolingo English Test have published a white paper about the DET. The contents will be familiar to most regular readers, but it is worth checking out if you are still puzzled about how the test was able to become so popular over the past five or six years. There is a tendency to assume that the DET is popular because test takers think it is really easy. But it goes deeper than that. Ignore those deeper reasons at your own peril.

BBC on Korea’s “Insane” English Test

Here’s the BBC article about Korea’s “insane” English test. The question about Kant is sort of silly, but the one about video games is similar to what you might find on the current TOEFL iBT (but not, seemingly, on the revised version set to launch next month). Even the relationship you might use to deduce the answer (“as in…” & “the difference is…”) wouldn’t look out of place on a modern TOEFL test. I suspect some of the people complaining about the writing style here would say the same things about a TOEFL test. Many people aren’t great at dealing with academic writing these days.

Hong Kong to Test NET Teachers

I read in the South China Morning Post that native English speakers hired to teach in Hong Kong schools via the NET grant now need to score 7.5 or more on the IELTS (no alternative tests are accepted).

Stories about native speakers who must take an English test are always good for a few clicks.

I recently spoke to a pharmacist who earned his degree in the United Kingdom, but must submit a TOEFL score before he can start working in the United States.

On the Differing Fortunes of ETS and the College Board

The College Board’s most recent 990 filing is now available via ProPublica. It covers the year ending December 2024. Some readers may find it interesting to compare and contrast the post-pandemic fortunes of that organization with those of its sometime partner, the Educational Testing Service.

It looks like the College Board has now fully bounced back from its pandemic-era doldrums. After four fairly lean years, program service revenues now sit at just over one billion USD. That’s down 42 million from 2019.

In contrast, program service revenues at ETS are currently down 273 million dollars from their pre-pandemic high.

Likewise, the value of the assets controlled by the College Board have increased 55% since 2019 to reach a staggering 2.5 billion USD.

Meanwhile, in the same time frame, the value of the assets controlled by ETS have dropped by 15% and now sit at a total of 1.7 billion dollars.

I crunched some numbers and put them into an article on LinkedIn. Note that these numbers will all be obsolete in just a couple of weeks when the next audit of ETS is published.

Anyways. I’m asked now and then why I care about this stuff, so I’ll repeat my usual point that both of these untaxed and sometimes unregulated organizations still have a proud impact on the lives of millions of young people around the world.

Picture of the Week

Just in time for the retirement of my little TOEFL book, someone shared this picture of their well-loved copy on Facebook.

In a world where online courses cost two or three hundred dollars, my $15 paperback ($2.19 for the ebook) found a bit of a niche for itself over the past couple of years. Students can still buy it on Amazon if they are preparing for the current TOEFL, but I will probably switch it to archival status early next week.

We’ll chat about the TOEFL changes, as usual. But also about the HOELT tender and the ongoing IELTS issue. And whatever else comes up. All are welcome.

Last month we talked about the testing scene in East Asia (China and Korea) specifically, and about fifty people turned out. One participant wrote afterwards and said “To be honest, the last 1.5-hour webinar was so much more worthwhile than all the workshops I have participated in over the past few years.”

The HOELT Tender has been published. And I mean the actual tender, not another Request for Information.

Here’s a link. The key takeaway is that the Home Office seeks “a fully remote language testing service.” Says the tender:

“HOELT will move to a digital-by-default approach, while upholding the rigorous standards necessary to support our immigration decisions. The successful provider will deliver a service that combines technological innovation with security assurance, enabling customers to demonstrate their English language skills with confidence and convenience throughout their UK immigration journey.”

And:

“The Supplier shall provide a fully managed Remote Testing service.  HOELTs shall be delivered securely in any location where the Customer chooses to take the test, provided secure test conditions are met.”

Test centers aren’t even mentioned in the publicly available version of the tender, as far as I can tell. How do you like that? The final RFI mentioned the possibility of 268 physical test centers. Note that the bidders will get a longer form of the tender, which could reference the possibility of test centers being included.

I’m not terribly familiar with the tender process, so it is absolutely fascinating to me how the document breaks down the weighting that will be given to certain criteria that make up the “Quality” of each bid. Do check that out to learn how the winner will be selected. Just note that price is pretty important too.

A winner will be selected in November of 2026.

A few other key details:

1. The contract will run from ’26 to ’31. With a possible extension. That’s the same as in the final RFI… but the start date has been pushed back about four months to December of 2026.

2. The tender will not be divided into lots, though the possibility of collaborative bids is mentioned.

3. As expected, the HOELT will be available in 2-skill and 4-skill versions.

4. The estimated value remains £680,000,000 (excluding VAT) and £816,000,000 (including VAT).

Needless to say, this is good news for the team from the Duolingo English Test. They likely have the most experience with and enthusiasm for this sort of remote testing. But it is also good news for a variety of smaller providers who also have ample experience with remote testing. It probably isn’t great news for the IELTS partners, who have been pretty vocal in recent months regarding perceived shortcomings of certain online-only tests.

What About Solo Bids?

If anyone here has the ear of the IELTS partners, it is worth asking if they still plan to bid on the HOELT as a group. It’s likely that they will do so – a bid that brings a sort of “white label IELTS” to the table would definitely have a lot of gravitas and psychometric heft. That said, given that test center delivery doesn’t seem to be a factor anymore, the partners suddenly have less need for each other. Any one of them could easily fulfill both the delivery and development requirements on their own. In the early days of the HOELT we talked a lot about how the massive test center networks of IDP Education and the British Council were huge advantages, but obviously they don’t matter any more.

If you are interested in the possibilities, do take a moment to explore some of the tests that the partners currently develop and deliver on their own. Cambridge University Press & Assessment, for instance, has a very nice test called LinguaSkill. I’d say that it leans “traditional” (if you will forgive the term). It is delivered digitally at test centers but could probably be adapted to a remote format. The British Council, meanwhile, runs the Aptis test which is also pretty traditional. This product is mostly taken at test centers, but a remote version has been available for some time. IDP Education Ltd has a test called Envoy. It is mostly unknown at the moment but has a lot of modern features (adaptive content, 90 minute duration, wholly AI scoring, scores in two hours, etc) that could help it stand out from the crowd. I think proctoring of the Envoy test is async, but I couldn’t confirm that by looking at the website today.

Now, regular readers know that the IELTS partnership has endured for 36 years without any hint of disagreement or conflict between the partners. So I’m pretty sure that a joint bid will take advantage of that history of positive cooperation. But, as I said, it is worth asking for clarification.

It should also be mentioned that the winning test is unlikely to be (strictly speaking) the IELTS test, the Duolingo English Test, the PTE Test, or whatever. It will build on the research and designs of one of those tests, but will be adapted to meet the requirements of the Home Office.

IELTS Administrations Go Awry

Just after I published the last newsletter, we got word that a bunch of IELTS administrations were messed up. Here’s a quick summary of the situation:

1. The results of certain IELTS tests administered between August 2023 and September 2025 have been retroactively adjusted. The tests were administered around the globe.

2. An email sent to test takers (pictured below) indicates that reading and listening scores have been impacted, though social media posts by test takers suggest that it is mostly reading scores that have been changed.

3. According to the IELTS partners, the adjustment is because “IELTS identified a technical issue that led to a small proportion of test takers – less than 1% – receiving incorrect results.”

4. Some scores have gone up, some have gone down.

5. Impacted test takers are being offered refunds and free retakes.

6. Results from both IELTS-Academic and IELTS-General administrations are affected, but only those done on computer, at test centers. Results from paper-delivered tests and tests taken at home are not affected.

7. A Zendesk page has been set up to help test takers and score users navigate the situation.

The nature of the “technical issue” is not stated.

It is stressed that “over 99%” of tests were administered without any problems. That sounds nice, but remember that over the course of 25 months 1% of IELTS tests amounts to more than seventy thousand administrations. Perhaps greater clarity on the number of impacted tests is required from the partners.

Obviously, this is not good. Consider the consequences of an IELTS score that is too low. Many test takers will have missed out on educational and professional opportunities. Some may have failed to meet requirements related to immigration and residency requirements. This is why I’m so hard on test companies in much of what I write. When they mess up, the consequences can be far-reaching and profound.

If given the chance, I would ask the following questions:

1. What exactly was the problem?

2. Exactly how many tests were affected by this problem? “Less than 1%” could mean anywhere from 1 to 70,000 tests.

3. Were any administrations from before August 2023 impacted by the problem? IELTS has contacted test takers who must be given new score reports. But what of test takers whose results have expired since they took the test? Did any of them receive incorrect scores?

4. Are there any instances where a test was incorrectly scored, but a band score change was not necessary (and, by extension, the test taker was not contacted)? If so, how many? And if so, were these included in the “less than 1%” figure given to score users?

5. Were any IELTS UKVI administrations (SELT) impacted by this problem? If so, has the Home Office been informed of cases where received scores are inaccurate?

I was quoted in this PIE News article by Kimberley Martin about the issue. Here’s my quotation, without the scaffolding of the article:

“This story highlights how important it is for test makers to identify problems as quickly as possible so that test takers have sufficient time to protect their interests. In this case, the IELTS partners have indicated that some test takers received scores that were lower than they deserved. As a result, many people around the world may have missed out on life-changing academic and professional opportunities for which they needed a particular IELTS score. Others may have been left in a position where they were unable to meet requirements necessary for immigration or residence purposes. It may be too late for some of these individuals to get back on track. I feel for those people.

Conversely, those who were given scores higher than they deserved may have quickly found themselves struggling in academic environments which they were not prepared for.

Testing companies serve as gatekeepers for academia and for immigration. When they mess up, the consequences can be far-reaching and profound.”

The IELTS E-Mail

Here’s a capture of the email sent to affected test takers:

ETS Expands Board, Adds PE Influence

Sometime in October, the Board of Trustees of the Educational Testing Service was expanded from 14 to 16 members. New to the board are a couple of individuals with backgrounds in private equity – Philip Alphonse of Vitria Group and Victor Hu of Lumos Capital Group. Also joining is entrepreneur Pramath Raj Sinha. Departing is James Pellegrino of the University of Illinois at Chicago.

And Also Adjusts Leadership Roster

The leadership masthead at the ETS website has been updated.

A few additions are worth noting:

1. Janet Garcia is now listed as president of ETS. This appointment was announced a few months ago.

2. Chief Assessment Officer Isabelle Gonthier has been added to the masthead.

3. Paul Gollash has shifted from “Senior Vice President of K-16” to “Senior Vice President of TOEFL and the GRE.”

And a few subtractions:

1. Chief Information Officer Steve Santana is no longer included in the masthead.

2. Nor is Senior Vice President, Global Enterprise Operations Management Dennis Stetzel.

This comes a few months after similar adjustments, covered in an earlier edition of this newsletter.

TOEFL Bypass for Some Taiwanese Students

I read this article about a pilot program in Taiwan that will allow America-bound students to bypass traditional English proficiency testing if they complete a certain amount of coursework in English beforehand (at certain universities). The program was developed by the Taiwanese Ministry of Education in collaboration with Fulbright Taiwan, and has already partnered with some pretty big schools in the USA: Arizona State University, New York University, Teachers College at Columbia University, and the University of Maryland. Eight students will participate in the pilot in February, and the program is expected to fully launch later in the year.

This could end up being a pretty big deal in terms of the business of English testing. Taiwan (like most places in East Asia) is part of the TOEFL heartland, so much so that all of the articles I’ve read about the program specifically refer to it as a way to bypass TOEFL requirements… as if other tests aren’t even considered by students. Our friends at ETS will surely be keeping an eye on the pilot. I suppose if it goes well, universities in other markets will be champing at the bit to secure similar waivers.

According to last month’s Open Doors release, there are about 23 thousand Taiwanese nationals studying in the USA.

CELPIP vs IELTS in Canada

Over on the blog, I wrote an article about how CELPIP seems to be closing the interest gap in Canada.

On Test Lengths and Trajectories

I also wrote an article about how test length might play a role in determining the popularity of tests over time.

Cost of IELTS to Drop in China

The cost of taking the IELTS in China will drop by 180 RMB in a few weeks. That’s about $25 USD. After the drop it will cost 1990 RMB to take the test (about $280 USD). It also seems that the UKVI version will have the same price as the standard version (a drop of 230 RMB). Cancellation and change fees will also be lowered.

It is unclear if this is an effort by the British Council to make their product a bit more attractive as the TOEFL test kicks off a high-profile renewal process, an effort by the NEEA to make the test more consumer friendly, an impending IDP launch, or something else entirely.

Let me know if you see any price changes in your market. Here in Korea, the price was hiked in the summer.

(FYI: It currently costs 2100 RMB to take the TOEFL in China, which is about $295 USD)

Cambridge Issues Annual Report

Cambridge University Press & Assessment has issued an annual report. One page is dedicated to IELTS. It references the ELT Journal article we’ve discussed here many times.

IELTS Suspended in Iran?

As I write this out, I’m hearing that administrations of the IELTS have been suspended in Iran. Probably for a long time. I’ll talk to some friends in that country and cover it in the next newsletter.

Photo of the Week

I took this photo at the Youngpoong Bookstore at Mario Outlet in Seoul. It appears that Hackers Education Group is first out of the gate with updated books in one of the test’s biggest markets.

Anyway, the focus will be the testing scene in East Asia. That, obviously, is a pretty massive market for a lot of tests. It is also a market where the legacy players still reign supreme. Will that continue? We’ll see. I’ve arranged for friends from China and Korea to join the chat.

And, as always, we’ll begin by gabbing about all the latest news, including the ongoing IELTS hullabaloo.

All are welcome. The Office Hours chats are mostly unstructured, and all viewpoints are welcome. The primary focus is community building. The largest bloc of participants seems to be people who help students prepare for tests, but we’ve also had quite a few people from test firms pop in to participate. And even a few brave souls from the worlds of recruiting and admissions!

PS: I’ll have more coverage in the next news roundup, but if you want to read more about the IELTS stitution, start with this PIE News item.

Subscribe now

The IELTS partners have published a new guide to assessment literacy and picking English tests. It’s pretty good.

They suggest asking five questions:

1. Is there research to support test validity?

2. How are speaking and writing assessed?

3. How do your tasks align with academic demands?

4. What security measures are in place?

5. Is it possible to review how test taker performance is assessed against specific criteria?

These are all excellent questions. Score users should certainly ask them. One might also read Goodine’s Guidelines.

From a business perspective, these questions seem to highlight how IELTS Official continues to find itself in a tight spot. They are dealing with competition from more contemporary tests like the PTE Academic Test and Duolingo English Test which are successfully putting forth the argument that shorter items (what IELTS calls “limited-response items”) have a role to play in snapshotting the language proficiency of an applicant (even if only in combination with longer items). On the other hand, they also face competition from fairly traditional tests like LANGUAGECERT which, as recent events have suggested, may be supported by much more stringent security measures than is the IELTS.

ETS Cuts 757 More Jobs

The Educational Testing Service will lay off 757 employees at the end of the year. According to a report in the San Francisco Chronicle, this comes as Measurement, Inc takes over the scoring of several tests included in ETS’s longstanding CAASPP-ELPAC contract.

That contract, which runs through 2027, is quite lucrative for ETS, though it is surprising how much of the work appears to be carried out by subcontractors. For instance, for the fiscal year ending September 2023, Cambium Assessment was paid 26.8 million dollars for subcontracted services associated with the contract.

I suppose these workers are getting slightly better treatment than TOEFL raters, who (according to reports) have seen their shifts dry up without getting actual termination notices. Combined, though, it seems like something close to a thousand jobs cut. That’s hard times.

Below is a copy of the WARN notice sent to the State of California. I put on my fake journalist hat last week and requested a copy. Note that although the notice was filed in California, the workers are based around the USA.

Garcia Confirmed as ETS President

ETS has finally confirmed the appointment of Janet Garcia as president of ETS. Garcia currently serves as CEO of PSI Services, a position she will continue to hold as she takes the reins at ETS.

CEO Amit Sevak will remain with ETS, perhaps with more time to devote to public speaking engagements.

It may be worth keeping an eye on the masthead over on the ETS website to see if any other roles change hands in the days ahead.

IDP Education Holds Annual General Meeting

You can read a transcript of IDP Education Ltd’s Annual General Meeting right here. It’s mostly quite dull, but I suppose it is interesting to see how IDP’s future depends on governments becoming more open to immigration in the near future. That’s a dicey proposition in a world where voters around the globe are becoming more nativist and more nationalist with every passing day.

IDP Education: Takeover Target?

Meanwhile, a new research brief from Macquarie Analysts places IDP Education at the top of a list of 16 likely takeover targets. As noted in the Australian, “private equity firms have been on the march this year in terms of privatisations of ASX-listed companies.” IDP is mentioned as a likely target due to steep declines in its share price over the past year (though they aren’t actually as steep as indicated in the article).

Over the years, many test watchers have observed the IELTS Official partnership, and wondered how its dysfunction might be increased. To me, handing one third of the partnership over to a private equity fund could be the best possible approach. An American fund would be the most efficient choice, of course, but a domestic fund would certainly do in a pinch.

British Council Selling “Everything”

It seems that the British Council isn’t making much progress in its quest to get the £197mn it owes to the FCDO written off. In the meantime, they are making cuts and selling assets. Speaking to the Foreign Affairs Committee a few days ago, CEO Scott McDonald said:

“…we are now selling everything the British Council has that we are able to sell. I talked about estates last time; we have 42 properties around the world and we are selling, or attempting to sell, every single one that is possible to sell. We are selling that bit of the art collection that we are able to sell under the terms of the museum code. We sold the school we had in Madrid. We do not have anything else.”

That school in Madrid, by the way, was sold for £50 million. It had been operated by the British Council since the 1940s. But even accessing the proceeds of the sale is proving to be tricky.

Test watchers might be asking if McDonald really is in the process of selling “everything” that can be sold. Given that the British Council’s IELTS operations in India were sold to IDP Education for £130 million just a few years ago, some might be wondering if we might see the rest of their IELTS business on the auction block in the near future.

The answer is: probably not.

Per the current agreement between all three IELTS partners (Cambridge, BC and IDP) the partners can only sell or transfer their share to one of the other founding partners. The British Council couldn’t, for instance, sell its IELTS business to a deep-pocketed competitor like Pearson or PeopleCert.

Meanwhile, IDP isn’t in a position to acquire expensive assets right now, especially one that might come with a price tag that exceeds its current market cap. And Cambridge, it is safe to assume, is content to continue developing the test while collecting a lucrative per-test fee from administrations carried out by the other partners.

Intelligibility and Comprehensibility Research

New from Studies in Language Assessment is an article by Daniel Isbell, Dustin Crowther, Jieun Kim and Yoonseo Kim which looks at the speaking tasks included in British Council’s Aptis Test. Specifically, it compares official Aptis scores to assessments of intelligibility and comprehensibility from laypeople recruited from the “Prolific” online research participant pool. Scored responses from 50 test-takers were selected and those participants rated their comprehensibility on a 1-9 scale. For intelligibility, the participants were asked to transcribe the recordings. Those transcriptions were compared to criterion transcripts created by the research team. Among other things, the authors note that “to the ears of layperson listeners… speakers that earned higher Aptis scores were more intelligible and easier to understand.”

I would love to see more of this kind of research. I suppose a similar study could be done of test-taker responses to just about any English test.

I Took the GateWay Test

I took the Gateway English Test from English3 a few days ago. I had a lot of fun with it. I’m not a psychometrician, so I can’t evaluate the validity of items, but I was happy to see that the English3 team put some thought into designing interesting tasks. There are some negative points, but I’ll save those for the end of this post, lest anyone gets the wrong idea about me.

A few notes:

1. This is a 90-minute test with a $99 price tag. Proctoring is asynchronous, which I guess puts it in the “contemporary affordable” quadrant with DET and PEXT. Results come in five days or you can pay extra to get them more quickly.

2. The test includes meaningfully integrated tasks! They include listening to a lecture and answering a question about it orally, listening to a conversation and giving an opinion about it in writing, and listening to a “zoom” call and summarizing each of the speaker’s points in writing. These are fun.

3. Content is mostly “academic” with some “campus life” stuff. I didn’t spot any non-campus “daily life” content. Reading and listening passages resemble what you might find on other tests.

4. Since this is a 90-minute test, there is still time for a complete essay. Indeed, there is quite a lot of written and spoken production. The 90-minute length gives designers room to include quite a lot of speaking and writing, if that’s what they value. Going with a 90-minute length is a tough decision in a world where 60-minute tests seem to be the future… but that extra half hour does give designers a certain amount of freedom.

5. The test uses the same on-screen note-taking widget as the ITEP. I like that.

6. The test starts with 5 unscored speaking questions. Responses are shared with score users.

7. The list of accepting schools seems to skew toward faith-based institutions, which is really interesting. While taking the practice test I sort of sensed content that might appeal to the CLT folks, but didn’t pick up on any of that in the actual test.

8. One almost senses that if the TOEFL team had a bit more time to think about their relaunch, they might have come up with a product sort of like this one. This is a fun, non-threatening test that includes an extended writing task, a ton of speaking, meaningfully integrated tasks and a splash of “campus life” stuff.

Meanwhile, some of the not-good stuff:

1. Security seems dated. This is an asynchronously proctored test that utilizes neither a secure browser nor a secondary camera. I think institutions expect a bit more in 2025.

2. Scoring is wonky. The maximum score is supposed to be 600 points. I scored 605 points. My listening score was 680/600. Whoops.

3. Instead of linking my scores to the CEFR, my score report just contains the letters “CEFR.” Hmmm.

Remember the APIEL Test?

While I was searching the warrens of the College Board’s website, I stumbled upon something I didn’t know I needed: this technical manual for the AP International English Language Test! Developed for the College board by ETS and offered from 1997 to 2002, the APIEL Test was used by some schools in the United States alongside TOEFL and IELTS to confirm English fluency when admitting international students.

Testing enthusiasts might find the content of the manual somewhat interesting. This was a three-hour test (plus time for instructions) and if you squint at it long enough you can see how it may have been influenced by research being done in support of the TOEFL iBT. For instance, this was the first ETS test with both a speaking section and a writing section. Meanwhile, in true AP fashion, the reading and listening items are sourced from real-world publications rather than the somewhat stilted constructions we see in tests today. I even spotted an uncredited excerpt from “The Spy Who Came in From the Cold.”

Perhaps some readers of this space contributed to the development of the test and have some memories to share.

This test actually came up in a conversation I had at DETcon 2024; another attendee reminisced about how ETS held similar events in the 90s to promote the product. Recall that back then ETS ran most of the College Board’s testing programs.

This seems to have been the College Board’s only foray into the world of high stakes English testing (aside from their early management of the TOEFL). I’ve often wondered if they have ever considered getting into this market. There is still some potential for profit, I think. Just off the top of my head I can count TEN firms trying to make money on high stakes testing for college admissions in the UK, but only four trying to profit from the much larger US market.

Duolingo Q3 Results

Duolingo published Q3 results last week. Revenue from the Duolingo English Test amounted to $9,649,000 for the quarter. If we divide that number by the cost of taking the test ($70) we get a total of 137,842 tests for the quarter. That’s a decline of 17% compared to the same quarter last year. This figure tracks with what we’ve seen so far this year: Q1 was down 17% and Q2 was down 12%.

Just keep in mind that this is a rough estimate, as some people pay less for the test by purchasing a bundle and others pay more for the test by purchasing express scoring. Others pay nothing at all by using a waiver from the Duolingo Access program (25k waivers were distributed last year).

How does this compare to the industry as a whole? By my data, PTE volumes were down 10% in the first half of 2025, while IELTS volumes at IDP were down 11% for the first half of 2025.

IDP will publish a new half-year report in February.

We might learn about IELTS volumes at the British Council if they publish their annual return in January, as required (but they probably won’t). ETS doesn’t publish TOEFL test volumes, but keep an eye out for the release of the organization’s annual audit in early January.

UVIC to Accept Duolingo English Tests

The University of Victoria will accept Duolingo English Test scores starting in May of next year. I think they were the largest non-accepting school in Canada last time I did a scan of Canadian institutions.

Duolingo Room Scan: It’s Fine

I finally got around to experiencing the new room scan for the Duolingo English Test. The room scan is mostly painless. It resembles that of the British Council’s EnglishScore test: the test taker just has to slowly spin around their room while holding the secondary camera (their phone). There is also a second scan, which hasn’t been discussed much in this space. To complete it, the test taker first points the phone at their keyboard, and slowly moves it towards the space behind their computer. It is a little clunkier than the main room scan, and the test taker might attempt it a couple times before doing it properly.

Some have argued that requiring test takers to have a compatible smartphone is burdensome. This is true, to a certain extent. That said, many test takers will find that it alleviates the burden of completing room scans by carrying a laptop around their apartment. Remember that most test takers don’t have a cool-guy Macbook like the people who design and market tests. Instead, they have a barbell of a Windows machine with a bunch of peripherals connected via USB and a battery that has to be plugged in all the time because it maxes out at a 5% charge. Doing a room scan with such a machine really sucks. Ask me how I know.

Oxford Rejecting New TOEFL (For Now)

Officials at the University of Oxford have posted the following note on the school’s website:

“Changes to the TOEFL iBT test are being introduced on 21 January 2026. The University will not accept TOEFL tests taken from that date to meet the English language condition until a review of the revised test has been completed.”

It is unclear when this review will be completed.

TOEFL Score Report Oddity

Some folks have reported being charged $39 (each) to send additional TOEFL score reports. It currently costs only $29 when I try to send scores from my account. I don’t know if ETS is doing an A/B test, or if someone in Gurgaon pressed the wrong button.

No More Free Paper TOEFL Score Reports?

According to the TOEFL Naver blog, in January ETS will end the practice of sending paper score reports to test takers at no charge. Currently, score reports are mailed to all test takers at no extra cost when the request is made at the point of registration, with express shipping available for an additional fee. Come January, only express shipping will be available.

Based on the current fee schedule, that means test takers who need a score report will be asked to pay an additional $25 at the point of registration, or $60 if one is requested after the test.

Social media posts suggest that there is still quite a lot of demand for paper score reports. They also suggest that standard (free) delivery is unreliable in a lot of places. And that most test takers are unaware that those usually come in through their national postal service.

Some test takers will be unhappy about paying more (per recent price hikes) and getting less. On the other hand, there will be fewer cranky test takers complaining about not getting their score reports.

If memory serves, IELTS Official still provides paper score reports in many markets, but has begun phasing them out in certain instances. For example, the British Council no longer provides them to people who take the computer-delivered IELTS in Korea. Pearson and Duolingo, meanwhile, do not provide paper score reports at all.

Following this change, LANGUAGECERT may be the only firm that still provides a paper score report to all test takers at no extra cost.

Stray Thoughts on Test Volumes

Between June 2021 and June 2022, IDP Education administered about 1,900,000 IELTS tests. During that same period, Pearson administered about 591,000 PTE tests.

Between June 2024 and June 2025 IDP administered about 1,293,800 IELTS tests (a 600k drop). During that same period, Pearson administered about 993,000 PTE tests (a 400k increase).

This according to the annual reports published by each firm.

Apples and oranges, I suppose. But one is left to wonder when the number of tests administered by Pearson will surpass the number administered by IDP education. One is also left wondering if Pearson (which seems to focus on a narrower range of markets) already does more tests in India than IDP.

(By my crude calculations, DET jumped from 494k to 684k in the same time frame)

Picture of the Week

I took the new “English Express Test” from Pearson. That’s their new product intended for students applying to American schools. It basically exists in the “contemporary affordable” category that Duolingo has dominated in recent years.

I liked the test. Practice materials are free and seem to be plentiful. The UX is pretty smooth and I was happy to see some integrated tasks.

Read on for a detailed look at the test.

(I suppose it is worth mentioning that I write the same sorts of reports for clients, but at a much greater level of detail. You can call me if you are interested.)

Pre-Test

1. Free practice tests are delivered within the secure app used on test day. I think I wrote positively about this feature back in 2023 when this test was a Versant product. And I stand by that – but I realize that some test takers won’t like it. It means that you have to follow most of the test day rules just to take a practice test. You must disable your second monitor, you must unplug or switch off extra microphones and headphones, etc. And your webcam will be activated during the whole practice test (so I felt obligated to put a shirt on). This places some amount of cruft between the test taker and the practice tests they want to take. It likely limits the variety of operating systems and devices the practice test can be accessed on.

2. The practice tests are scored! Even the writing and speaking. It takes 15 minutes, though.

3. The practice tests are just 30 minutes long. I suppose that Pearson, like Duolingo, will eventually realize that it is best to just deliver full-length practice tests.

4. I took two free practice tests and they were different! I don’t know how many variations there are, but this is the future. Test companies that are leaning heavily into making money from prep should realize that the days of selling practice tests for $55 a pop are coming to a close.

5. The test costs $70.

6. The user account is blissfully clean and uncluttered. I’m in love.

The Test

1. This test is not dissimilar to the TOEFL iBT. It has some similar question types and some similar design philosophies. Many of the tasks are similar, and in general both seem to favor short form items. As I’ve h

2. There is not much of what one might call “academic English” on this test. Less than the Duolingo English Test. And less than the new TOEFL. I counted about two such questions. They made up about 4 minutes of the 60 minute test. There is plenty of “campus life” English, but that’s not the same thing. This is not a complaint, but it might be a point of controversy as some people argue that “academic English” is a thing of nebulous definition. Remember that this test hits the market just a few months before the TOEFL test relaunches in a format that drastically de-emphasizes traditional “academic English.”

3. Note-taking is not permitted.

4. The test UI is clean and pretty. Prettier than average. But it should include a button to increase the text size. Or they should just make everything frigging huge, like on the Duoligo test.

5. I was surprised by the number of British and Australian accents I heard, considering that this is a test for only those seeking to attend American schools. I imagine they would be better off replacing these with non-native accents, to be honest.

6. One of the speaking tasks requires test takers to speak for 60 seconds, but there is no timer on the screen.

7. There are some tasks which we might call “integrated.” I’m happy to see that Pearson is doing its best to include integrated tasks, despite the short length of the test. The most notable is the “summarize a conversation” task where the test taker must listen to a fairly long conversation between three people and then summarize it orally.

The Security

1. The test uses both a secure browser and a second camera. I like that the second camera is implemented via a web-based application instead of a whole app that must be installed.

2. No room scan, though.

Post Test

1. Unofficial results came in 15 minutes! This feature could set the test apart from Duolingo’s product. My official scores were reported in about 48 hours, as promised. Will try to append a copy of my score report to this article.

2. The test doesn’t send a “test successfully completed” e-mail at the end. Test makers should realize that students feel anxiety if they don’t get such a message (even if they can learn as much in their user account).

3. I can’t find any research supporting the design and/or validity of the test.

   

UK Visa Changes set for January 8

Those changes to English requirements mentioned in the White Paper are set to come into effect January 8. Requirements for applicants in the Skilled Worker, High Potential Individual and Scale-up routes will increase from B1 to B2.

The change is part of a collection of changes which the Home Office says will replace “Britain’s failed immigration system” with something better.

The Home Office (and press) refers to this as requiring “A-Level standard” of English. Maybe they are trying to make a comparison to earning an A-Level in a foreign language, and not that immigrants will have to be able to understand Wuthering Heights or Sense and Sensibility.

City & Guilds Sells Everything

City & Guilds Institute has sold its assessment business to PeopleCert (they own LanguageCert) for an undisclosed amount. According to FE Week, CEO Kirstie Donnelly and almost all of City & Guilds’ 1,400 staff will move over to the PeopleCert organization. They will leave behind a very wealthy charity which will, one assumes, be seeking a new role to play in the world. PeopleCert has a press release out.

Recall that PeopleCert earlier acquired the City & Guilds English assessment business and from that formed the LanguageCert group of tests.

I’ve written quite a few times about non-profit assessment organizations selling their assets to buyers from the for-profit world. There will be more sales to come, I’m quite certain.

Leadership Changes at ETS

There are some leadership changes at ETS to ring in the new financial year. PSI’s CEO Janet Garcia is now also president of ETS. She’s joined at ETS by PSI colleague Isabelle Gonthier, who now serves as Chief Assessment Officer of both firms. Gonthier’s podcast “Tried and Tested” is my absolute favorite bit of related media, and I sure hope it continues despite a possibly increased workload.

Meanwhile, the masthead at the ETS website no longer lists the firm’s Senior VP of Global Mobility Solutions Rohit Sharma (who has confirmed his departure from ETS) nor Chief Marketing & Innovation Officer Michelle Froah.

I don’t like to mention people below the c-suite by name, but a few other changes caught my eye:

1. ETS’s Country Manager for Japan has shifted to an advisory position. That appears to make it a clean sweep as far as country managers for key TOEFL and GRE markets go. As I mentioned in an earlier post, there may be an effort to install people with business backgrounds in these positions instead of people with careers in education.

2. Several senior directors with expertise related to institutional partnerships have decamped to LanguageCert.

3. Most of the friendly faces that test prep folk will know from the ETS Global workshops for teachers have departed. That’s kinda sad.

TOEFL Price Hikes This Month

Big price hikes for the TOEFL in most of the world took effect this month. I started writing them all down alphabetically, but the list was too long. I do know that the price increased by $25 in Albania (to $250) and $45 in Burkina Faso (to $255). That’s as far through the alphabet as I got.

Switzerland retains its title as the most expensive place to take the test, at $475 (up $5).

The price in India is now ₹18,000 including tax.

If you have a country that matters to you, let me know and I’ll compare the current price with last month’s price. Note that EU countries are harder to track because the new sticker price is inclusive of VAT, while the old sticker price was exclusive of VAT.

But Someone at ETS has a heart: prices were cut in Palestine, Ukraine and Sudan.

University of Glasgow Rejects New TOEFL?

It appears that the University of Glasgow will not accept scores from the revised TOEFL iBT Test for postgraduate studies. Every program has its own page, but I now see that “only tests taken before January 2026 can be accepted for entry” has been added to several of them.

New TOEFL iBT Technical Manual

There is now a Technical Manual for the revised TOEFL iBT Test. You can find it via its home on the website of the ETS Research Institute.

Many readers will be most interested in its detailed descriptions of the items included on the new test.

A few frequent questions have been answered:

1. The reading and listening sections will each contain 35 scored questions in total. The reading section might contain 15 unscored questions, while the listening section might contain 12 unscored questions.

2. The exact number of questions of each type is listed.

3. Everyone gets the same mix of items in the routing module, of course. The easy reading module doesn’t contain an academic reading passage and the easy listening module doesn’t contain an academic listening passage. The hard reading module contains one academic reading passage and the hard listening module contains two academic listening passages.

4. Human raters will not score every speaking and writing response. Just some of them. The manual notes that “[f]or responses where the automated scoring lacks confidence or encounters difficulty, human raters step in to provide scores, ensuring reliability across all responses. In addition, a random sample of responses is regularly reviewed by certified human raters to ensure quality and inform model updates.”

TOEFL, Pearson and the NABP

I noticed that Pearson administers both the NAPLEX and FPGEE exams for the National Association of Boards of Pharmacy. I’m curious if they might try to leverage that relationship as the NABP convenes a task force to determine if the TOEFL is still suitable for the purpose of judging the English fluency of foreign-trained pharmacists.

Trinity College Launches ISE Digital Test

I read in the EL Gazette that Trinity College London’s new “ISE Digital” test launches this month in markets around the world.

New Print Guides for PTE Academic Test

I was happy to discover that Pearson has published updated print guides to the PTE Academic test (one for students, one for teachers). They cost $86 AUD a piece and only seem to be sold from Pearson’s Australian web store. But they do exist. Apparently.

I am still a big believer in print books because they can be stocked by libraries around the world. Libraries remain a major source of free test prep – the long reservation queues for prep books at many libraries speak to that.

PTE: “Strong Performance”

Pearson’s Nine Month trading update describes the PTE as having “strong performance” in Q3. Dunno what that actually means, but there ya go.

(Said one wag on LinkedIn: “‘Strong performance’ is always code for “no context will be given, please keep guessing.”

Want to Support This Newsletter?

So I keep getting financial pledges via substack from people who want to support this newsletter. That’s really awesome, but due to draconian banking regulations here in Korea, it is unlikely that I will ever be able to turn on paid subs and claim those pledges.

If you really want to support my work, feel free to visit my Ko-Fi page, where you can easily make a donation via credit card or PayPal or something else. Ko-Fi is actually better than Substack for creators. One cool person has already made a donation.

Picture of the Week

This reader-submitted picture is good for a chuckle, if you look closely.

Pearson has launched the new “Pearson English Express Test” in select markets. According to my notes, this launch comes about one month earlier than originally planned.

The test may be available in my market. If so, I’ll do my best to take it in the weeks ahead. I gotta send my passport to the shop pretty soon, so that might slow me down a bit.

Pearson’s express test seems to represent an effort to compete in the same space that Duolingo currently has all to itself. It is a one-hour test, is remote-only and costs $70. I believe that scoring is mostly AI and that proctoring is asynchronous. Scores can only be sent to institutions in the United States.

A short list of accepting schools is available here.

University of Cambridge Won’t Accept Scores from New TOEFL

In case you missed the special edition, I wrote about The University of Cambridge not accepting TOEFL scores, and Duolingo adding a room scan over here.

Cambridge on Remote Testing

I read an article out of Cambridge published by the Higher Education Policy Institute. It’s a curious read. This part struck me:

“The proposed introduction of a Home Office English Language Test (HOELT) raises the stakes still further. The Home Office has indicated an interest in at-home invigilation. While innovation of this kind may appear to offer greater convenience, it also risks undermining quality, fairness and security. The HOELT process must be grounded in evidence, setting high minimum standards and ensuring robust protections against misuse. High-stakes decisions such as the creation of HOELT should not be driven by cost or convenience alone. They should be driven, instead, by whether the system enables talented students to succeed in the UK’s competitive academic environment, while safeguarding the country’s immigration processes.”

This is a fair and principled stance. And sincerely held, I’m sure. But to go the route of telling the Home Office that the test they are desirous of (a fully remote at-home test) is the wrong test is a curious move. It could backfire. I’m imagining a restaurant-goer who orders a salad only to be told by the chef that he ought to get a steak instead.

What’s more curious is that Cambridge has been offering an at-home IELTS test for more than five years. And with nary a peep about security concerns. Indeed, to my eye it seems to have more reliable security than the traditional test center administration. Hundreds of people take this version of the IELTS every day. So why not lean into that?

Duolingo’s bid is completely beatable. Even by the standards of at-home testing Duolingo’s approach is new and scary. I can picture the HOELT tender being won by a testing firm that openly states its love for at-home testing but presents the sort of conservative game plan that the Home Office is probably looking for.

A result where a smaller testing firm emerges to snag the tender might be Cambridge’s worst nightmare. And a source of great amusement for the Duolingo team.

LanguageCert Research Report

LANGUAGECERT has published a new article by Leda Lampropoulou, Irene Stoukou, PhD and David Coniam called “The Relationship between Language Complexity and Test-taker Achievement on a High-Stakes Test of Writing.” It compares lexical, syntactic and discourse features of 2,746 test taker responses from the LanguageCert test at the B1, B2 and C1 levels.

THE on Tests

Times Higher Education published a long article by Helen Packer about English testing for higher education in the UK. It is pretty evenhanded, inasmuch as everyone got a chance to say their piece. I liked it.

TOEFL Reading Deep Dive

I took a deep dive into what we know so far about the new TOEFL reading section. You can find it on YouTube.

No New TOEFL for Australia and New Zealand?

I was reading the instructions for “TOEFL Access” (as one does) and I saw a notice that “Australia and New Zealand will not be affected by the significant updates to the TOEFL iBT® test which are being made after January 21, 2026 that include a new 1–6 score scale and adaptive testing capabilities.”

NABP Examines TOEFL

The National Association of Boards of Pharmacy (NABP) has convened a task force to determine whether foreign trained pharmacists should still be required to submit TOEFL scores as part of their credentialing process.

The resolution establishing the task force notes that “questions have been raised as to whether the TOEFL iBT effectively measures English proficiency in reading, listening, speaking, and writing for this purpose.” And that “concerns have been raised regarding whether this requirement may be a potential barrier to licensure.”

The task force will meet in November.

Regular readers know that credentialing of foreign trained pharmacists is one of the most lucrative on-shore use cases of the TOEFL iBT. The TOEFL is currently the only test accepted for this purpose and, as I’ve discussed here, it is common for would-be pharmacists to take the test 10, 20, 30 or even 50 times. ETS has enjoyed this lucrative monopoly since at least 2008, according to my research. Its loss would be significant. And, not for nothing, it would be catastrophic for the online TOEFL prep community.

Some have suggested that ETS has taken this use case for granted over the years.

The resolution establishing the task force was passed at the end of May, so it isn’t a response to recently announced changes to the test. But it comes at a bad time for ETS. The NABP will be forced to examine a wholly different test than what they originally signed up for. And officials will surely be aware that back in 2024 their regulatory counterparts in Canada stopped recommending that provincial pharmacy boards accept TOEFL scores.

Given the importance of this use case, I am quite certain that alarms sounded across Lawrence Township the moment the task force was announced. I’m sure that the TOEFL product owner scrambled a team of B2B pros to ensure that his test remains acceptable to the NABP for many years to come. But maybe I’m wrong. In any case, you know how to reach me if you need a consult.

It is safe to assume that other test makers took notice of the task force. The folks at Michigan Language Assessment, for instance, have long been focused on English testing for medical professionals and have plenty of experience wrangling state licensing boards. Their flagship product has a vague resemblance to the outgoing TOEFL iBT which the NABP has been devoted to for so many years, but without the little annoyances which make some people think of it as a “barrier to licensure.” I wouldn’t be surprised to hear that our friends in Ann Arbor responded to this development even more quickly than ETS did.

New TOEFL Official Guide in January?

A new Official Guide to the TOEFL will be published in January, apparently.

AI on the New TOEFL?

At the ETS Japan event about AI and human scoring last week, I asked again about the use of AI scoring on the new TOEFL. I hesitate to mention it (because the TOEFL team seems apprehensive about this issue) but I got the clearest suggestion yet that ETS will not send every test taker response to a specific human rater for scoring. This will be a tough row to hoe as ETS has spent the past five years extolling the virtues of having humans individually score every response. Anyway, I could be wrong; maybe they’ll stick with humans for the long run or on a transitional basis for the first few months of the test.

ETS Partners with Socra (Riiid)

I saw that Korean ed-tech Socra AI (known as Riiid until about four weeks ago) has entered into a licensing deal to create a TOEFL prep platform called Santa TOEFL. Socra says that the platform will “[integrate] official ETS content, providing learners worldwide with updated practice tests and learning solutions aligned with the January 2026 TOEFL revision.”

It is starting to look like new licensing deals will be a bigger part of the TOEFL’s revenue stream as ETS seeks to reboot and reinvent its flagship product. I wrote a few days ago about the Official TOEFL AI platform launched recently with a partner – as yet unnamed – in China. If this is indeed part of the strategy, it is probably a good idea. To my eye, test forms for the new TOEFL seem a heckuva lot cheaper to develop than test forms for the old version.

But this isn’t totally new. TOEFL already enjoys longstanding partnerships with Chinese firms like New Oriental who for years have paid big bucks to license retired test forms. They also license retired tests to Chosun Ilbo (weird choice) in Korea which are sold to consumers for $44 a pop. A partnership with the Japanese office of the Council on International Education Exchange to license the same test forms ended a few years ago.

Not that it matters, but Socra/Riiid is an interesting case. During the pandemic, that firm raised an eye-watering $170 million dollars of series-D funding from SoftBank, mostly on the back of a good-but-not-great AI-powered TOEIC prep app.** I think they used the money to buy up a more traditional English study platform, the management of which ended up running the whole merged company. Is that what they call a reverse takeover?

ETS 990 Forms Published

ETS’s 990 form for the year ending September 2024 is now available via Propublica. Some of this information was available in the audit published earlier this year, but much of it is new. A few things are worth mentioning:

1. Program service revenues were 908 million dollars. That’s down 10 million from 2023. It is the lowest in 15 years, as far as I can tell.

2. ETS finished the year in the black with a net income of 22 million dollars.

3. But that’s partially because they sold 95 million dollars of publicly traded securities. That’s the biggest sell-off of assets since they unloaded Prometric in 2018.

4. Salaries paid to key executives roughly doubled to 22 million dollars for the year.

5. CEO Amit Sevak saw his pay more than double, reaching 1.7 million dollars for the year. That included an incentive based bonus of about 500 thousand dollars.

6. Two other current executives crossed the million dollar mark.

7. Three former executives were paid more than a million for the year.

8. The biggest of the buyouts are listed this year. They are pretty big.

9. PSI Services LLC (Domestic) is on the books as “ETS Digital, LLC.” It is listed as being worth 553 million dollars. Apparently it lost money during the year.

10. PSI (Domestic) was given a loan or loan guarantee by ETS to the tune of 223 million dollars.

11. Here’s a puzzler: there is a thing on the books called EDUCATIONAL TESTING SERVICE DOMESTIC HOLDINGS INC. This has been around for some time. It is a C corp. Usually it is valued at a few million dollars. Now it is valued at 502 million dollars. What does it contain? I don’t know. All of the existing subsidiaries (Vericant, Kira, Pipplet, etc) are listed separately, so it isn’t them. I suppose it consists of investments made by ETS Capital (Applyboard, etc) and other securities.

I published some commentary on this data on LinkedIn.

Want to Support This Newsletter?

So I keep getting financial pledges via substack from people who want to support this newsletter. That’s really awesome, but due to draconian banking regulations here in Korea, it is unlikely that I will ever be able to turn on paid subs and claim those pledges.

If you really want to support my work, feel free to visit my Ko-Fi page, where you can easily make a donation via credit card or PayPal or something else. Ko-Fi is actually better than Substack for creators. One cool person has already made a donation.

Photo of the Week

The story moved me to write, once again, about the question of synchronous vs asynchronous proctoring of at-home tests.

You can read what I produced over here (just scroll to the third item).

My thesis on that day was that while both methods include the use of human proctors to ensure the security of a test, live human proctoring of remote tests raises concerns related to both privacy and overall test taker experience.

Following the publication of that item the testing firm in question terminated the proctor’s employment and reached out to the test taker with an apology. While it is nice to get a win now and then when it comes to test taker advocacy, it is impossible to celebrate when the details surrounding the win are so unsettling.

The HOELT and Remote Proctoring

This month’s Request for Information notes that the Home Office is exploring “remote proctoring as the primary mode of delivery” for the HOELT, a test which will soon be the only way those seeking UK visas can prove their fluency in English. Since this may usher in an era of even greater use of remote testing for high stakes purposes, I want to use this soapbox to, once again, say my bit about proctoring.

I’ll begin with a snippet of what I wrote back in July:

…the days of live proctoring are likely numbered. It seems increasingly obvious to me that asynchronous proctoring (AI-based security during the test combined with a human review after the test) is the future of at-home testing. This is Duolingo’s current approach. ITEP uses it too. It appears that Pearson will follow suit when they launch their new “Pearson English Express” Test later this year. I assume that many other test makers are currently trying to figure out how to implement their own async systems. A decade from now, I don’t think any at-home English tests will utilize live proctors.

That was July. In September, one is left to speculate whether the requirements of the HOELT tender might push back wide scale adoption of async proctoring to a much later date. After all, the Home Office may not be following this issue as closely as we are. They may be insisting on approaches to test security that were best-practice five or ten years ago.

So here are a few things decision makers may wish to keep in mind. I apologize to regular readers, who have been watching me beat this particular horse for half a decade.

What is the Impact on Personal Privacy?

As I’ve described above, live proctoring raises concerns about test taker privacy. Many live proctoring systems currently in use require proctors to visually confirm the identification documents provided by test takers. As remote proctoring positions are increasingly offshored and outsourced, I worry about the ability of testing firms to screen the individuals who have access to this sensitive information and to conduct background checks as needed. Async proctoring doesn’t eliminate this concern entirely, but it does vastly reduce the volume of individuals who are required to view personal information of this type.

Smooth Sailing or Choppy Waters?

A traditional defence of live proctoring is that it ensures a smooth and professional test experience. But five years into the era of wide scale remote testing, I’m not sure I buy that argument anymore. Social media abounds with stories of proctors who have actually impeded the performance of test takers.

I’ve heard about proctors working from public transit. I’ve heard about proctors forgetting to mute their microphones, subjecting test takers to the sounds of their own busy office. I’ve heard about proctors giving test takers permission to break rules, which often results in a score cancellation somewhere down the road. I’ve heard about proctors interrupting test takers at inopportune moments to insist on a mid-test room scan while the clock continues to tick away. The list goes on.

I’ll pause now and illustrate this point with a quote from someone who took one of the most widely administered at-home tests a few months ago (you can probably find the original post on Reddit):

“Just took the at-home test today. My nearest testing location is a bit of a drive, so I thought it would be easier to take it at home. Big mistake on my end! I can’t speak for everyone, but I had a lot of issues mostly with the proctors. The first proctor I had couldn’t get their microphone to work, and that took about 45 minutes to figure out. I finally got fed up and asked to switch proctors. I got connected to somebody else and thankfully it worked, but they were really mean and I almost cried because of how frustrated I was. It took about an hour and 15 minutes before I finally could start the test.

I am not particularly good at taking standardized tests, so I knew it was going to be difficult. I actually felt adequately prepared before I took the test. But my testing anxiety plus the technical difficulties caused my brain to shut down and I just couldn’t answer the questions. When I retake it, I’m just going to suck up the drive and take it in person so I don’t have to deal with those technical problems.”

Here’s a screenshot (forgive me for the wall of text) from someone who struggled with mid-test room scans during a major at-home English test:

Here’s one more. Reading about how the proctor insisted on getting a peek at the test taker’s balcony and even their building’s lift left me feeling uneasy:

A friend has informed me that getting illicit peeks at elevators is a fetish. I should get new friends.

Anyhow, I’ve got many more stories if you want ‘em. I’ve been saving screenshots for years.

Now, I get that these are all just social media anecdotes. But five years of reading similar reports has left me unconvinced by the argument that live proctoring provides the promised benefits in terms of a smooth testing experience.

And, of course, these issues could be addressed through better proctor training. But I am still left wondering about the impact of turnover, offshoring and outsourcing among the proctor population. How much does that impede the ability to retain a properly trained staff? How can staff quality be monitored during a tender that lasts a decade or more? A better approach is, possibly, an intuitive UI that eliminates the need for human guidance in the first place and a forgiving approach to mid-test technical failures that allows for a quick re-test.

Just How Secure is it?

Most readers are already aware that proctors aren’t watching test takers on a 1:1 basis. Instead, a single proctor might watch multiple test takers simultaneously. Sometimes as many as 16 to 18, according to some reports.

Too many demands on a proctor’s time may impede their ability to provide a seamless testing experience. But even more worryingly, it is unclear how much security a proctor working under such conditions can provide. Accordingly, even when live proctoring is implemented, the most meaningful security checks are likely performed after the test, when the session is reviewed by a proctor working under more comfortable conditions. Which is something that the developers of async proctoring systems realized long ago.

That’s All

There you have it. This is likely my final word on remote proctoring, as I’ve been repeating myself for some time. If you wish to know more, I’ll leave you with a few links to recent publications in the academic press:

• Belzak, et al. in “Educational Measurement.”

• Isbell, et al. in “Language Assessment Quarterly”

And the last few years of my ongoing blog posts about remote testing issues:

• This Should Not be Happening (August 28, 2025)

• On Async vs Live Proctoring (August 1, 2025)

• Inappropriate Proctors and the Future of Remote Testing (July 30, 2025)

• Home Edition Hilarity (September 20, 2024)

• “We’re Not Trying to Test Their Patience” (March 17, 2024)

• “Educational Measurement” on Remote Proctoring (January 31, 2024)

• First Steps for Better Remote Proctoring (January 17, 2024)

• Dan Isbell on Remote Proctoring and Justice (January 15, 2024)

• Live Proctoring can be Frustrating (October 20, 2023)

Back to regular programming next week, I promise.

Want to Support This Newsletter?

It was pointed out that i yesterday’s letter I didn’t actually link to the Ko-Fi page I recommended readers use to support the newsletter. Here’s a repeat of that item, with the link in place:

So I keep getting financial pledges from people who want to support this newsletter. That’s really awesome, but due to draconian banking regulations here in Korea, it is unlikely that I will ever be able to turn on paid subs and claim those pledges.

If you really want to support my work, feel free to visit my Ko-Fi page, where you can easily make a donation via credit card or PayPal or something else. Ko-Fi is actually better than Substack for creators. One cool person has already made a donation.

Meanwhile, here’s a couple of items that didn’t make it into yesterday’s letter…

Cambridge Will Not Accept New TOEFL Scores

The University of Cambridge (undergraduate) will not accept scores from TOEFL tests taken after December 31, 2025. According to the university's website:

"changes to TOEFL Internet Based Test and Home Edition in January 2026 mean that it will no longer be suitable for entry to Cambridge"

Regular readers know that the TOEFL iBT will be heavily revised in January. As far as I can tell, Cambridge is the first school to signal that it will not accept the revised version.

Duolingo Gets a Room Scan

I saw in this Youtube video that the Duolingo English Test now requires a room scan. The required scan is a 360 degree motion (with a phone) akin to British Council’s EnglishScore test, plus a look at the space behind the computer.

I’ve never been a fan of room scans. I find them cumbersome and I’m unconvinced of their effectiveness. For what it’s worth, this one seems a bit less obtrusive than most of the scans I’ve carried out. It seems unlikely to result in a request to drag furniture around my house or to take down all of my Stone Cold Steve Austin posters.

I’ll take the test later this month and report back with details.

One wonders if this nod to more traditional at-home testing is a reflection of Duo’s new status as a co-frontrunner for the HOELT tender. Note also the reminder in this week's THE article that every DET test is reviewed by a human proctor (something that not everyone is aware of).

Subscribe now

The UK Home Office has published a fifth request for information regarding the Home Office English Language Test (HOELT). This one is a shocker. It notes that “the Home Office is exploring a ‘Digital by Default’ service, with remote proctoring as the primary mode of delivery and physical test centres available where remote solutions are not feasible.”

This could explain the curiously low number of test centers mentioned in the fourth RFI, which is again listed as just 268 in 142 countries.

A Home Office choice to go with remote proctoring by default might favor a smaller test provider – like LANGUAGECERT, Duolingo or ETS – heretofore considered an underdog in the race to win the tender. All three of those providers are well known for offering robust remote-testing options to test takers around the world.

On the other hand, the IELTS partnership (widely considered a front-runner to win the HOELT tender) currently offers remote tests only in select markets, while Pearson (another favorite) pulled the plug on its remote options back in 2024 shortly after stories broke about widespread cheating on the at-home PTE Test.

Of course this doesn’t mean remote testing is a sure thing. But it is worth paying careful attention to the possibility.

Want to Support This Newsletter?

So I keep getting financial pledges from people who want to support this newsletter. That’s really awesome, but due to draconian banking regulations here in Korea, it is unlikely that I will ever be able to turn on paid subs and claim those pledges.

If you really want to support my work, feel free to visit my Ko-Fi page, where you can easily make a donation via credit card or PayPal or something else. Ko-Fi is actually better than Substack for creators. One cool person has already made a donation. Thanks, Arno!

Widescale IELTS Cheating in Bangladesh

The Daily Sun (Bangladesh) has published a long article on IELTS exam leaks and organized cheating. I’ll post a link in the comments.

According to their report, two suspects have been arrested in Dhaka for their part in a cheating ring. The Sun reports that the suspects used leaked IELTS tests to prepare students for the test.

Says the article:

“Students who followed the racket’s instructions reportedly achieved their expected band scores, raising serious concerns over the integrity of the IELTS examinations in the country.”

This comes after a lengthy Daily Sun investigation of a separate IELTS cheating ring operating on the same principle. Basically, it is alleged that ringleaders acquired test forms the night before specific administrations of the IELTS test from a leaker operating within the IELTS partnership. After that, customers were (allegedly) secreted away to local hotels where they had 4.5 hours to memorize the correct answers. The next morning they (allegedly) took the test at a variety of test centers across the city.

Says the article:

“On 25 April, a correspondent stayed at Hotel Afford Inn in Uttara, where around 100 students were lodged overnight. Leaked questions were reportedly handed out at mid-night, after students were separated into teams according to their exam centres. The next morning, the students were transported by minibuses, microbuses, and CNGs to [test] centres including Patronas at Panthapath, Compass at Banani, IALC at Dhanmondi, and Penstone at Uttara. A similar operation was observed on 23 May at Hotel Central Inn in Motijheel, where 120-130 students were being prepared for the following day’s exam.”

Beneficiaries of this scheme confirmed (allegedly) that the leaked questions were identical to those on the real IELTS.

The Daily Sun report suggests that the aforementioned cheating rings “highlight serious flaws in exam security.”

The security of the paper-delivered IELTS has come to the forefront in recent months. Note that the IELTS Partnership pulled the plug on paper delivery of IELTS for UVKI in Bangladesh at the end of July.

The paper-delivered IELTS has also been eliminated or restricted in key markets of Vietnam, Pakistan, Sri Lanka, Malaysia and Nigeria. Meanwhile, test takers globally have recently been forbidden from accessing the paper-delivered test outside of their country of residence (or country of citizenship).

Some have speculated that these moves are connected to longstanding security concerns about the paper-delivered IELTS.

Earlier this year the UK Home Office raised concerns about possible cheating on IELTS tests, citing comments from HEIs that IELTS scores “don’t necessarily replicate the reality of the applicant’s English language ability.”

If you are still reading, do take a moment to peruse the whole Sun article. It contains the most succinct description I’ve ever read of the ol’ paper-delivered IELTS cheating technique (allegedly).

IDP Education Financial Results

IDP Education Ltd has released Full Year Financial results for FY 2025.

IELTS volumes are down 18% for the year for a total volume of 1,293,800 tests. Last year’s volume was 1.5 million tests. The peak for IDP was 2022, when 1.9 million tests were administered.

IELTS volumes are down 50% in India for the year. This comes after a 42% drop in India back in 2024.

Outside of India, IELTS volumes are up 2%. This is attributed to onshore testing in Canada and Australia.

IDP failed in its goal to launch IDP IELTS in China in the fiscal year.

Shares are way up since the release, perhaps because IDP has promised to significantly reduce costs.

Office Hours #5

Office Hours #5 will focus on the testing (and test prep) scene in Southeast Asia. I’ve arranged for short interviews (10-15 minutes) with experts based in Vietnam and Indonesia. Regular readers know that Vietnam is one of the fastest growing markets for English tests, and that student recruiters think of Indonesia as their next big gold mine. I’m interested in multiple perspectives from these markets, so if you are based in one of them I hope you’ll join us. And, of course, people from all over the world with an interest in these markets should attend.

Register here.

I would love to have a speaker from a third country in the region. Let me know if you might be able to contribute. A big test market like the Philippines is the logical choice, but I’m just as interested in hearing about what’s going on in smaller countries.

On top of that, we can talk about any relevant testing news that has broken since the last Office Hours.

The event will occur September 26. Just note that since this is a Zoom meeting (not a webinar) I can’t send reminders. You should manually mark the day in your calendar app of choice. As always, the chat will be totally casual and mostly free-form. The event will not be recorded.

How Testing Works in China

I read that the Oxford Test of English will soon be offered at test centers in China and Thailand. Congratulations to the team at OUP! That’s one step closer to my backyard!

China is an interesting case. Regular readers know that the two big English tests in that country – IELTS and TOEFL – are administered in partnership with the NEEA, a public agency associated with the Chinese Ministry of Education. Test registration is done through an NEEA portal. The NEEA collects registration fees and later passes them along to ETS and the British Council… after pocketing an unspecified amount.

This is probably good for test takers as the NEEA is known for blocking price hikes and upselling. It also mandates that registration fees be charged in RMB. It probably isn’t great for testing firms, as the NEEA is known for blocking price hikes and upselling. It also mandates that registration fees be charged in RMB.

Not all language tests are required to operate in partnership with the NEEA. This is because there are two types of language tests in the eyes of the Chinese regime – we might translate them as “educational tests” and “commercial language proficiency exams.” The former must be administered in partnership with the NEEA, while the latter can be administered with any local partner.

The biggest English test to operate without NEEA partnership is the PTE. Pearson partners with an on-shore company called Beijing Ensi (d/b/a Pearson VUE China).

In the case of Oxford, that partner will be a group called GEC.

A funny case is that of Prometric. They run the CELPIP with local partners, but run the CAEL with the NEEA. Which makes sense, as the CAEL is used for educational purposes, while the CELPIP is not.

There is a point to all of this. Anyone who cares enough to still be reading already knows that the IELTS test is run in China through a partnership between the British Council and the NEEA. IDP Education is not involved, but traditionally the British Council paid them a per-test royalty as a sort of booby prize in recognition of their partial ownership of IELTS.

Last year, IDP gave up that royalty and began administering the IELTS themselves. It is my understanding that their strategy was to brand themselves as a provider of a commercial language proficiency exam. This was done with an unnamed local partner referred to by CEO Tennealle O’Shannessy as “a respected professional examination service provider.” Someone once told me the name of that partner, but I’ve long forgotten. I could look it up, if anyone cares.

That went well for a few months. But by December, testing had ceased. IDP is now in negotiations to resume testing. I don’t know how those negotiations are going.

I’ve always wondered how Pearson managed to skirt the NEEA requirement. I know that it is used for both educational and non-educational purposes. Maybe that’s enough. But so are the IELTS and TOEFL tests, to some extent. And, surely, they would like to be free of the NEEA.

Too Much Testing in Australia?

Here’s a video from the Australian Broadcasting Corporation (ABC) about English tests in Australia and recent changes. Among other things, it touches on the fact that many visa holders in Australia test and re-test and re-test and re-test. Indeed, the recent financial results from IDP Education Ltd touched on how this kind of on-shore testing is a growth area for the IELTS.

New Duolingo English Test Demographic Data

Duolingo has published a new Demographic and Score Properties document, tracking Duolingo English Test takers for the year ending June 2025.

Notably, China now accounts for the largest percentage of test takers per test taker ID, at 12.88% of all test takers (down from 15.88% last year). India now accounts for the second largest percentage, at 12.43% (down from 18.91% last year).

Here are the top ten countries, with last year’s percentage in parentheses.

• China: 12.88% (15.88)

• India: 12.43% (18.91)

• Canada: 5.49% (4.35)

• USA: 4.23% (3.34)

• Brazil: 3.49% (3.42)

• Korea, Good: 3.39% (3.51)

• Mexico: 3.36% (3.12)

• France: 2.95% (2.09)

• Nepal: 2.71% (new to list)

• Pakistan: 2.40% (1.94)

Last year, China and India accounted for about 35% of all test takers. This year they account for about 25% of all test takers.

About 49% of test takers took the test to facilitate undergraduate studies, while about 35% took it to facilitate graduate studies.

Anyway. The document contains quite a lot of additional data. Do check it out. You can read about last year’s document over here.

Duolingo Scores Mapped to CLB…

A new study aligns Duolingo English Test scores with the Canadian Language Benchmarks.

…And to TOEFL and IELTS Section Scores…

Earlier this month the Duolingo English Test score comparisons page was updated. It now includes charts which compare DET and TOEFL subscores. This is useful information, but note that TOEFL subscores will switch to a new scale in January.

…Amidst Charm Offensive.

Apropos of various things, the Duolingo English Test team has published a charming new video highlighting their mission of “transforming testing for good.” It features Duolingo co-founder Luis von Ahn and others. I was happy to see DET creator Burr Settles make an appearance.

Duolingo is really good at this sort of thing. And it accounts for so much. Some people tend to miss that. Those who consume their output, though, are quite aware of it.

Is this video part of their HOELT charm offensive? Probably.

TOEFL Test and Score Data for 2024

The 2024 “Test and Score Data” summary for the TOEFL iBT is now available. This represents the first full year of the revised two-hour TOEFL test. A few notes:

The mean score (overall) is 86 points. That’s a one-point drop from last year.

The mean reading score is 21.9 (-0.4), the mean listening score is 22.1 (-0.2), the mean speaking score is 20.8 (-0.1), and the mean writing score is 21.1 (-0.1).

Among countries of interest (to me):

• The mean score in China was 85 (-1)

• The mean score in Korea was 86 (no change)

• The mean score in Japan was 72 (-1)

• The mean score in Brazil was 88 (no change)

• The mean score in India was 94 (+1)

• The mean score in Iran was 93 (-2)

• The mean score in the USA was 91 (-1)

Ontario Pharm Tech Programs Stop Accepting TOEFL Scores

Since July most of the popular pharmacy tech programs in Ontario (Canada) have stopped accepting TOEFL scores for admissions. This brings them in line with the National Association of Pharmacy Regulatory Authorities, which dropped TOEFL from its list of acceptable English tests in 2024.

See, for instance, Sheridan College, which I think has the most popular pharm-tech program in the country:

More Thoughts on TOEFL Essentials in Canada

I was quoted a few times in this PIE News article about TOEFL Essentials being approved by IRCC. I have a few thoughts:

1. Like most things in Canada, English tests are terribly expensive. As is often the case, this is partially due to a lack of competition. In introducing a whole new product to the market, ETS has the opportunity to really shake things up in terms of pricing.

2. Some readers may already know that Canada is a fairly large country. The introduction of a new test with a pre-existing network of test centers could mean the difference between some poor bastard taking a test close to home versus having to drive three hours to do so. Remember that ETS can offer the iBT and Essentials tests in the same test centers (and at the same time). That’s what Pearson does with the PTE-A and the PTE-Core tests.

3. ETS currently has a 0% share of this market, so the introduction of the TOEFL Essentials test is a huge win for the testing giant. This is notable because ETS doesn’t get a lot of wins nowadays. Kudos to the team that put this together.

4. Keep an eye on the specifications once the test is officially available for this purpose. It wouldn’t surprise me to see a few minor modifications get made. For instance, the test probably doesn’t need three adaptive levels, nor an adaptive writing section, nor an unscored video statement. The final version could more closely resemble the 2026 version of the TOEFL iBT.

What about TOEFL Essentials for Australia?

The Educational Testing Service funded a three-year study at the University of Melbourne “to examine the relevance of the task types of the new TOEFL Essentials test to migration to Australia.” I dunno how that study turned out, since it has not been published anywhere.

What about the Future of TOEFL iBT?

I want to cycle back to that pre-print from Maggie McGehee and Daniel Isbell which I wrote about a few weeks ago. Take a quick look at the data in Table 1. In the period under review (Fall ‘22 to Fall ‘23) 100 undergraduate students submitted ELP scores to the University of Hawaiʻi at Mānoa. Here are the tests they used:

• DET: 61

• TOEFL: 20

• IELTS: 19

And limiting ourselves to Fall ‘23:

• DET: 30

• TOEFL: 5

• IELTS: 8

So 61% of students submitting a score in the whole period opted to send a Duolingo English Test score. DET’s share reached 70% in the final semester under review, while the TOEFL had just a 12% share at that time.

What I wouldn’t give for numbers from 2019.

This is just a single less-selective school in the second most beautiful state in the union, but it speaks to the challenge that the TOEFL team faces as it implements a major relaunch of its flagship product.

That challenge being how to respond to a young person who knows that all his friends were admitted to good schools using a DET score, and who also knows that the DET is just a fraction of the price.

I wouldn’t know what to say.

Regardless, one can almost suss out the potential future (bright) of the TOEFL program. Some things come to mind:

1. Maybe US admissions are something of a lost cause, but there are a ton of jurisdictions that are reluctant to accept DET scores right now. Places like the big new receiving countries in Europe and Asia. Perhaps TOEFL could become the go-to “easy” test for these markets. I can envision it becoming something like the DET of the EU. I would mention Australia here, but the revisions have come too late to be accepted by the DHA.

2. On the other hand, maybe the domestic market can be recaptured. The new TOEFL is hella streamlined. Development, administration and scoring will all be much cheaper than for the old TOEFL. Why not turn TOEFL into a $75 at-home test and instantly recapture much of the market share lost to DET? Heck, even knocking the price down to a cool $99.99 would probably do the trick. A test center version could be maintained at a hefty price tag for folks who prefer it. To put it in somewhat nefarious terms, ETS could charge test takers in the lucrative Chinese market $300 a piece and let everyone else test for $100.

3. On the vestigial third hand, there are big players who greatly desire a bigger slice of the American pie. Pearson has gone to all the trouble of developing their own DET competitor (the PEET Test) which will launch next month. That test looks fantastic, but it will take an enormous amount of time, money and effort to gain widespread acceptance at institutions. It took Duolingo six years and a pandemic to gain a stable foothold. Why not just… well… I won’t say it. But you know what I mean. Everybody wins.

The cool thing is that all three of these avenues suggest a bright future for the TOEFL program. They just require boldness and fortitude.

TOEFL Compared to Duolingo English Test… One Last Time?

The ETS Research Institute has published a new study which compares the TOEFL iBT and Duolingo English Test (DET). It is dated April 2025 but is new to me and it appears to be new to the Internet. The study was carried out by Sara Cushing of Georgia State University.

Curiously, this study compares the DET to the current incarnation of the TOEFL iBT rather than the heavily revised version which will debut in January. One is left wondering if Professor Cushing was even told that she was researching a test that would cease to exist a few months following publication. Surely her time could have been better spent comparing the DET to the revised TOEFL.

The study is very critical of the DET. The author levels pretty strong criticisms at the test and the choices made by its designers.

Interestingly, many elements of the DET which are singled out for criticism will soon be found in the new TOEFL. Meanwhile, many aspects of the TOEFL which are specifically praised in the study will not be found in the new TOEFL.

For instance, Cushing praises the TOEFL for the inclusion of integrated speaking and writing tasks that reflect the environment students will find themselves in at university, for its use of lengthy reading and listening passages, for its exclusive focus on academic content, and for eschewing short item types that test low-level fluency.

In contrast, the DET is criticized for its lack of integrated speaking and writing tasks, for using too many short reading and listening passages, for testing students on non-academic content and for including short items that test low-level fluency.

Regular readers don’t need me to tell them that the revised TOEFL will not include meaningfully integrated speaking and writing tasks. Regular readers also know that it will include shorter reading passages than those currently found on the DET, and that many of them will be about non-academic subjects. And, needless to say, the new TOEFL will include many short items that test low-level fluency.

Do not take this as a complaint. The revised TOEFL is a wonderful test. I mention these things only to highlight how different it is from the old TOEFL and how much it resembles the DET.

In some ways, this report can be read as a sort of coda for the old TOEFL program, before ETS gave up the ghost and determined that the best way forward was to remake the test into something that very closely resembles Duolingo’s product.

Official TOEFL AI Prep Platform (China)

It is worth taking a moment to peer at the Official TOEFL AI Learning platform recently launched by ETS in China. It looks like TOEFL TestReady on steroids, and should be considered a pretty big deal.

Test takers have spotted advertisements for it at the big TOEFL test center in Shanghai. I’ll post a picture of one of those below. Note that purchase and registration requires a Chinese phone number. This isn’t meant for the rest of the world.

The price tag is an eye-watering 10,888 RMB (about $1500 USD). But given that spend on TOEFL prep in China can sometimes be quite high (and often comes from the Bank of Mom and Dad) I think it will be a pretty lucrative product for ETS. Also note that a lot of test takers spend two or three years preparing for this test.

An additional perk is that users of the platform can plug in their test reservation number and get their TOEFL scores a few days early.

What’s fascinating to me is that this is one of the biggest additions to the TOEFL portfolio in decades, and you’ll find not one mumbling word about it here on LinkedIn, on the ETS website or on the usual social media channels. I’m not sure anyone outside of the PRC even knows about it. This isn’t a complaint, of course. As regular readers know, I’ve been fascinated by the veil which separates test prep pros inside China and those outside of China.

I’m also vaguely curious how this platform will affect long-standing licensing deals which have sent dozens of retired test forms (along with SpeechRater and e-rater grading) to China-based partners like New Oriental. Will they be continued now that ETS is operating a competing platform? A decision to forego those deals and instead funnel all Chinese learners into a self-operated platform could be super-duper lucrative.

In any case, this all seems to be part of the strategic pivot back to the Chinese market which I’ve written about again and again.

TOEFL Institutes Fee to Update User Profile

The TOEFL program has instituted a “country change fee.” If test takers who have completed a home edition test wish to change the address in their TOEFL account to a new country they are now required to pay a $99 “country change” fee. I repeat: the fee is charged if you want to change your address after taking the test.

Needless to say, this is the silliest thing I’ve heard all month.

The pop-up message explaining the fee (pictured) from ETS is so poorly worded as to be incomprehensible. But test takers have reported that declining to pay the fee after changing their address has resulted in them being locked out of accessing (or sending) any existing TOEFL score reports.

As many readers know, the TOEFL test is often taken by individuals seeking to study at an educational institution outside of their country of origin. As a result, it is not uncommon for test takers to change their address sometime after completing a TOEFL test.

Photo of the Week!

I got this TOEFL prep kit in the mail a few months ago. Dated 2003, it is probably the last official prep product sold by ETS which targeted the original version of the TOEFL before it transitioned to the TOEFL iBT in 2005.

Duolingo’s Q2 results for 2025 were published a couple of weeks ago. The Duolingo English Test had revenues of $10,088,000 for the quarter. Since the test costs $70, we might estimate that it was taken about 144,114 times in the quarter. But this is a very rough estimate because not everyone pays $70. Some pay extra for faster results, some pay less by purchasing a bundle or buying through a partner organization. Others pay nothing thanks to the DET Access Program (which gave away about 25,000 free tests in 2024).

In Q2 of 2024 the test had revenues of $10,968,000 at $65 a pop, representing something in the ballpark of 164,584 total tests.

This suggests about a 12% decline in volume from last year.

Pearson Reports Interim Results

In my last email, I failed to mention Pearson’s interim results, which were published August 1. They reported that PTE Volumes were down in H1 by about 10% compared to H1 of last year. That likely represents a total of about 491,000 tests for H1 2025.

Volumes of all the big tests are down this year, which is not a surprise.

IDP Education is scheduled to publish full year results at the end of this month. ETS will likely file a 990 form later this month. That’ll be available to the public whenever the Internal Revenue Service gets around to processing it.

TOEFL Essentials Set for Canada

ETS has been named a recognized testing organization by Immigration, Refugees and Citizenship Canada, and the TOEFL Essentials Test will soon be accepted for immigration purposes. A press release from ETS confirms that the test will be offered at testing centers at that time.

ETS will be in competition with the IELTS General, PTE Core and CELPIP tests in this space.

This is a big win for ETS. Testing for Canadian immigration is a pretty big market, and ETS has basically none of that market at the moment.

More on TOEFL Raters

Speaking of TOEFL, the PIE News has dug into the issues surrounding a petition launched by TOEFL speaking raters who are seeking clarity regarding their status with ETS.

According to the petition, scoring of TOEFL speaking responses has largely been offshored to India, a process which began in December of last year. This comes after about twenty years of speaking responses being scored solely by raters working in the USA.

A scoring leader who went on the record with The PIE noted: “[t]he email that came in December was well crafted… they said they were going to be including global raters and that some would see their shifts increase and others would decrease.”

And also: “[c]ome February, lots of raters would email me and say they were only scheduled for five days out of the month, some having registered their availability for every day… [c]ome March, they might get 10 days for the month, but a day and a half before their shift, they’d get an email saying that half of their shifts were cancelled.”

That scoring leader noted that “shifts for US test raters had all but dried up in April, though test scorers training the new global raters continued receiving shifts.”

Those training shifts had mostly concluded by June.

The article describes some quality-related concerns that said scoring leader (along with an anonymously quoted leader) has with the change.

The PIE contacted ETS for comment, who responded, in part:

“In response to these changing demands, we made the strategic decision in late 2024 to expand our scoring capabilities beyond the US. This shift allows us to better serve a global testing population, increase operational flexibility, and uphold the quality and efficiency our customers expect.”

—

Personally, I’m sure that the overseas raters will eventually perform up to the same standards as the former American workforce. That said, it seems like the workforce in question has been treated somewhat shabbily in recent months.

LSAT Cheating out of Control

The LSAT will be suspended in China for an indefinite period of time following the October administration of the test. This is due to excessive cheating in that market.

The Law School Admission Council (LSAC) has indicated that testing may resume when they can figure out how to offer in-person testing in China, but I don’t think that will be an easy task. Currently, all LSAT testing in China is remote, with proctoring services provided by Prometric.

One can easily grasp the relevance of this story. Recall that Pearson pulled the plug on at-home language testing in China back in 2023 for the same reason. Shortly thereafter they quietly eliminated at-home testing entirely.

Updated IELTS Test Stats Now Available

IELTS test statistics for 2024/25 are now available (I think they were published earlier this month). Interested readers can now access updated demographic and test taker performance data. Nothing unusual caught my eye, but let me know if you see something interesting.

Transparency in Testing

There is an excellent article in the Review of Education by Nathaniel Owen and Ananda Senel about enhancing transparency in high stakes English tests. I encourage everyone to check out the whole thing, but I want to jump to the end of the article, which includes a few recommendations for test makers. I’ll paraphrase them below (forgive me for any errors in comprehension). They call for:"

1. Increased transparency when it comes to how much time is allocated to test takers. This means explaining the rationale behind the amount of time given, and suggesting how test takers can handle time-related limitations.

2. Better communication regarding technology and related quality assurance. This includes clear system requirements and troubleshooting instructions.

3. More streamlined communication with test centers and with test takers. This includes clear in-test instructions and explanations of features, like adaptivity, which can sometimes be complex.

4. Increased cultural and linguistic accessibility. This could include providing instructions in local languages for test center staff, and recognizing the challenges that test takers might have with accent variety.

5. Transparent communication of how fairness is embedded in test design.

6. Accurate resources that clearly explain the demands of the test and demystify it in a general sense.

These are pretty solid recommendations. Nobody is doing all of them right now, but experienced test watchers will probably recognize that the tests which are on the upswing in terms of market share are doing more of them than the tests which are on the downswing.

I think most of us will agree that test takers often struggle to decode the requirements of their chosen test. I’ve written here how the TOEFL scoring rubric tells test takers to use “a variety of syntactic structures.” I am a great admirer of Professor Chomsky, but this is not a useful phrase when it comes to a scoring rubric.

I’ve also written about how – with a few exceptions – at-home testing has been pretty frigging terrible. Test takers are regularly told that their tests have been terminated due to reasons that the test owner won’t actually reveal to them. Remote proctors seem to enforce rules in a haphazard way, depending on whether or not they actually know what they are. At least one at-home test has included on-screen instructions requesting that its rules be broken. This is not transparency.

People who have been reading my work the longest will recall the “Saga of the Altruism Question,” which is pretty much my origin story as a commenter on testing. I won’t go into it here, but do encourage you to revisit the linked blog post. In some way, transparency has been the throughline of everything I’ve written over the past 15 years.

There is no regulatory body ensuring transparency in language testing, so test makers can be as transparent or as opaque as they like. But if they want to stand out in an increasingly competitive field, increased transparency is probably a smart choice.

Me on Livestream

Here’s me on the TEFL Development Hub Livestream.

Duolingo’s Interactive Writing Task

Here’s a post from the official Duolingo English Test blog about the DET’s multistage interactive writing task. It summarizes a long article in Language Testing. Twenty-five page explorations of individual tasks give me great joy.

The article explains that this task is an attempt to address the fact that “most large-scale writing assessments do not reflect real-world and classroom-based composing, where writers develop their ideas over multiple drafts.”

In the DET’s task, test takers are first given an initial prompt on a specific topic. Next, AI is used to identify which pre-identified themes the test taker’s response has addressed. Finally, a follow-up prompt is presented which asks the test taker to address one of the themes they have not touched on. It’s slightly more nuanced than my summary suggests, but I think you get the point.

On the livestream linked to above, I talked about how some test takers prefer human interlocutors because they can potentially tease out their English skills with follow-up questions, making it possible for them to get full credit for their fluency in the language. Here we have a style of item that could perhaps replicate the same effect in writing assessment. The authors note that “[w]ith the advancement in AI technology, Interactive Writing using prompting that is specifically keyed to the content of an initial response offers greater opportunities for measuring test takers’ full writing abilities…”

Scenario Based Language Assessment

Here’s a video from Soo Hyoung Joo of Columbia University about her work on scenario-based language assessment. It describes how an AI character can help test takers deal with a scenario (about physics, in the example case). The performance and collaboration of the test takers can then be observed and assessed.

English Tests and Foreign Currency Loss

A new article from Vientnam.vn touches on the issue of “foreign currency loss” stemming from widespread use of international English tests for domestic university admissions in Vietnam. It estimates that in 2025 over 300,000 IELTS tests will be taken in that country. Each administration of the IELTS costs VND4,664,000 (about $177 USD), the majority of which is said to flow outside of the country. The author also touches on how the cost of test preparation exacerbates this issue.

A domestically developed alternative to the IELTS, known as the VSTEP, is available but not widely accepted or used. Vietnam is one of a few markets where IELTS testing volumes are still increasing year after year.

In earlier posts I’ve written about how Seoul National University in Korea developed the TEPS Test partially due to concerns that overuse of the TOEFL and TOEIC tests for domestic admissions were leading to currency outflows at a particularly sensitive time (the late 1990s IMF Crisis). By all accounts, the TEPS Test successfully reduced the use of foreign tests for domestic purposes. Eventually, universities were blocked from using tests in this way and this all became a non-issue.

Interestingly, ETS has long been blocked from selling its “TPO” practice tests for the TOEFL directly to consumers in Korea. Instead, they must be purchased via a local partner. This was a big deal in the earliest days of the TOEFL iBT test before the development of a robust network of domestic prep providers.

Connecting English Scores to Academic Outcomes

Some readers might be interested in an article by Maggie McGehee and Daniel Isbell which will appear in a future installment of “Studies in Language Testing.” It examines the relationship between English test (DET, TOEFL, IELTS) scores and academic outcomes at the University of Hawai‘i at Mānoa (UHM).

According to the article, “mean GPAs and proportions of student withdrawal/probation were similar for the DET and other tests (or no test), and no differences among them were statistically significant.”

I was happy to see the authors note that “[f]or students admitted unconditionally with higher ELP scores, there were no statistically significant correlations between ELP scores and first year GPA.”

As has been discussed in this space, there are many things which impact student outcomes. The authors note the existence of students with really high ELP scores, but surprisingly low GPAs. Go figure.

Interestingly, though, it is mentioned that “[f]or those admitted conditionally, with ELP scores reflecting a lower range of English proficiency, stronger, positive, and statistically significant correlations emerged for students submitting IELTS or TOEFL scores, with low or negative (but not statistically significant) correlations for DET takers.”

It is pointed out a few times that the sample size was pretty small, so further study is needed.

There is also some good stuff in here about the business of English testing, but I’ll touch on that in another post. Check out Table 1 for a preview, though. You’ll spot which tests are frequently used… and which are not.

I Took the iTEP Test!

I took the iTEP test a few weeks ago. You can find some comments over on the blog.

Picture of the Week