I spent much of this month traveling.  On Pacijan Island in the Philippines I stayed at an accomodation called “Camotes Cay Hideaway.”  That’s a one-room property with a really nice view of the sea.  I understand that it was built about thirty years ago as a summer getaway for a chiropractor who practiced in Cebu City for many years.  He passed away recently and his getaway is now used by anyone who needs a quiet escape for a few days.  I think the furniture still in use was all his… and also his books remain.  I snapped a picture:

And another:

Judging from his collection of textbooks, this fellow was the oldest of old school chiropractors.  I not sure I believe in the efficacy of this particular school of thought, but I wish I could have gotten the “flying seven” from him before his passing.

Perhaps after I pass into the next world by collection of books on the history of ETS will be read by strange tourists forevermore.

Not a whole lot of relevant reading this month.  Sorry!  But a few things are worth mentioning:

  • Still following along with the Norton Library Podcast, I read Mary Shelley’s “Frankenstein.”  For a fairly old book, it’s still really accessible.  I would recommend it to language learners who want exposure to some classic literature.  It is easy to find, but make sure to get a copy with basic annotations to guide you through the tricky parts (and to explain some of the many literary allusions in the text).  You could even get the super cheap Wordsworth Classics Edition of the book, which probably has enough notes for most readers.
  • I read the August 29, 2023 issue of The New Yorker, which included a reprint of a very long 1979 profile of the silent film star Louise Brooks.  Film fans might enjoy this one.  Others may not.
  • I put in a purchase request at my local library for the 5th edition of They Say, I Say. To my surprise, the library got a copy!  This is a lovely book that can be a lot of help to students beginning their university life who need guidance about writing argumentative essays.  This edition includes a new chapter on revising essays (which is a welcome addition) and a chapter on writing research essays (which should probably be the subject of a whole ‘nother book).  It also includes a couple new sample essays.  I’ve already endorsed the fourth edition of the book in this blog, but I mention the new one here just in case anyone wants to ask me questions about it.  Note that they didn’t get the “with readings” version, so I don’t know about changes to those.

That’s all for now.  But more in September.  Keep reading.  Let me know if you have any recommendations.  I’ll track them down eventually.

Continuing my interview series, here’s a chat with Martin Chan, who has been teaching TOEFL prep in China for about ten years. This interview went long (and is really inside-baseball at times) but it is a fun one because TOEFL is thriving in China and there is a lot of demand for his services.  Below is the video.  After the jump are a few highlights. 


1. From Martin’s perspective, the TOEFL Test is still the most popular English test in China, though the IELTS is at about the same level.

2. He doesn’t see a ton of demand for help with the Duolingo and Pearson tests, and from his perspective their popularity may have declined a bit since pandemic restrictions were lifted.

3. People love the recent changes to the TOEFL Test, but some people worry about how errors are now weighted in the reading section.

4. 1:1 classes seem to be the most common prep method in China.

5. We tried to explain the reasons for increases to the average TOEFL score in China. We speculated that it might be a result of technology, great prep resources, and changes to the demographics of test takers. Note that this conversation occurred before the release of 2022 test score data (which showed another significant increase).

6. More specifically, we talked about how much students in China benefit from greater access to retired TOEFL tests and to implementations of the SpeechRater and e-rater AI (all licensed from ETS).

7. It is still really hard to book seats at test centers in China (for all English tests). Many students resort to buying seats from resellers who charge an extra fee.

8. The NEEA. What’s up with that?

9. We discussed the ban on private tutoring (for regular school subjects) that went into effect back in 2021. This is still impacting the tutoring industry. But, interestingly, it has created new business for TOEFL and IELTS prep as it is something of a workaround for parents who want their kids to do well in their regular English classes.

10. Some parents in China continue to consider study abroad destinations other than the USA due to the dreadful state of relations between those two countries.

IDP Education has just published its financial report for the year ending June 2023. Some highlights:

  1. IDP delivered 1.93 million IELTS tests during the year. That’s up 1% from 1.91 million tests last year.
  2. Revenue from testing is up 7%. Revenue growth is up mostly because the average cost of taking the IELTS increased by six percent during the year.
  3. IELTS revenue in India is down 5%. No word on the decline in testing volume, but it must be more than 5%, given price increases. I guess the other companies are successfully chipping away at IDP’s dominance in this market. Good for them.
  4. IELTS revenue is up a big 18% outside of India. Testing volumes are up 22% outside of India.
  5. These figures do not include IELTS tests administered by the British Council. Remember that IELTS is a two-headed beast, and that test takers can opt to get the test from IDP or BC (though not in every market).  Those figures will be available later this year.

It is fascinating to see a test like the Pearson PTE grow by 50% in a year, while the IELTS grows by only about 1%. We are certainly witnessing a sea change in the way language testing is done. I predict a more consumer-friendly environment moving forward (but I fear consolidation).

TOEFL score data for 2022 is now available!  Get the data over here.

The mean TOEFL score for all test takers in 2022 was 88, the same as 2021.

  • The mean reading score was 22.8 (+.4)
  • The mean listening score was 23.0 (+.4)
  • The mean speaking score was 20.9 (-.2)
  • The mean writing score was 21.6 (+.1)

Among the countries I track closely…

  • The mean score in China was 90 (+3)
  • The mean score in Korea was 86 (no change)
  • The mean score in Japan was 73 (no change)
  • The mean score in Brazil was 89 (-1)
  • The mean score in India was 95 (-1)
  • The mean score in the USA was 93 (no change)

That big jump in China comes as no surprise.  Scores in China have increased significantly since changes to the TOEFL in 2019 which shortened the test and introduced an at-home option.

Here are per-country changes since 2018 (the last full year of the old four-hour TOEFL, which was available only at test centers).

  • China: +10
  • Korea: +2
  • Japan: +2
  • Brazil: +2
  • India: no change
  • USA: +3

I did a follow-up interview with my friend and TOEFL Tutor Han Joon Lee about some of the teaching methods he uses in Korea. This one might be a bit too inside-baseball for some, but I think a few readers here are interested in how test prep is done around the world. A few highlights:

  • Overall, he thinks the TOEFL is a pretty good test. Test-takers build their fluency while preparing for the test. They do tend to improve in all four key skills.
  • He really likes the massive vocabulary books that are available in Korea. Memorizing those 2000 words is helpful for the reading section. This somewhat surprised me.
  • He thinks that practice reading sets are useful, but test-takers can go overboard. Instead of blowing through 100 sets before taking the test, it is often better to use just a few dozen and examine them at a sentence-by-sentence level. Test-takers might consider spending a couple of hours on a single reading passage and really understand the function of each sentence.
  • Dictation and shadowing are key for the listening section. It can be useful to spend a few hours working with just a short passage.
  • Non-TOEFL conversation practice is really, really useful. Seek it out.
  • Feedback, feedback, feedback. People need speaking and writing feedback before they take the test. Tutors need to be harsh.
  • Student buy-in matters. Students need to understand that high scores require a lot of work. This doesn’t seem to be a problem when it comes to test-takers in Korea.

I’ve got a great spit-take at about 17:50.

Distressing news today regarding Pearson’s PTE Academic Test. Everyone ought to read this report in The PIE News, but I’ll provide a few highlights:

  1. Some universities have raised concern about test-takers getting “full or very high scores” on the at-home version of the PTE Academic Test.
  2. Pearson has begun withdrawing and cancelling scores of some test-takers.
  3. Some universities in the UK have withdrawn offers from applicants with canceled scores. Students in China and South Asia seem most heavily impacted.
  4. Several universities in the UK have stopped accepting scores from the at-home version of the PTE Academic test, including the University of Edinburgh, the University of Sussex and the University of Southampton.
  5. Pearson has stopped offering the at-home PTE test in China.

There is a lot of cheating on high stakes tests taken at home. Not just for English tests, of course, but for all tests. Testing companies are working hard to stop it, but I’m not sure it will ever be eliminated completely. As long as test-takers are using their own Windows or Apple machines to take the tests, remote access will be an ongoing concern. Even Amazon, which obviously has access to infinitely more resources and talent than Pearson, has been pretty frank about the amount of cheating that goes on in its at-home AWS certification exams.

Duolingo just reported its Q2 results! Revenue from the Duolingo English Test was 9.80 million dollars in the quarter, up from 8.03 million dollars in Q2 of last year. Note that the cost of taking the test is now $59, up from $49 in 2022.

At $59 a pop, we can assume that the test was taken about 166,254 times in the quarter. The real number is probably a bit higher, given the presence of occasional discounts and freebies. This is about the same as Q2 of last year.

I was a bit surprised by the drop in both revenue and test-takers compared to Q1 of this year, when the company reported revenue of 9.97 million (from perhaps 203,000 tests).

That said, historic SEC filings suggest that Q2 is generally soft for the DET (Q2 saw the fewest DET test-takers in both 2021 and 2022).

It has been argued that the new “Writing for an Academic Discussion” task requires test takers to write fewer words overall, every grammar or language use mistake they make will have a greater impact on their score.  I figured I would run a few tests in the e-rater to see if this is the case.

First up, this response got a score of 5.0:

“I like the ideas noted by Claire and Kelly, but I feel that the only way to truly solve this problem is to build better schools.  Parents want their children to be educated at the best possible places. It is almost impossible to find great schools in the country. The government currently has money for good things like science laboratories and nice libraries in cities so they should not ignore rural areas all the time. When rural schools like the one I attended lack even basic educational supplies like computers and sports equipment, parents who are concerned about their kids go to bigger places.”

If I submit the same response with all of the commas removed I get a score of 4.0.

If I submit the original response but with ONE spelling mistake (truely) I get a score of 4.0.

So does a small number of individual errors have the potential to impact the score of a test taker?  It would seem so.

However, it isn’t as clear cut as it seems.  While the human raters give only whole number scores, it seems like there the e-rater is using decimals under the hood.  I imagine that my original response is a “low 5” answer (maybe just a 4.5, rounded up).

Here is an answer where I retain the spelling mistake, but beef up my vocabulary usage:

“I respect the ideas noted by Claire and Kelly, but I feel that the only way to truely solve this problem is to construct better schools.  Parents want their children to be educated at the best possible facilities. It is almost impossible to find great schools in the countryside. The government currently provides funding for beneficial things like science laboratories and lavish libraries in cities, so they should not ignore rural areas all the time. When rural schools like the one I attended lack even basic educational supplies like computers and sports equipment, parents who are concerned about their kids depart for bigger places.”

This time, my score goes back to 5.0.  I can use the same technique to overcome the penalty for not using commas.  Actually, I can use the same technique to overcome the problem of not using conjunctions, which I have written about elsewhere.  I bet it works both ways.

Decimals, right?  I gain a few tenths of a point for improving my vocabulary and that compensates for the few tenths of a point I lost for the spelling mistake.  We’ve seen a similar thing in some implementations of the SpeechRater AI.

Anyway.  This might be useful for people preparing for the test.  Test-takers ignore the quirks of automated scoring at their own peril.

I got a report from a TOEFL Home Edition test-taker that made me very happy. The individual launched the test, and during the checkin procedure a pop up appeared asking them to shut down a specific application that was running on their system (in this case, I think it was Asus Hotplug Controller). They quickly did so, and the test began.

Easy as that.

From 2020 until three weeks ago, I received report after report about tests simply being terminated, with no reschedule or refund granted.  Those made me very sad.  

Of all the changes to the TOEFL implemented by ETS on July 26 this is the most important. By far. More than anything else, this change will ensure the long term viability and success of the TOEFL iBT. As I said above, I’m happy.

One question remains: were tests terminated in the past because of innocuous things like the Asus Hotplug Controller?  I frigging hope not.  If anyone has an answer, though, you know how to reach me.

Those changes to the Student Direct Stream (SDS) are now in effect. The IELTS monopoly is broken, and students may now also use scores from the TOEFL, PTE, CAEL and CELPIP tests. This is very, very good for students. Monopolies, as we know, are bad.

I just noticed one other change, which is that students using IELTS scores are no longer required to achieve a score of 6.0 in each of the four skills. Now they only need to achieve an overall score of 6.0. That’s probably too low, but I guess it won’t really matter too much as few schools (other than the sketchy degree mills) will admit students with band scores below 6.0.

I also noticed that students submitting an IELTS score need 6.0, students submitting a TOEFL score need an overall score of 83. That seems a bit high. According to the conversion tables published by ETS, an IELTS score of 6.0 matches a TOEFL score of 60.

Similarly, students submitting a PTE score need an overall score of 60. But according to the conversion tables published by Pearson, an IELTS score of 6.0 matches a PTE score of 46.

Again, this probably doesn’t matter since accepting schools do a better (but not perfect) job at properly matching score requirements but I think it is worth noting.

Pearson recently published its half year results for 2023. They include a report that the Pearson Test of English (PTE) was taken 606,000 times in the half year, way up from 344,000 in the first half of 2022.

As I’ve written here in the past, the growth of the PTE is really amazing. Remember that for all of 2019, it was taken just 547,000 times.

I imagine that the PTE is now the second most popular English test, by test-taker volume. If the PTE Core (nee “PTE Essential”) variation gets approved for more markets, Pearson could possibly move into the top position.

One day I will make a blog post that lists the estimated popularity of all of the major English tests.