I often hear things like this:

I’ve been speaking English for twenty years, and I only got a 78 on the TOEFL.  The test is bullsh–t.

And things like this:

My crazy uncle Bob is a native speaker, and he only got a 22 in the reading section.  The test is obviously unfair.

But here’s the thing.  The TOEFL is not just an English test.  I know, “TOEFL” is supposed to stand for “Test of English as a Foreign Language.”  But check the ETS website.  They don’t use that name anymore.  You won’t find it in the Official Guide anymore.  These days, TOEFL doesn’t stand for anything.  It’s just the “TOEFL Test.” 

Even though Uncle Bob is a native speaker, it is likely that he’s incapable of functioning in an academic environment.  Indeed, most native speakers aren’t.  I suspect if you pull 100 random people off the street in the USA and give them the same TOEFL reading set, their average score will be quite low.

Keep that in mind before you get frustrated by your TOEFL score.  The score is meant to predict your performance in an English-medium university.  That’s it.  It isn’t a test of how well you communicate in English in general.  

This is what test designers refer to as “validity.”  They argue that you can’t just toss a bunch of grammar and vocabulary questions on a test (like the original TOEFL) and expect the scores to be useful for any real purpose.  The questions need to have a connection to the users of test scores, and be valid for their intended purpose.

In 2008 ETS published a 370 book to make this case (“Building a Validity Argument for the Test of English as a Foreign Language,” Carol A Chapelle, et al).  Ask your TOEFL teacher if they have a copy.

In a more recent book, Chapelle says:

Test developers, researchers, and anyone responsible for assessing human capacities would readily agree that validity is their central concern.  Similarly, teachers, employers, students, parents and researchers want the tests they use to be valid, and they expect professionals in educational and psychological testing to know how to evaluate a test’s validity. (“Argument- Based Testing in Validation and Assessment,” Carol A Chapelle)

This is why the TOEFL is really hard, and this is why the TOEFL is really popular with university administrators.

And this sort of thing is why there are a whole bunch of different tests.  The IELTS General Test is intended to assess language skills needed by immigrants and people in non-academic training.  The TOEIC test is meant to assess the potential of people to function in business settings.

Heck, ETS seems to be all-in on this concept.  They recently purchased Pipplet, which makes boutique language tests.  Including tests for: customer service agents, consultants, retail workers, medical professionals, startup employees… and more!  Some day they’ll have a different test for everyone!

What Does This Mean for Students?

It means that you need to prepare for the TOEFL.  Don’t just count on your amazing English skills. You can’t just prepare for a couple of days and expect an amazing score.  You need to study. 

The Future

Some people think this is an old-fashioned idea.  Maybe they are right. Maybe language tests won’t be so hung up on this conception of validity in the future.

In fact, according to the IPO documents recently published by Duolingo, that company wants its “Duolingo English Test” to be used for university admission, immigration AND workforce placement!  The idea of an identical test for all three purposes seems antithetical to the above definition of validity.  This sort of idea is why the Duolingo folks seem to be able to get a rise out of ETS in a way that the IELTS and Pearson people cannot.

Notify of

Inline Feedbacks
View all comments