ELT: Teach to the Test!

Alright, so if you balked at the title, I admit that I’m not really going to suggest that you teach to the test. That can undermine your pedagogy and compromise your values. But in adult ed and other ESOL settings, we often find ourselves in a tricky predicament: to satisfy funders or bureaucratic stipulations (BS, for short), we have no choice but to use standardized tests to demonstrate our students’ progress.

And these tests, often, are, unfortunately, well, forlackofabetterword—alright I’ll just come out and say it—bad. No need to name names or acronyms, but the problem fits a common profile: We see our students progress drastically, and they feel that progress themselves, but when the results of these tests come in, that progress just isn’t reflected. It’s frustrating. It’s disheartening. In some cases it can be of great consequence to the future of the program.

So though I’m not going to tell you to teach to the test, if you’re stuck with a test that simply isn’t working, you’ve got to do something. I want to propose that we can teach toward the general direction of the test, teach testward, in a testerly direction, teach north by northtest.What I propose is that we can teach in a way that ensures alignment between testing and your course without contravening your teacherly ethics. I’ve employed this method myself, with excellent results, teacherly ethics entirely intact.

The idea isn’t all that complicated or secret. Plenty of teachers out there are surely doing it already.

Analyze the Test

Take a look at the test and determine the original intent. However bad the test may be, some serious experts spent a lot of time designing it to assess some skills that somebody deemed to be important. And they likely weren’t completely off-base when they did so. Run through the test and identify the major areas of focus, both in terms of language and content. What skills and strategies are being tested and how are they tested? If there’s a listening section, are students listening for details or main idea? Maybe both? Are the questions selected response or constructed response? Are they straightforward or do they demand some level of inference? How are their answers scored? Do grammar and mechanics count?

Compare With the Syllabus

Now that you’ve got a sense for the test, take a look at your own course objectives, syllabus, and materials. Hopefully most of what is assessed in the test is covered in your course, but most likely there will be some minor holes and discrepancies. You will almost certainly identify some problems with the test, even more than you had seen before (and here it’s good to cut the test publishers a little slack and remember that assessing language in a quantitative way is an endeavor that involves some fundamental, insurmountable challenges).

You may also, however, notice some oversights or discrepancies in emphasis in your own syllabus, materials, or activities. For instance, I once found that the listening section of the standardized test my students had to take was so challenging for my students because the recordings used in the test involved more authentic language with speech disfluencies such as hesitation phenomena and false starts, even distracting background chatter, whereas the listening sections of our core text were built around clearly articulated complete sentences, more akin to written language than natural speech.

Make Appropriate Instructional Shifts

It’s discrepancies like these that we want to focus on: The issue was relatively easy to address (we incorporated some podcasts as supplementary listening texts), and the change was a clear improvement. Not all shifts will be as easy as this one, but you’re seldom going to come across problems whose solutions are completely unacceptable to you. It’s going to be to emphasize grammar a little more when scoring, to incorporate more of this kind of activity, or to introduce this structure a bit earlier in the curriculum.

If you come across some larger discrepancies, there are ways to mitigate the need for larger instructional shifts. One of the tests that I was once stuck with put what I considered to be undue emphasis on the what-I-consider-to-be-useless-in-the-year-2015 skill of letter writing. A nice skill, to be sure, but when prioritizing the many skills my high beginning students needed to develop in their 6 class hours per week, letter writing ranked pretty low. Moreover, points were awarded (or not) based not on arbitrary conventions (like the placement of the date and mailing address) that have little if anything to do with English language proficiency.

I certainly wasn’t about to scrap any of our course content to include a unit on letter-writing. On the other hand, digital literacy is a component in our program, as are career readiness and systems navigation skills. It wasn’t much of a stretch to compare the conventions of email with those of paper correspondence. Likewise, it wasn’t a big problem to introduce a rudimentary version of a cover letter a bit earlier in the scope and sequence than we otherwise would have. We also had students bring in letters from utility companies and immigration services as relevant reading texts when appropriate in the units they were studying.

Though the discrepancy between test content and course objectives was larger in this case, a handful of small changes scattered throughout the curriculum was plenty to make sure our students were able to demonstrate their skills on the required test.

Also teach some basic test-taking skills, and teach them in the context of other valuable, closely related or transferable skills, such as forms literacy or following written instructions.