Re-learning that really works?

Three years ago, under a new principal, my high school leaped head first into the “re-learning, re-take” trend sweeping public education.

The philosophy is simple and has undeniable logical appeal. If Michael performs poorly on a test, he should have the opportunity to re-learn the material and improve his score. After all, we get more than one attempt on the driver’s test, the ACT and the bar exam. Why should one test, one moment in time, define what Michael knows about the structure of the brain or the use of figurative language?

The trouble, as always, came with the implementation. I won’t get into the time-consuming process of writing dozens of new, equivalent tests – that’s a subject for another day.

The real trouble is the “re-learning”. Teachers know how to present lessons. We explain new material, provide practice opportunities, assign reading, offer review quizzes, and at our best, we engage students in active learning and simulations to make it stick.

But if all of that doesn’t work – evidenced by a student’s 55% grade on a unit test – what are we supposed to do next? What exactly is this “re-learning,” and how much re-teaching does it require?

This is where things got vague. Our administration deferred to us as professionals, which is both a compliment and a trap. I like being trusted to know what to do, but not when I actually don’t know what to do.

I struggled with this process semester after semester, trying lots of ideas with my subject-matter colleagues, finding nothing that helped more than a handful of students improve their test performance. The rest either stayed the same — or performed worse.

The first strategy I tried was to have students review the original test and explain their misconceptions. That didn’t work, so the next semester I asked students to identify the learning objectives they hadn’t mastered and complete additional practice. That was mostly paperwork.

Then I required tutoring time. I spent hours before and after school trying to re-explain concepts, but honestly I didn’t know how to make “re-teaching” different from the original, unsuccessful teaching. If they’re still not getting it, then what?

Then I read Make it Stick, and it planted a seed. The authors explained that students typically perform poorly on tests because they don’t know how to study. Most students study by re-reading and highlighting the textbook or notes, and research has found these strategies completely ineffective.

What my students need, I realized, is not necessarily new instruction, but new ways to work on retention. The authors offer lots of terrific research-based advice, but one strategy that stuck with me was self-questioning. You learn better if you write yourself a practice test, wait a few hours, then force yourself to take it.

I was so excited about this finding that I shared it with my students right away at the beginning of the semester. I encouraged them to self-test before the first test. They ignored me.

Then I realized, there is a better time to teach study skills: Re-learning time. After a low test grade shows up in the grade book, many students feel a newfound sense of urgency about study skills.

Last week, I was finally able to implement this new re-learning strategy – which I’m calling “20 Questions” – in my AP Psych class. The results were definitely encouraging.

Here’s how it works: After tests are handed back, students have the opportunity to review the original test and identify learning objectives they have not yet mastered. Then, their job is to write a new test for themselves, measuring their own understanding of just those objectives, using all open-ended questions.

For example, if Michael missed three questions about defining variables in a research study, he might write a question like: Describe a psychology experiment and identify the independent and dependent variables.

Students must show me their questions before they answer them, and I help make sure their questions are clear and specific – in other words, answerable. On their own, they take the test, grade it, and bring it in when they are ready for the re-take.

Last week, 10 of my 35 1st hour students took the first Psych re-take, and 9 of them improved their scores. The average improvement (including the one student whose score fell) was 8 percentage points, and the median was 10. Here are a few specific score improvements

  • 66% to 90%
  • 86% to 100%
  • 50% to 63%

The results are preliminary, of course, but I have never seen this kind of success with any other re-learning method — not even close.

I realize I have barely begun to test this process. I’d like to do a controlled study: randomly give half of the students a different retake strategy and half of the students the “20 Questions” strategy, to see if the results hold up. Unfortunately, my sample size is pretty small, and I don’t want to disadvantage any students by giving them a poor strategy.

For now, I will continue to use “20 Questions” and see if I get the same results. It would be such a relief to finally have a re-learning strategy that really works.