In my seventh year of teaching, I ran an experiment I should have run in my first. I split my classes: one group reviewed notes before the test, the other took low-stakes practice quizzes on the same material. No extra study time. Same content.
The quiz group scored 23% higher.
I felt sick. How many students had I failed by not knowing this sooner?
The Most Replicated Finding in Learning Science
Retrieval practice — pulling information from memory rather than reviewing it — has an effect size of 0.93 according to Hattie’s meta-analysis of educational interventions [1]. For context, an effect size of 0.40 is considered the threshold for “significant impact.” Retrieval practice is more than double that.
The landmark study: Roediger and Karpicke (2006) had students study prose passages. One group re-read the material. The other practiced recalling it. After one week, the retrieval group remembered 80% versus 36% for the re-reading group [2].
That’s not a marginal improvement. That’s the difference between mastery and forgetting.
Why Re-Reading Feels Effective (But Isn’t)
When you re-read your notes, something insidious happens. Your brain recognizes the information — “Oh yes, I remember this” — and creates an illusion of knowing [3]. Recognition and recall are completely different cognitive processes. You can recognize something without being able to recall it, the way you can recognize a face but not remember the name.
Highlighting is even worse. Dunlosky et al. (2013) reviewed ten popular learning strategies across decades of research. Highlighting and re-reading were rated as having “low utility” — the two most common study methods are among the least effective [4].
How I Implemented It (And What Went Wrong)
First attempt: 10-question quiz every class. Students hated it. It felt like punishment. Compliance dropped. I almost abandoned the idea.
What worked: reframing it as a brain dump, not a quiz.
“Take 3 minutes. Write everything you remember about yesterday’s lesson. No notes. No judgment. Just dump.”
Then compare with a partner. Then I reveal the key points. Total time: 6 minutes. No grades. No stress.
The results showed up on unit tests within three weeks.
The Practical Playbook
Daily (5 min): Brain dump at start of class. “What do you remember?” No notes, no textbook. Compare with partner.
Weekly (10 min): 5 low-stakes quiz questions from the past week. Show answers immediately. The act of trying to remember — even getting it wrong — strengthens memory [2].
Unit-level: Every test includes 20% questions from all previous units. This forces spaced retrieval, preventing the learn-and-forget cycle that makes cumulative exams feel impossible [5].
Student Resistance Is Data
Students will say it’s hard. They’re right. Retrieval practice feels harder than re-reading because it is harder — and that difficulty is the learning. Bjork calls these “desirable difficulties” [6].
Show students their own data after three weeks. When they see their scores improving, resistance evaporates. I’ve never had a student complain about brain dumps after the first month.
The Question That Keeps Me Up
We’ve had this evidence since 2006. Roediger’s work has been replicated dozens of times across ages, subjects, and cultures. Yet the dominant study method in most schools is still “read the chapter again.”
Why? Probably because retrieval practice feels wrong. It feels like you’re failing when you can’t remember something. But that feeling of failure is literally your brain building a stronger trace. The research is clear. The implementation is simple. The only barrier is inertia.
References
[1] Hattie, J. Visible Learning. Routledge, 2009. Updated meta-analysis: visiblelearningmetax.com
[2] Roediger HL, Karpicke JD. “Test-enhanced learning.” Psychological Science, 17(3), 249-255, 2006. DOI: 10.1111/j.1467-9280.2006.01693.x
[3] Koriat A, Bjork RA. “Illusions of competence in monitoring one’s knowledge.” Journal of Experimental Psychology: Learning, Memory, and Cognition, 31(2), 187-194, 2005.
[4] Dunlosky J, et al. “Improving students’ learning with effective learning techniques.” Psychological Science in the Public Interest, 14(1), 4-58, 2013. DOI: 10.1177/1529100612453266
[5] Cepeda NJ, et al. “Distributed practice in verbal recall tasks.” Psychological Bulletin, 132(3), 354, 2006.
[6] Bjork RA. “Memory and metamemory considerations in the training of human beings.” In Metacognition: Knowing about knowing, MIT Press, 1994.