Testing Effect

Category: Memory

The Testing Effect is the finding that retrieving information from memory (taking a test, using a flashcard, recalling an answer) locks it in far better than passively re-reading it, even though re-reading feels like the smarter move.

How it works

When you retrieve a fact instead of re-reading it, you force your brain to reconstruct the memory, and that act of reconstruction strengthens both the memory and the route back to it. Rereading skips this entirely: it feels smooth and familiar, which your brain misreads as "I know this," a metacognitive trap Roediger and Karpicke labeled the illusion of competence. The harder the retrieval (short answer beats multiple choice, recalling from a blank page beats recognizing), the bigger the long-term payoff, as long as you eventually get feedback on what you missed. The effect is strongest on delayed tests: right after studying, rereading can actually look better, which is exactly why people keep choosing the strategy that loses in a week. Rowland's 2014 meta-analysis of 159 comparisons pinned the average boost at g = 0.50, a solid medium effect that shows up across word lists, prose, lectures, and real classrooms.

Where you'll see it

  • Karpicke and Roediger (2008, Science): four groups learned 40 Swahili-English pairs. The groups that kept getting tested on every pair recalled about 80% a week later. The groups that stopped testing items once correct (and just restudied them) recalled 33 to 36%. Restudying a known item was nearly worthless; retrieving it was everything.
  • Roediger and Karpicke (2006): students read a prose passage, then either reread it or took a free-recall test. Rereading won on an immediate test. But after two days and again after a week, the tested group crushed the reread group, and the rereaders had confidently predicted they would remember MORE. That is the illusion of competence in one experiment.
  • Medical and law students live this: Anki and other spaced-flashcard apps are built entirely on the testing effect. The USMLE and bar-exam crowds who grind active recall cards outperform peers who reread outlines and highlight, despite the highlighters feeling more productive.
  • In real college courses, the Adesope, Trevisan, and Sundararajan (2017) meta-analysis of 188 experiments found practice quizzes beat restudying by about d = 0.51, and beat doing nothing by d = 0.93. Low-stakes weekly quizzes, not just final exams, drive the gain.

Where it comes from

This is not a 21st-century discovery. Francis Bacon noted it in 1620, and William James mused on it in 1890. An early demonstration came from Edwina Abbott's 1909 work, where recall practice beat pure restudy for memorizing material. Arthur Gates (1917) ran it at scale, testing schoolchildren on biographies and finding that time spent reciting from memory beat time spent rereading. Herbert Spitzer's 1939 study of about 3,600 (3,605) Iowa sixth-graders confirmed it in classrooms. Then the idea mostly sat dormant for decades until Henry Roediger and Jeffrey Karpicke revived it with their 2006 Perspectives on Psychological Science paper "The Power of Testing Memory" and their 2008 Science paper, which turned "the testing effect" into one of the most replicated findings in learning science.

How to counter it

Close the book and produce, do not skim. After each section, shut the material and write everything you can recall onto a blank page from memory. The blank page is the whole point: recognizing an answer you are shown is easy and teaches you almost nothing, while dragging it out of an empty head is the reconstruction that locks it in.

Score yourself against a delay, not against fluency. Rereading feels smooth right after studying and that smoothness is a lie about what you will remember in a week. Quiz yourself a day or three later, then grade what you actually got right, because the only number that counts is unaided recall on a delayed test.

Always check what you blanked on. Retrieval without feedback lets you rehearse your own errors, so after every recall pass go back and mark exactly what you missed or fudged. Feed those specific gaps back into your next quiz instead of restudying the whole section.

Ban the highlighter and the reread. The moment you think "I already know this, I'll just look it over again," that is the illusion of competence talking. Turn that section into a question you have to answer cold instead, and only move on once you can generate it without the page in front of you.

The tell

You catch yourself thinking "I already know this, I'll just skim it again" while highlighting or rereading, and it feels smooth and effortless. That fluency is the tell. Easy, familiar, and confident is exactly how material feels right before you blank on it in the exam.

Related biases

References

  1. Roediger, H. L., & Karpicke, J. D. (2006). The Power of Testing Memory: Basic Research and Implications for Educational Practice. Perspectives on Psychological Science, 1(3), 181-210
  2. Karpicke, J. D., & Roediger, H. L. (2008). The Critical Importance of Retrieval for Learning. Science, 319(5865), 966-968
  3. Rowland, C. A. (2014). The Effect of Testing Versus Restudy on Retention: A Meta-Analytic Review of the Testing Effect. Psychological Bulletin, 140(6), 1432-1463
  4. Adesope, O. O., Trevisan, D. A., & Sundararajan, N. (2017). Rethinking the Use of Tests: A Meta-Analysis of Practice Testing. Review of Educational Research, 87(3), 659-701
  5. Gates, A. I. (1917). Recitation as a Factor in Memorizing. Archives of Psychology, 6(40), 1-104