Test-Enhanced Learning: Taking Memory Tests Improves Long-Term Retention
Roediger, H. L., & Karpicke, J. D. · 2006
grade Brctindependentreplicatednumbers spot-checked
Sample
2 lab experiments, undergraduates
Population
Undergraduates, lab; ~250-word prose passages; free-recall outcome.
Design
The canonical crossover demonstration. Study-only vs study+test (no feedback); final free-recall at 5 min / 2 days / 1 week.
Key findings
THE crossover result: at 5 minutes restudy beats testing; the benefit of testing emerges ONLY on delayed tests, where it is large (~61% vs 40% at 1 week). Retrieval is a DURABILITY method, not an immediate-performance one.
Genetic confound
Retention outcome.
Replication notes
Crossover widely replicated; underpins 'desirable difficulties.' Caveat: lab prose, response-congruent free recall, no transfer measure.
DOI / URL
10.1111/j.1467-9280.2006.01693.x
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Final recall at 5 MINUTES | % recalled | study WINS (~81% vs ~75% for testing) | researcher-designed | immediate; lab; proximal | active-alternative | end-of-treatment | domain-skill |
| Final recall at 1 WEEK | % recalled | testing WINS (~61% vs ~40% for study) | researcher-designed | delayed 1wk; lab; proximal | active-alternative | under-1yr | domain-skill |
Cited by
- Retrieval practice (the testing effect)strong supportconf: highgc: low