Transfer of Test-Enhanced Learning: Meta-Analytic Review and Synthesis
Pan, S. C., & Rickard, T. C. · 2018
grade Cmeta-analysisindependentunreplicatednumbers spot-checked
Sample
192 effect sizes / 122 experiments / N=10,382
Population
Mostly lab and classroom; verbal/factual and some applied materials.
Design
The definitive transfer meta for retrieval practice. Categorizes transfer by type (format, context, untested-but-studied material).
Key findings
Retrieval transfer is real overall (d=0.40) but concentrated in response-congruent formats (~0.58); transfer to untested-but-studied material is the WEAKEST category (~0.15-0.20) and near-zero without feedback. Genuine far transfer thins sharply — retrieval mainly strengthens the material it directly tested.
Genetic confound
Retention/transfer outcome; no g.
Replication notes
Anchors the response-congruency critique: retrieval mainly re-expresses aligned material.
DOI / URL
10.1037/bul0000151
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Overall transfer of test-enhanced learning | d | 0.40 [0.31,0.50] | researcher-designed | mixed; delayed | active-alternative | unclear | near-transfer |
| Response-congruent format transfer | d | ~0.58 (largest) | researcher-designed | delayed | active-alternative | unclear | near-transfer |
| Transfer to UNTESTED-but-studied material | d | small/weakest (~0.15-0.20), near-zero without feedback | researcher-designed | delayed | active-alternative | unclear | near-transfer |
Cited by
- Retrieval practice (the testing effect)strong supportconf: highgc: low