Retrieval practice (the testing effect) Spaced repetition
Fifth graders remembered more history a week later after quizzing themselves than after rereading
In two fifth-grade classes in Italy, children read short history texts and then either reread them or filled in gaps from memory, checking and correcting their answers until all were right. A week later, in a surprise open-question test, the self-testing children remembered clearly more. Spreading the study session four days after the first reading, instead of one day, made no measurable difference.
| Paper | Retrieval practice enhances learning in real primary school settings, whether distributed or not |
| Authors | Laura Franzoi, Veronica Cembrani, Claudio Mulatti, Barbara Treccani |
| Published | 8 August 2025 · Frontiers in Psychology 16:1632206 (Brief Research Report) |
| Study type | Quasi-experimental classroom study (two intact classes) · Peer-reviewed |
| What we read | Full text read. We read the open-access full text. |
| Summarized |
What they did
Over about seven weeks of normal history lessons, the classes went through four phases, each with a new 590 to 660-word text on ancient history. After a first reading, children studied again either 1 or 4 days later, for up to 30 minutes, by rereading or by a fill-in-the-gap test with feedback until 100% correct. Seven days later they had 10 minutes for an unannounced open-question test that did not count towards grades.
Who took part
33 children aged 10 to 11 in two fifth-grade classes (15 and 18 pupils) at one primary school in Italy, taught by their usual teachers.
What they found
- Self-testing beat rereading in phases 1 and 2: 67.0% vs 48.6% correct on the final test (F(1, 25) = 41.42, p < .001).
- Phase 3 confirmed it with the same spacing for both: 67.7% for testing vs 41.3% for rereading (η² = .254).
- Spacing on its own made no significant difference: 45.9% after a 4-day gap vs 44.6% after a 1-day gap (F < 1).
Effect size: Testing vs rereading: η² = .154 (phases 1–2) and .254 (phase 3). Spacing: not significant.
Limitations
- Small sample from one school, with no power analysis.
- Classes, not individual children, were assigned to conditions, and phases 3 and 4 were not counterbalanced.
- Only a one-week delay; the spacing contrast (1 vs 4 days) was modest, and other spacing studies do find benefits.
- Testing included feedback until every answer was right, so the benefit combines retrieval with correction.
What this means for learners and parents
- For 10 year olds, a short "fill in what you remember, then check" round beats reading the page again.
- Getting every answer right in the end, with feedback, is part of what made it work.
- Don't read the null spacing result as "spacing doesn't matter"; it is one small study with a small gap. Little-and-often remains well supported overall.
Put it into practice (free tools)
Citation
Laura Franzoi, Veronica Cembrani, Claudio Mulatti, Barbara Treccani (2025). Retrieval practice enhances learning in real primary school settings, whether distributed or not. Frontiers in Psychology 16:1632206 (Brief Research Report). https://doi.org/10.3389/fpsyg.2025.1632206
Related summaries
- Quizzing boosts classroom learning: a meta-analysis of 222 studies · Chunliang Yang et al., 2021 · Systematic review and meta-analysis
- The classic study: testing yourself beats rereading after a week · Henry L. Roediger III et al., 2006 · Lab experiments
- Spacing beats cramming: the classic review of 317 experiments · Nicholas J. Cepeda et al., 2006 · Review and meta-analysis