Retrieval Practice
Test yourself, don't just re-read
What the Research Says
Retrieval practice means pulling information out of your memory rather than putting it back in. Every time you successfully recall a fact, the memory trace gets stronger. Re-reading your notes feels productive, but it creates a false sense of confidence — the information looks familiar without actually being learned.
Evidence Base
| Study | Finding | Impact |
|---|---|---|
| Roediger & Karpicke (2006) | Students who practised retrieval remembered 80% after one week; those who re-read remembered 36% | More than double the retention |
| Dunlosky et al. (2013) | Major review of 10 revision strategies rated practice testing as HIGH utility | One of only two top-rated techniques |
| Karpicke & Blunt (2011) | Retrieval practice outperformed concept mapping for learning science texts | Works even for complex material |
| EEF Cognitive Science Report (2021) | Recommended retrieval practice as one of six key principles for effective learning | Endorsed for UK classroom use |
Step-by-Step Method
-
1
Read one topic — Spend 5 minutes reading through a single lesson or section. Don't highlight or make notes yet — just read to understand.
-
2
Close your notes — Put everything away. No peeking. This is the crucial moment where learning actually happens.
-
3
Brain dump on a blank page — Write down everything you can remember about the topic. Use bullet points, sketches, equations — whatever comes to mind. Aim for 5–8 minutes.
-
4
Check and highlight gaps — Open your notes and compare. Use a different colour pen to fill in anything you missed or got wrong. These gaps are your priority for next time.
-
5
Repeat on the gaps — Next session, focus your retrieval attempt on the parts you couldn't remember. Keep cycling until the gaps are filled.
When and How Often
One full cycle takes about 18 minutes. Do 2–3 cycles per revision session, covering different topics. Use retrieval practice every time you revise — it should be your default method, not an occasional extra.
The atomic model changed over time. First there was a solid sphere model — I think that was Dalton. Then Thomson discovered the electron and made a plum-pudding model. Rutherford did an experiment that showed there was a nucleus in the middle. Then Bohr added orbits. Chadwick discovered neutrons at some point. The modern model has a nucleus with protons and neutrons, and electrons in shells or energy levels.
The atomic model changed over time. Dalton (early 1800s) proposed a solid sphere — all matter made of indivisible atoms. Thomson (1897) discovered the electron using cathode rays, producing the plum-pudding model (positive "dough" with negative electrons embedded). Rutherford (1909–1911) fired alpha particles at gold foil — most passed straight through, but some bounced back at large angles, proving there is a tiny, dense, positively charged nucleus. Bohr (1913) refined this by proposing that electrons orbit the nucleus in fixed energy levels (shells) and can move between them by absorbing or emitting specific amounts of energy. Chadwick (1932) discovered the neutron by bombarding beryllium with alpha particles, explaining why the nucleus was heavier than protons alone could account for. The modern model treats electrons as a cloud of probability rather than fixed orbits.
What this reveals: The first attempt had the sequence right but missed every date and the mechanism of each discovery. Dates (1897, 1909–11, 1913, 1932) and the gold-foil experiment details become the focus of the next retrieval cycle.
Common Mistakes
-
Re-reading instead of recalling — Reading your notes five times feels like revision but barely strengthens memory. Close the book and test yourself.
-
Peeking at notes too soon — The struggle to remember is the point. If you look at your notes the moment you get stuck, you skip the part that builds memory.
-
Only testing what you already know — It feels good to recall easy facts, but the real gains come from practising the topics you find hardest.
-
Not returning to the same topic — One brain dump is a start, not the finish. You need to revisit each topic multiple times at increasing intervals (see Spaced Repetition).