In The Science of Repetition we saw that repetition is non-negotiable. But research has a less famous finding: repetition is at its weakest when it is done the same way every time. Changing the context a little each time — the place, the sentence, the task, the order — measurably improves both recall and the ability to use what you learned in new situations.
Divers and two rooms
Godden and Baddeley (1975) had scuba divers learn word lists on land and underwater. Words were recalled about 40% better in the environment where they were learned: land-learned words on land, water-learned words underwater. Memory encodes information together with the context it was learned in. That's not good news — in real life you never meet a word in the same place twice.
Smith, Glenberg and Bjork (1978) showed the fix. Students who studied a word list twice in the same room recalled an average of 16 words when tested in a third room; students who studied in two different rooms recalled 24. Same time, same list; only the setting changed, and recall rose by 50%. Instead of being "locked" to one context, the knowledge was tied to several cues.
Motor skills: the group that never practiced the target hit it best
Kerr and Booth (1978) trained children to toss beanbags. One group practiced only at a 3-foot target; the other practiced at 2 and 4 feet and never tried 3. On the 3-foot test, the variable group was more accurate. Shea and Morgan (1979) found that people who practiced three movement sequences in blocks (AAA BBB CCC) did better during training but worse on the test, while those who practiced in a mixed order (ACB BAC…) did worse in training and clearly better on the test. This is the contextual interference effect: struggling during practice is a sign of durable learning.
Robert Bjork named this phenomenon "desirable difficulties". Variability forces the brain to rebuild the solution every time, and rebuilding is retrieval itself.
Cognitive learning: learning painters by mixing them up
Kornell and Bjork (2008) showed participants paintings by 12 artists. One group saw each artist's works back to back; the other saw the artists interleaved. When asked to identify the artist of paintings they had never seen, the interleaved group was clearly better. Strikingly, most participants believed the blocked schedule had worked better. Rohrer and Taylor (2007) found the same effect with math problems: students who practiced with mixed problem sets scored more than twice as high on a test one week later as those who practiced in blocks.
Language: the same sentence or different sentences?
- Webb (2008): Learners who met a target word in 10 different sentences scored higher on meaning and usage tests than learners who met the same word 10 times in a single sentence. Equal repetitions; the difference was contextual variety.
- Bolger, Balass, Landen and Perfetti (2008): When rare words were presented in varied contexts, learners built a more abstract, transferable sense of the word's meaning; a single context "glued" the meaning to that one sentence.
- Nakata and Suzuki (2019): Learners who studied semantically related words in an interleaved order remembered them better than those who studied in blocks — contextual interference applies to vocabulary too.
The logic: if you always see a word in the same sentence, your brain memorizes the sentence, not the word. If you see it in a restaurant, in a job interview and in a news headline, the one thing they share — the word's meaning — is what stands out.
In practice: how to vary your repetition
| Variable | Weak repetition | Strong repetition |
|---|---|---|
| Sentence | Always the same example | A new sentence each time, ideally one you wrote |
| Task | Always flipping cards | Build from the definition, spell from audio, fill a gap, use it in a dialogue |
| Order | Topic-by-topic blocks | Words from different topics mixed |
| Direction | Always English → your language | Both directions, alternating |
| Setting | Always at a desk | On the train, walking, out loud |
Word Hype's four modes implement exactly this table: the same word is built from its definition in Word Board, typed from audio in Listen & Spell, picked inside a paragraph in Fill in the Blank, and used inside a conversation in Scenarios. The smart review queue brings the word back; the change of mode changes the context. For the timing and number of repetitions, go back to The Science of Repetition.
References
- Godden, D. R. & Baddeley, A. D. (1975). Context-dependent memory in two natural environments: On land and underwater. British Journal of Psychology, 66.
- Smith, S. M., Glenberg, A. & Bjork, R. A. (1978). Environmental context and human memory. Memory & Cognition, 6.
- Kerr, R. & Booth, B. (1978). Specific and varied practice of motor skill. Perceptual and Motor Skills, 46. Shea, J. B. & Morgan, R. L. (1979). Contextual interference effects on the acquisition, retention, and transfer of a motor skill. Journal of Experimental Psychology: Human Learning and Memory, 5.
- Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In Metacognition: Knowing about knowing. MIT Press.
- Kornell, N. & Bjork, R. A. (2008). Learning concepts and categories: Is spacing the "enemy of induction"? Psychological Science, 19. Rohrer, D. & Taylor, K. (2007). The shuffling of mathematics problems improves learning. Instructional Science, 35.
- Webb, S. (2008). The effects of context on incidental vocabulary learning. Reading in a Foreign Language, 20. Bolger, D. J., Balass, M., Landen, E. & Perfetti, C. A. (2008). Context variation and definitions in learning the meanings of words. Discourse Processes, 45. Nakata, T. & Suzuki, Y. (2019). Mixing grammatical and semantic categories in vocabulary learning. Studies in Second Language Acquisition, 41.