Why more attempts in one sitting don't buy much durability
It is tempting to think of memory like a muscle: each retrieval is a repetition, so more repetitions should mean more strength. The evidence points somewhere less intuitive. When researchers compared expanding retrieval schedules (tests spaced closer together at first, then further apart) with equally spaced ones, the expanding schedule produced better short-term recall but the equally spaced schedule won out over the long term [1]. The repeated attempts were present in both conditions; what differed was their timing.
That result is the clearest signal that repetition count is not the engine. If simply racking up more retrieval attempts drove long-term retention, the schedule with more closely packed early attempts should have held its advantage. Instead, the advantage flipped as the retention interval lengthened, which suggests the benefit comes from the conditions surrounding each attempt rather than the tally of attempts [1].
The first retrieval attempt may do the heavy lifting
One explanation offered for these results is that the difficulty of the very first retrieval attempt is critical [1]. A test you can answer easily — because you just read the material — is a weak signal to the brain that this information is worth keeping. A test you have to struggle through, after some forgetting has set in, is a stronger one. On this view, delaying the initial test is what makes later retention better, and the abstract explicitly frames this as a possible reason the expanding schedule underperformed long term [1].
This is a proposed mechanism, not a settled one. The authors draw an analogy to massed repetition, which is known to produce poor retention, and suggest the same logic may apply to retrieval attempts that come too easily and too close together [1]. The practical upshot is that the value of a retrieval attempt depends on how much effort it demands, not just on the fact that it happened.
What longer training studies add
Studies that track learners over weeks rather than a single session tell a compatible story. In a four-week training study comparing expanding and equal-interval retrieval schedules, the timing of the first retrieval attempt mattered more than the relative spacing of the subsequent attempts in determining long-term retention [3]. That is a narrower claim than "repetition doesn't matter," but it sharpens the picture: once you have decided to retrieve repeatedly, the early attempt is where the leverage is.
Work on mathematics learning reaches a similar conclusion from a different direction. When researchers varied both the amount and the spacing of retrieval practice, increased spacing led to better retention across a semester [4]. The study was designed to examine repetition and spacing together, so it does not cleanly separate the two, but its spacing result lines up with the broader pattern: distributing attempts outperforms piling them up [4].
So does repeating retrieval matter at all?
Yes — but the strongest evidence for repetition comes from comparisons where spacing is held roughly constant or not the focus. One line of work reports that continuing to retrieve items during learning improved long-term retention by more than 100% compared with dropping those items from further testing [2]. That is a large effect, and it shows that repeated retrieval is far better than no further retrieval. The caveat is that this comparison does not isolate repetition from spacing, so it cannot tell us how much of the gain is due to the extra attempts versus their timing [2].
A neural study offers a possible glimpse of the mechanism. When participants underwent repeated tests, lesser similarity in brain activity patterns across those tests was associated with better long-term retention [5]. In plain terms, the brain responded a little differently each time rather than replaying an identical pattern, and that variability tracked with stronger memory later. Because the repeated testing happened within a single day, this design also does not separate repetition from spacing [5], so the finding is best read as a clue about what repeated retrieval does rather than proof of how much it contributes.
Put together, the evidence supports a practical rule: repeated retrieval is clearly better than none, and the difficulty and timing of the attempts shape how much durable benefit you get. If you can only change one thing, change when the attempts happen — especially the first one — rather than adding more of them back to back [1][3].
Sources used in this answer
Expanding retrieval practice promotes short-term retention, but equally spaced retrieval enhances long-term retention.
Expanding retrieval produced better short-term retention, but equally spaced retrieval produced better long-term retention, with the difficulty of the initial retrieval attempt proposed as critical.
Repeated retrieval during learning is the key to long-term retention
Repeated retrieval during learning improved long-term retention by more than 100% relative to dropping items from further testing, though the design did not isolate repetition from spacing.
Retrieval practice over the long term: Should spacing be expanding or equal-interval?
In a four-week training study, the timing of the first retrieval attempt mattered more than the relative spacing of subsequent attempts for long-term retention.
How the amount and spacing of retrieval practice affect the short-and long-term retention of mathematics knowledge
Increased spacing of retrieval practice led to better retention of mathematics knowledge across a semester, in a design that examined amount and spacing together without isolating repetition.
Lesser neural pattern similarity across repeated tests is associated with better long-term memory retention
Lesser similarity in brain activity patterns across repeated tests was associated with better long-term retention, in a design with repeated testing within a single day.