Proactive interference accumulates with the number of previously learned similar items, not a single one-off effect

What it is

Proactive interference (PI) is when material learned earlier gets in the way of learning or retrieving newer material — the direction runs "old interferes with new." The point to stress here is that it isn't a one-time hit: interference accumulates with the number of previously learned similar items — the more similar prior content someone has learned, the harder it becomes to cleanly remember the new content, rather than a fixed-size penalty from having learned old content just once.

Why it happens

If new and old material share the same retrieval cue (the same entry point, the same menu location), every previously learned similar item becomes a competing candidate under that cue. The more candidates pile up, the more crowded that single cue becomes, and the more competitors the new item has to beat to be picked out cleanly. Interference isn't caused by any one old memory in isolation — it's the crowding effect produced by everything accumulated under that cue over time. This is why the longer someone has used the same feature entry point, and the more versions they've lived through, the more likely a new operating method is to get pulled off course by old habits.

Studying it

The classic paradigm is multi-round paired-associate learning: participants successively learn several structurally identical sets of pairs (repeatedly assigning the same cue words to different new answers), testing recall of the most recent set after each round. Recall of the newest set systematically declines as the number of previously learned sets increases. The independent variable is the number of prior similar sets learned; the dependent variable is recall/accuracy on the newest set. A real-world analog is a workflow whose rules get repeatedly changed but keep the same structure (adjusting the value rule for the same field over and over) — you can observe whether user memory for "the current rule" also degrades as the number of prior versions experienced grows.

Where it stops holding

For proactive interference to actually occur, the new and old material need to substantially overlap under the same cue or context — if the newly learned content is clearly different in structure or category from what came before, interference sharply weakens or disappears, though the mechanism behind that release is developed elsewhere. The accumulation also isn't unbounded — it typically plateaus after a certain number of similar prior items rather than continuing to worsen indefinitely with count. Most evidence also comes from paired-associate experiments completed within minutes to tens of minutes; applying the accumulation slope measured in these short-timescale experiments directly to real scenarios spanning years and multiple product iterations is an extrapolation, not direct evidence.

Applying it

• Audit controls whose position, entry point, or shortcut meaning has been redefined multiple times over the product's history — if the same button location has meant three different things in succession, expect long-tenured users to show a markedly higher error rate at that spot than users who've only ever known one meaning, rather than assessing risk purely by "is this a new feature."
• For entry points with this kind of high historical reuse, prioritize clearer current-state cueing (a temporary highlight explaining "here's what this does now") rather than assuming users will naturally retain the latest meaning.
• Validation: group users by tenure or number of past versions experienced and compare error or help-request rates at the same change point — if error rate rises systematically with the number of versions experienced, proactive interference's cumulative effect is at work and should be addressed directly, rather than attributing the problem to "the new feature being hard to use."

Related

• Same group: A6.16.2 Switching to a new category clearly different from the old content can partly release the buildup of proactive interference · A6.16.3 Retroactive interference grows with the similarity between new and old content, and shrinks as the difference grows · A6.16.4 Successive version iterations stack proactive interference, leaving memory for the newest version the most fragile
• Nearby: A6.14 Retrieval from long-term memory · A5.16 Habituation and cue wear-out
• Search terms: proactive interference · interference theory · cue competition