The serial position effect - why you remember the first and last things in a session

Sit down and study a long list of anything - vocabulary, dates, concepts in a lecture - and you'll remember the beginning and end of that session far better than the middle. This isn't a personal focus problem; it's one of the more consistently replicated findings in memory research, and it has a real, practical fix.

Primacy and recency, the two halves of the effect

Items early in a study sequence get an advantage (primacy) because they've had more total rehearsal time by the end of the session - you've encountered them, then kept encountering new material while the early items sit in the background getting reinforced. Items at the very end get their own advantage (recency) simply because they're still fresh in short-term memory when the session ends. Everything in between gets neither benefit, and it shows: recall for mid-session material is reliably worse in controlled studies, not just in how it feels.

Why this matters for how you structure a session

If you always study the same material in the same order - front of the deck to back, every time - the middle of that fixed sequence is quietly under-learned relative to the ends, session after session. The material doesn't rotate through "getting the primacy or recency boost," it's stuck in the disadvantaged middle every single time you review it in that order.

The actual fix: change the order, or break up the session

Shuffling the order material appears in, or splitting one long session into a few shorter ones, both work by giving different material a turn at being "first" or "last." A feed-based review that pulls cards without a fixed sequence sidesteps the problem entirely - there's no consistent middle to lose, because there's no fixed order for a middle to exist in. See interleaving vs. blocking for a related but distinct ordering effect - interleaving is about mixing topics, this is about position within a single pass.

Where this compounds with spaced repetition

A scheduler that resurfaces cards on individually-tracked intervals, rather than reviewing a whole fixed list start to finish each time, naturally avoids a stable "middle" - each card's position relative to the others shifts session to session as they move through their own interval ladder. That's a secondary benefit worth knowing about on top of what spaced repetition actually does for the misses it's more directly built to catch.