Look a little closer
After hearing a list and being asked to recall its items in any order, people generally retrieve more from the beginning and end than from the middle. This U-shaped relationship between list position and recall is the serial-position effect. The early advantage is called primacy and the late advantage recency.
Early items benefit partly from having room to be processed. When the first word appears, few competitors are present, so it can be silently rehearsed or connected with meaning several times. As more words arrive, rehearsal must be divided among them. Changing presentation speed or preventing rehearsal can therefore alter the size of the primacy advantage.
The final items are temporally close to the moment recall begins. Their recent context overlaps strongly with the present retrieval context, and some remain especially accessible, so they often come to mind first. A classic account assigned this advantage to short-term memory, while modern models show that changing temporal context and retrieval dynamics can also produce recency over longer intervals.
A distractor before recall helps reveal that primacy and recency do not respond identically. If participants count backward for a short period instead of answering immediately, the end-of-list advantage can shrink dramatically while much of the early advantage remains. Recent accessibility and the stronger encoding of initial items are sensitive to different conditions.
The finding is not a law that people always remember only the first and last things they encounter. List length, presentation rate, familiarity, meaning, attention, delay, intervening activity and the type of memory test can all reshape the curve. A distinctive or emotional event in the middle can also overcome the ordinary disadvantage of its position.
The kind of memory test matters as well. The serial-position curve is prominent in free recall, where people produce items without cues. Recognition among alternatives and reconstruction of the original order demand different retrieval processes. Applying a list-memory result directly to interview order, advertising placement or another complex real situation therefore requires additional evidence.
The serial-position effect shows that memory is not a recorder assigning every item equal strength. The beginning offers extra opportunities for rehearsal; the ending benefits from closeness to the retrieval moment. Middle items often receive less of either advantage. Alongside what we encounter, memory preserves something about where in time we encountered it.
