Tide table · when to expect high water
Recency Effect
The tendency to remember the most recently presented items better than earlier ones, especially in immediate recall tasks.
The lunar pull · the condition that drives the surge
Why the water moves
The classic account, rooted in the Atkinson-Shiffrin multi-store model, holds that recency arises because terminal items are still resident in a limited-capacity short-term store when recall is requested; they can be output directly without retrieval from long-term memory. A brief post-list distraction task displaces these items from the short-term buffer, eliminating recency while sparing primacy (which reflects long-term rehearsal). Alternative single-system accounts (e.g., Wixted & Ebbesen, 1991) argue that recency reflects relative retroactive interference: the last items have no subsequent items to interfere with them.
The dual-store interpretation (Atkinson & Shiffrin, 1968) competes with temporal distinctiveness and ratio-rule accounts (e.g., Wixted & Ebbesen, 1991; Brown, Neath, & Chater, 2007) that explain recency within a single memory system.
Prediction reliability — replication
The serial position curve, including the recency effect, is one of the most robust phenomena in cognitive psychology, replicated thousands of times across variations in materials, participants, presentation modalities, and list lengths. Meta-analytic support is implicit in the sheer volume of converging studies reviewed in Greene (1986) and more recently in Unsworth & Engle (2007). The recency effect also appears in recall of U.S. presidents, advertising messages, and clinical assessments of amnesia.
Recorded tides
Serial position curve. Recall probability was highest for the first 3–4 items (primacy) and the last 8 items (recency), with a flat middle section. The recency effect persisted even in lists of 40 words.
Murdock, 1962 · window: n = 103 undergraduates
Distraction eliminates recency. Immediate recall showed both primacy and recency effects. The 30-second delay eliminated the recency effect but left primacy intact, suggesting recency depends on short-term memory.
Glanzer & Cunitz, 1966 · window: n = 46 (Experiment II)
Recency in juror judgments. Incriminating evidence presented by the last witness produced a 69.8% conviction rate versus 48.4% when presented first—a robust recency effect. Memory for the critical evidence partially mediated the effect.
Costabile & Klein, 2005 · window: n = 184 (Experiment 1)
King tides · logged extremes
Juror verdicts swayed by late-presented evidence · 2005
Costabile and Klein (2005) demonstrated in mock-juror experiments that a wiretapped confession presented by the final prosecution witness produced a 69.8% conviction rate, versus 48.4% when presented first. The recency effect persisted even when the evidence was ruled inadmissible.
Tide warning · the safe window
Introduce a brief delay or distraction task before making a decision that was informed by recently presented information, or deliberately review all items before rendering judgment.
Glanzer and Cunitz (1966) showed that a 30-second backward-counting task eliminated the recency effect, suggesting that a brief delay can level the playing field between recent and earlier information.
Reading the gauge in the wild
Watch for situations where the last item heard, seen, or read disproportionately influences judgment—such as the final candidate in an interview lineup, the last witness in a trial, the most recent stock tip, or the last data point in a presentation.
Neighbouring waters
- Primacy EffectBoth primacy and recency are components of the serial position curve, but they arise from different mechanisms: primacy from enhanced rehearsal and long-term encoding, recency from short-term memory availability.
- Availability HeuristicThe recency effect contributes to the availability heuristic by making recently encountered information more accessible in memory.