Stimulus-specific memorability effects go beyond memory
Undergraduate Just-In-Time Abstract
Poster Presentation 56.346: Tuesday, May 19, 2026, 2:45 – 6:45 pm, Banyan Breezeway
Session: Undergraduate Just-In-Time 3
Schedule of Events | Search Abstracts | Symposia | Talk Sessions | Poster Sessions
Darius Suplica1, Igor Utochkin1, Edward Awh1; 1University of Chicago
Stimulus-specific memorability refers to the finding that recognition memory performance is enhanced for specific items, independent of study and retrieval context. Nevertheless, recent work has raised the possibility that memorability effects are not specific to memory, per se. For instance, the efficiency of low-level perceptual discriminations – in the absence of memory demands – is higher for memorable stimuli (Deng et al., 2024). Following this thread, we investigated whether memorability affects visual search performance when recognition memory is not a limiting factor. Critically, memorability appears to have two distinct facets: “Hittability” and “false-alarmability” refer to the probability of correctly recognizing previously studied items, or falsely recognizing unstudied items, respectively. While both facets can strongly impact memory performance, they are virtually uncorrelated at the item level. Thus, we independently manipulated the hittability and fa-ability of targets and distractors in four separate samples. Visual search performance was strongly affected by both factors. Subjects were faster to find more hittable (rm-ANOVA, p = 1 x 10-11) and less fa-able targets (p = 6.9 x 10-10), with no effect of distractor memorability (hittability p=0.18, fa-ability p=0.08). Experiment 2 replicated these findings while employing a continuous manipulation of hittability and fa-ability within a single set of stimuli. Once again, subjects were faster to recognize more hittable and less fa-able targets (linear mixed-effects model, hittability p=7x10-17, fa-ability p=1.2x10-6), with a significant (p=2x10-19) negative interaction between the two. In line with the argument that hittability and fa-ability are distinct aspects of memorability, search performance was better explained by keeping hittability and fa-ability as independent factors, rather than combining them into single d-prime measure (WAIC=-2954 vs -3048). Therefore, stimulus-specific memorability powerfully affects performance in tasks that do not require recognition memory, motivating further work on precisely which computations are impacted by memorability.
Acknowledgements: The authors were supported by NSF grant #2402391 to E.A.