The idiosyncratic nature of how individuals perceive, represent, and remember their surroundings and its impact on learning-based generalization.

Abstract

The current study adopted a multimodal assessment approach to map the idiosyncratic nature of how individuals perceive, represent, and remember their surroundings and to investigate its impact on learning-based generalization. During an online differential conditioning paradigm, participants (n = 105) learned the pairing between a blue color patch (CS+) and an outcome (i.e., shock symbol) and the unpairing between a green color patch and the same outcome. After the learning task, the generalization of outcome expectancies was assessed to 14 stimuli spanning the entire blue-green color spectrum. Hereafter, a stimulus identification task assessed the ability to correctly identify the CS+ among this stimulus range. Continuous and binary color category membership judgments of the stimuli were assessed preconditioning. We found that a response model with color perception and identification performance as sole predictors was preferred to contemporary approaches that use stimulus as a predictor. Interestingly, incorporating interindividual differences in color perception, CS identification, and color categories significantly improved the models’ ability to account for different generalization patterns. Our findings suggest that insight into the idiosyncratic nature of how individuals perceive, represent, and remember their surroundings provides exciting opportunities to understand post-learning behaviors better.

The assessment of gender differences in perceptual fear generalization and related processes

“…However, no gender differences were observed in tasks related to a contrast detection threshold, visual search, orientation discrimination, the Simon effect, and four other visual illusions (Shaqiri et al, 2018). In this study, we consistently measured realtime estimations of geometric size using a rating scale-a methodology demonstrated in recent perceptual generalization research (Yu et al, 2023;Zaman et al, 2022Zaman et al, , 2023 for effectively tracking the dynamic shifts in geometric perception during fear learning and generalization processes. Our findings revealed gender differences in both perceptual intercept and slope parameters through a statistical linear model, indicating a pattern where women exhibited greater perceptiveness than men.…”

“…Recent advances in research on fear generalization suggest that a multitude of mechanisms can contribute to generalization behavior; an exclusive focus on fear responses renders researchers unable to scrutinize these mechanisms (Struyf et al, 2015;Yu et al, 2023;Zaman, Chalkia, et al, 2020;Zaman et al, 2023). Congruently assessing (variations in) stimulus perception, fear and safety learning, and generalized fear responses enables to account for inter-and intra-individual differences regarding the latter (Struyf et al, 2017;Zaman et al, 2022;Zaman, Struyf, et al, 2019.…”

“…Congruently assessing (variations in) stimulus perception, fear and safety learning, and generalized fear responses enables to account for inter-and intra-individual differences regarding the latter (Struyf et al, 2017;Zaman et al, 2022;Zaman, Struyf, et al, 2019. It also yields generalization indexes that are comparable across individuals as they control for differences regarding learning, perception, and memory (Yu et al, 2023;Zaman et al, 2023). In a recent paper, Yu et al (2023) investigated latent mechanisms in fear generalization behavior via a computational generalization model that parameterizes several key processes that contribute to generalized responding, including the speed of learning (Rescorla & Wagner, 1972) and the strength of a similarity-based generalization process (Shepard, 1987) while accounting for differences in stimulus perception.…”

Successful generalization of conceptual knowledge after training to remember specific events

“…After optimization, prototype and exemplar model fits were compared at the individual level. A number of studies have demonstrated how group-level trends in generalization can be obtained by superimposing subgroup trends (F. G. Ashby et al, 1994;Lee & Livesey, 2018;Lovibond et al, 2019), highlighting the need for individualized assessment of generalization behavior (Zaman et al, 2023). Model fits were compared to each other and to chance to estimate the best fitting categorization strategy for each individual participant.…”

The role of perception in generalization: Commentary on Zaman, Yu, & Verheyen (2023)

“…Variations in the gradient of learned responses are usually interpreted as differences in the underlying cognitive process of generalization. A recent study by Zaman, Yu, and Verheyen (2023) seeks to challenge this view, arguing that generalization is best modelled by perceptual factors and that individual differences in perception are a primary driver of generalization. In this commentary, we outline issues in the methodology and analysis of Zaman et al, and show that their key result is not robust to the addition of theoretically-informed alternative models.…”