Individual differences in stimulus identification, rule induction, and generalization of learning.

Abstract

In the field of stimulus generalization, an old yet unresolved discussion pertains to what extent stimulus misidentifications contribute to the pattern of conditioned responding. In this article, we perform cluster analysis on six datasets (four published datasets and two unpublished datasets, included N = 950) to examine the relationship between interindividual differences in (a) stimulus identification, (b) patterns of generalized responding, and (c) verbalized generalization rules. The datasets were obtained from online predictive learning tasks where participants learned associations between colored cues and the presence or absence of a hypothetical outcome. In these datasets, stimulus identification and expectancy ratings were assessed in separate phases to a range of colors varying between blue-green. Using cluster analyses on performance during stimulus identification, we identified different subgroups of participants (good vs. bad identifiers). In all six datasets, we found a close relationship between the pattern of stimulus identification and the shape of the expectancy gradient across the test dimension between the identified subgroups. Furthermore, participants classified as good identifiers were more likely to report a similarity generalization rule than a relational or linear rule, suggesting that individual differences in stimulus identification are related to individual differences in generalization rules. These findings suggest that greater consideration should be given to interindividual variability in stimulus identification, inductive rules, and their relationship in explaining patterns of generalized responses.

Multiple pathways to widespread fears: Disentangling idiosyncratic fear generalization mechanisms using computational modeling

“…The inverse pattern may be observed because of sampling bias where there are far fewer individuals in the first dataset classified to these two groups than in the second (40% versus 72.5%). Regardless of which scale is generally more relevant to the generalization behavior, our results show that the perceptual variability (Figure 1) is one of the contributors to the generalization behavior for some individuals and hence should not be neglected [21][22][23][24][25] .…”

“…Recent research on generalization [21][22][23][24][25] has demonstrated the existence of perceptual variation between and within individuals. Hence, perceptual variation may affect inter-stimulus similarity if perception, rather than physical characteristics, determines the coordinates of stimulus representations in a multidimensional psychological space.…”

“…Without proper learning, organisms are incapable of generating distinct and consistent responses. However, even when CS-US learning is firmly established, generalized responses can still occur as a result of imperfect perceptual discrimination between the CS and newly encountered stimuli [21][22][23][24][25] , providing the second potential mechanism for generating flatter or noisy response gradients. In earlier research, the relationship between insufficient perceptual discrimination and generalization behavior has been discussed 26,27 .…”

“…Conditioned responses to test stimuli were strong when they were falsely perceived as the CS while attenuated when the stimulus was perceived as different from the CS (even during CS trials). This inflated the inferred extent of fear generalization as response gradients were much narrower after accounting for discrimination errors [21][22][23][24][25] . In sum, rather than assuming a single mechanism by which humans generalize their responses, a more plausible account is that generalization occurs as a result of a variety of cognitive and perceptual processes 28 .…”

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

“…While this work has demonstrated the profound impact of response rules, it failed to identify what caused participants to develop different types of response rules despite similar learning experiences. A recent study found latent group differences in the ability to correctly identify the CS+ (Zaman, Yu, & Lee, 2022). Through cluster analysis, participants allocated to a group with good CS identification performance had a much higher probability of reporting a similarity-based response rule.…”