Rule-based generalization and peak shift in the presence of simple relational rules

Abstract

After discrimination learning between two stimuli that lie on a continuum, animals typically exhibit generalization on the basis of similarity to the physical features of the stimuli, often producing a peak-shifted gradient. However, post-discrimination generalization in humans usually resembles a monotonically increasing (e.g., linear) gradient that is better characterized as following a relational rule describing the difference between the stimuli. The current study tested whether rule-based generalization could be disrupted by reducing the applicability of a relational rule on test. We compared generalization following a difficult categorization task between a group who could use their rule consistently throughout test (Group Consistent), and a group who could only apply their rule effectively on 50% of test trials and thus could only use it inconsistently (Group Inconsistent). Across two experiments, a peak shift was found in the Inconsistent group and a monotonic gradient in the Consistent group. A post-hoc sequential analysis revealed that the Inconsistent group produced both peak-shifted and monotonic gradients as a function of whether or not the relevant rule was applicable on the previous trial. Reducing the applicability of a rule on test thus appeared to lead participants to revert to generalizing on the basis of similarity. Our results suggest that humans learn about the physical features of the stimuli alongside relational rules, and that rule- and similarity-based learning can interact in determining generalization.

Modelling generalisation gradients as augmented Gaussian functions

“…This monotonic pattern of generalisation can take either a sigmoidal or more linear form and has been found in multiple human studies following the same discrimination procedures that usually produce peak shift in animals (J. C. Lee et al, 2018; J. C. Lee & Livesey, 2018; Livesey & McLaren, 2009; S.…”

“…The first test compares responses at the location of the gradient peak against responses for the CS+, and the other compares the peak to the end point of the dimension (e.g., J. C. Lee et al, 2018; J. C. Lee & Livesey, 2018; Livesey & McLaren, 2009).…”

Easy-to-hard effects in perceptual learning depend upon the degree to which initial trials are “easy”

“…The recent work of Lee and Livesey (2018) combined with the results reported here suggest an interesting possibility for why attentional spotlighting has remained such a favored account of easy-to-hard effects: Strong attentional spotlighting effects have the potential to mask effects of other learning processes. Easy-to-hard effects unlikely to be related to attentional spotlighting are generally smaller (e.g., Cohen's d = 0.90 for Experiment 1; Liu et al, 2008) compared with the effect of knowing the relevant dimension(s) (e.g., Cohen's d = 2.58 for Experiment 5; Pashler & Mozer, 2013).…”

“…Others have come to similar conclusions based on recent work in category learning. Lee and Livesey (2018) trained participants to categorize circles that varied in color and size such that paying attention to a single dimension was encouraged. Afterwards, participants were tested either in a condition where they could consistently apply a rule based on the relevant dimension (e.g., small circles are Category A, but large circles are Category B), or in a condition where such a rule led to inconsistent performance.…”

Evidential diversity increases generalisation in predictive learning

“…For example, following discrimination training where two perceptually similar stimuli predict different outcomes or belong to different categories, participants generalise in a manner that suggests that they have learned about relations between stimuli (e.g., “greener than”) rather than just physical characteristics (e.g., Lee, Hayes, & Lovibond, 2018; Lee & Livesey, 2018; Livesey & McLaren, 2009). Furthermore, these patterns of dimensional generalisation (highest responding at one extreme of the dimension) often correspond to explicit (i.e., verbalisable) rules describing the basis of generalisation (e.g., “the greener the stimulus, the higher the likelihood of the outcome”; see Lee et al, 2018; Wong & Lovibond, 2017).…”