Does practice in category learning increase rule use or exemplar use—or both?

Abstract

Categorization research has demonstrated the use of both rules and remembered exemplars in classification, although there is disagreement over whether learners shift from one to the other or use both strategies simultaneously. Theoretical arguments can motivate predictions for both rule use and exemplar use increasing with more practice. We describe a single large experiment (n = 190) that manipulated the number of training items (category size), the number of presentations of each training item, and the similarity between the training and the transfer stimuli in order to discover when rules and exemplars are most likely to be used. Results showed that rules and exemplars both influenced classification and that exemplars were used more often with smaller categories, with more training on items, and when test items were similar to training items. There was no consistent evidence of a shift from rule-based to exemplar-based categorization with more learning. Importantly, we found a number of conditions in which rules and exemplars were both used, even within individual participants. We discuss our results in terms of hybrid models of classification.

Tracking prototype and exemplar representations in the brain across learning

“…Thus, exemplar representations may also form to support judgments requiring specificity. Precedence for co-existing representations also comes from neuroimaging studies of spatial processing (Brunec et al, 2018), episodic inference (Schlichting et al, 2015), and memory for narratives (Collin et al, 2015 (Desimone, 1996;Gonsalves, Kahn, Curran, Norman, & Wagner, 2005;Henson, Shallice, Gorno-Tempini, & Dolan, 2002) Prior work has shown that individuals may use both rules and memory for individual exemplars throughout learning without strong shifts from one to the other (Thibaut, Gelaes, & Murphy, 2018). Others have suggested that there may be representational shifts during categorylearning, but rather than shifting between exemplar and prototype representations, early learning may be focused on detecting simple rules and testing multiple hypotheses (Johansen & Palmeri, 2002;Nosofsky et al, 1994;Paniukov & Davis, 2018), whereas similarity-based representations such as prototype and exemplar representations may develop later in learning (Johansen & Palmeri, 2002).…”

The Social Route to Abstraction: Interaction and Diversity Enhance Performance and Transfer in a Rule‐Based Categorization Task

“…It is thus possible to learn the alien categories by an exemplar strategy either by recalling previous exposures and feedback to the same exemplar or relying on similarities to previously seen exemplars. Indeed, there are observations suggesting that participants often rely—at least partially—on such exemplar strategies, even when categories are defined entirely by rules (Chubala et al., 2016; Hu & Nosofsky, 2021; Rodrigues & Murre, 2007; Thibaut et al., 2018). In order to get an indication of the potential efficacy of such strategies in the context of the alien categorization task, simulations based on the GCM provide a useful baseline.…”

“…Second, as an additional way to query participants' emerging rule inductions, we performed an analysis only on the first block of stimuli of each session, that is, before any stimulus exemplars have been repeated. Above-chance performance on these initial exposures could indicate that participants are inducing rules and generalizing from other exemplars (but see also Chubala et al, 2016;Hu & Nosofsky, 2021;Thibaut et al, 2018). Together with details about motivation and specification of priors and model checks, these additional analyses are provided in the Supplementary Material (link is provided by the end of the section).…”

“…In the low‐level complexity condition, the nutritiousness of an alien was constituted by a single feature (arms up), while a dangerous and nutritious alien would have arms up, eyes on stalks, and spots. This meant that category members shared particular features and thus an element of basic, perceptual similarity and the task could be solved relying also on exemplar‐based learning (Chubala, Johns, Jamieson, & Mewhort, 2016; Hu & Nosofsky, 2021; Thibaut, Gelaes, & Murphy, 2018). In contrast, in the highest complexity condition, a category would be defined by relations among multiple features such that nutritiousness, for instance, would be marked by any combination of two out of four features (e.g., either eye on stalks and fat legs, spots and blue color, or eye on stalks and blue color).…”

Representando Las Conectivas Lógicas en Una Tarea De Igualación a La Muestra

“…Algunos conceptos pueden ser delimitados de manera muy precisa mediante una regla o criterio que separa los ejemplares positivos de los negativos (e.g., Bourne, 1970;Bruner et al, 1956;Feldman, 2003;Thibaut et al, 2018;Vernouillet et al, 2021). En el presente trabajo son de particular interés aquellos conceptos en los que la regla definitoria especifica dos o más atributos de estímulo y un tipo de relación entre dichos atributos (Bourne, 1970).…”