Category learning processes in the light of variability: Insights from a self-regulated category learning task

Category learning helps us to make sense of the complex world around us. In contrast to conventional laboratory experiments, individuals often possess the opportunity to self-regulate their category learning process and determine when they acquired sufficient knowledge about the categories. We investigate how category variability, i.e., how much the exemplars vary from one another, influences search length in a novel category learning task. In this task, participants sampled category exemplars from two categories until they felt they could categorize the objects. We hypothesized that a higher variability within the focal and the counter-category increases the search length, here the number of sampled exemplars. In three experiments, we demonstrate that as the variability within the categories increases, participants sample more extensively. This finding suggests that individuals weigh the cost of sampling against environmental demands, adapting their information search to cope with the uncertainty arising from variability. Moreover, our findings reveal that the variability of the counter-category significantly influences the search and learning process. In summary, self-regulated category learning enables participants to adapt their learning behavior to environmental factors like category variability. Importantly, our study underscores that the acquisition of a category is intricately linked with the experience of a counter-category. In a nutshell, by exploring the dynamics of category variability in a novel task format, this study contributes insights into how learners adapt to the categories’ characteristics and regulate their learning behavior with respect to the context.