Investigating Learning of Similarity-Consistent vs. Similarity-Inconsistent Categories using Feedback vs. Observational Methodologies

Category learning is crucial for organizing information to be used throughout our daily lives. One way category learning is assessed in a laboratory setting is by testing participants’ ability to sort faces into family groups. Previous work in our lab (Ashby et al., 2023) assessed how physical similarity of faces impacts category learning by comparing learning when category members share physical features (similarity-consistent categories) and they do not share physical features (similarity-inconsistent categories). Similar-consistent condition yielded a higher learning rate than the similar-inconsistent condition, which showed minimal learning despite requiring only 9 face-name associations. Given that minimal learning was demonstrated from the similar-inconsistent condition, we aimed to alter the learning process to determine if differences in training could yield more substantial results. Findings from other studies suggest variation in participants’ learning rates depending on the task (Henricksson & Enkvist, 2018; Levering & Kurtz, 2014), thus this study will compare observation learning against feedback learning to determine if participants’ ability to learn similarity-inconsistent categories differ based on differences in methodology. The two forms of methodology we will use are observation learning and feedback learning. The participants assigned to observation learning will be shown black-and-white faces and their corresponding categories, whereas the group assigned to feedback learning will be tasked with guessing which family each face belongs to and will receive feedback on their responses. All groups will complete no-feedback categorization test after training. Categorization will be compared using a 2x2 between-subjects ANOVA test, with experimental condition (similarity-consistent, similarity-inconsistent) as a between-subject factor and learning condition (observation learning, feedback learning) as the second between-subject factor.