Although significant progress has been made with respect to our understanding of categorization and concept learning behaviour in adults, much less is known about how this key capacity plays out with respect to more restrictive populations. This is unfortunate because much may be revealed about the nature of concept learning by examining the limits exhibited by special populations. With this tenet in mind, in what follows, we investigated key aspects of concept learning in terms of the unexplored comparative performance of three populations-adults, adolescents, and adolescents with attention-deficit/hyperactivity disorder (ADHD). To do so, we employed a novel parainformative experimental task involving categorical stimuli with four objects defined over three dimensions. The learning difficulty ordering for these types of three-dimensional stimuli has proven robust and has been replicated by several researchers. Indeed, in our experiment, we observed the same concept learning ordering in adults and adolescents, but not in ADHD adolescents: For example, the latter group showed greatly impaired categorization performance on stimuli characterized by an "exclusive-or" rule. However, categorization performance on such type of stimuli indicated good reliability in discriminating between adolescents with and without ADHD (receiver-operating characteristic, ROC = .82). We accurately predicted and accounted for these results using generalized invariance structure theory (GIST; Vigo, 2013. The GIST of concepts. Cognition, 129, 138-162), which posits that organisms detect invariance patterns in stimuli that are necessary precursors to concept formation.
Keywords: Attention-deficit/hyperactivity disorder; Categorization; Cognitive modelling; Concept learning; Generalized invariance structure theory; Individual differences learning.