Electronic Thesis and Dissertation Repository


Doctor of Philosophy




Dr. John Paul Minda


Executive functions are important for learning rule-based (RB) categories, as well as non-rule-based (NRB) categories (e.g., categories learned implicitly, without a verbal rule). However, executive functioning is known to decline with age, leading to age-related deficits in category learning. The current thesis examines RB and NRB category learning in older adults using category sets that vary in difficulty (e.g., rule complexity, number of stimulus dimensions, salience of stimulus dimensions). In Chapter 2, older adults and younger adults completed three category sets (simple single-dimensional RB, disjunctive RB, and NRB). Older adults learned the simple, single-dimensional rules quite well. In contrast to younger adults, older adults found complex disjunctive RB categories harder to learn than NRB categories because of the executive functioning demands associated with complex rule learning. In Chapter 3, I introduced a pre-training procedure prior to the disjunctive RB and NRB categorization task used in Chapter 2. This was done in an effort to reduce task demands, as to minimize age-related categorization deficits. Both RB and NRB category learning improved among older adults following pre-training, but the improvements to RB learning were more drastic, suggesting that executive functioning plays a heavier role in RB learning. In Study 1 of Chapter 4 I used a difficult, single-dimensional RB category set (i.e., the correct rule is based on the less salient stimulus dimension) and a NRB category set to further examine category learning in normal aging and to better understand the types of strategies used by older adults. Relative to younger adults, older adults struggled with learning both the RB and NRB category set because they used suboptimal rules during the RB task and a RB strategy during the NRB task. In Study 2 of Chapter 4, I used a pre-training procedure to familiarize older adults with the stimulus dimensions of the RB category set, reducing the executive function demands of the task. Pre-training improved RB accuracy and the consistency with which older adults applied the rule. Across all studies, executive functioning abilities were associated with RB and NRB category learning. Overall, the results from this thesis help to better understand the locus of age-related categorization deficits and offer a method of reducing these deficits.