The goal of the experiments was to evaluate an explanation of set-alternation costs based on episodic memory principles. The assumption is that performance of any task is a consequence of memory retrieval processes that involve representations of specific prior experiences (Kolers, 1976; Leboe, Whittlesea, & Milliken, 2005; Neill & Mathis, 1998; Tenpenny, 1995; Whittlesea, 1997; Whittlesea & Jacoby, 1990). When the Event 1 and 3 targets mismatch the retrieval of the Event 1 memory episode is not entirely appropriate for performing the Event 3 task. The interference due to a partial match between Events 1 and 3 might be the source of set-alternation costs. Results of Experiment 1 revealed larger costs in the high probability set-alternation condition. The high probability set-alternation condition encouraged retrieval of Event 1. However, because the targets of Event 1 and 3 mismatched the retrieval of Event 1 interfered with the processing of Event 3’s task-set. In other words, the interference due to a match in task-sets but a mismatch in targets generated costs. If set-alternations costs originate from a partial match between Events 1 and 3, increasing the amount of overlapping information between these events should reduce costs. The findings of Experiments 2 and 3 showed reduced set-alternation costs when there was a target identity match between Events 1 and 3. Lastly, Experiment 4 showed that set-alternation costs are larger when the retrieval of the Event 1 memory episode is obstructed. That is, costs were larger when there was a combination of obstructed Event 1 retrieval and a partial match between Events 1 and 3.

The goal of the experiments was to evaluate an explanation of set-alternation costs based on episodic memory principles. The assumption is that performance of any task is a consequence of memory retrieval processes that involve representations of specific prior experiences (Kolers, 1976; Leboe, Whittlesea, & Milliken, 2005; Neill & Mathis, 1998; Tenpenny, 1995; Whittlesea, 1997; Whittlesea & Jacoby, 1990). When the Event 1 and 3 targets mismatch the retrieval of the Event 1 memory episode is not entirely appropriate for performing the Event 3 task. The interference due to a partial match between Events 1 and 3 might be the source of set-alternation costs. Results of Experiment 1 revealed larger costs in the high probability set-alternation condition. The high probability set-alternation condition encouraged retrieval of Event 1. However, because the targets of Event 1 and 3 mismatched the retrieval of Event 1 interfered with the processing of Event 3’s task-set. In other words, the interference due to a match in task-sets but a mismatch in targets generated costs. If set-alternations costs originate from a partial match between Events 1 and 3, increasing the amount of overlapping information between these events should reduce costs. The findings of Experiments 2 and 3 showed reduced set-alternation costs when there was a target identity match between Events 1 and 3. Lastly, Experiment 4 showed that set-alternation costs are larger when the retrieval of the Event 1 memory episode is obstructed. That is, costs were larger when there was a combination of obstructed Event 1 retrieval and a partial match between Events 1 and 3.