Predictable Switching Between Recognition Memory and Magnitude Judgement Tasks: Effects and Implications

Task switching and interruption effects—slower and (often) less accurate responses when a task changes compared to that when a task remains the same—have been investigated from both theoretical and applied vantage points (e.g., Altmann & Trafton, 2007; Jersild, 1927; Rogers & Monsell, 1995). The task switching research has typically used simple tasks with high stimulus-response (S-R) overlap, but there is a need to use different methods and tasks to test the boundaries of task switching effects and the theories used to explain them (Logan, 2003). This thesis examined the costs of switching between a recognition memory task, which is a more complex task than those typically used, and a magnitude judgement task (for the number of dots in a spatial array), which is the type of simple task that has been used (e.g., Altmann, 2002; Monsell, Sumner & Waters, 2003). Across seven experiments, participants switched between the recognition and magnitude tasks in predictable 1, 2, or 4-trial runs. The first two experiments examined task switching effects on recognition memory performance, with Experiment 2 investigating whether specific recognition processes (i.e., recollection and/or familiarity) were affected by switching tasks. Experiment 3 investigated the recovery from a task switch for both tasks and included a visual, task switch "reminder" cue in an attempt to improve switching performance. Finally, Experiments 4A, 4B, 5A, and 5B examined evidence for two well-known task switching phenomena, the practice effect and the preparation effect. The results led to four critical conclusions: (1) switching between two tasks with minimal S-R overlap produced significant RT and accuracy switch effects; (2) the cost to recognition memory accuracy did not reflect an impairment to controlled recollection processes; (3) the magnitude and persistence of task switching effects changed as a function of practice within an experiment; and (4) there was little evidence that participants began to switch tasks in advance of stimulus presentation. The results reported in this thesis provide a clear example of task switching driven by the type of stimulus (word or dots), where a change in stimulus type (i.e., from word to dots array or vice versa) initiated the time consuming process of retrieving/activating the appropriate task set. Future research will need to clarify whether the stimulus-driven nature of switching between the recognition and magnitude tasks remains when using different task switching paradigms and when S-R overlap is reintroduced.