Interactions in short-term implicit memory and inhibition of return

Research presented in this dissertation examines the nature of interactions that arise between visual features in short-term implicit memory. When stimulus features change from trial to trial, repetitions and alternations of each of the component dimensions will interact to influence reaction time. Often, for a stimulus defined by multiple features, when the task-relevant feature repeats from the previous trial responding is fastest when the task-irrelevant features also repeat. When the response feature changes, responses are faster when the task-irrelevant features change as well. Nine experiments were conducted. Experiments 1 through 4 used stimuli defined by two features, using various combinations of Position, Color, Geometric Shape, and Letter Shape, and each feature served as both a task-relevant and task-irrelevant dimension between experiments. These experiments found that different features will influence each other asymmetrically and that influence may be partially determined by how quickly each features is processed. We also examine the duration of interactions through kernel analysis (Maljkovic & Nakayama, 1994; 1996; 2000). By eliminating the need to respond on every trial, Experiments 5 and 6 explore the role that executing a response plays in creating interactions. Interactions still appear between features in a cue-target design, but their strengths are not identical to those that arise in continuous responding tasks. Finally, Experiments 7, 8, and 9 argue that Inhibition of Return (IOR) appears to be a special case of this larger set of interactions that are generic to reaction time sequences where stimulus features change from trial to trial. IOR has been interpreted primarily in terms of mechanisms that prevent the return of attention to locations where it has recently been allocated, however we propose a new account of the classic IOR data pattern that is based on the ways in which object features are integrated in short term implicit memory.