Varying temporal contiguity and interference in a human avoidance task

Temporal contiguity and number of interfering events were manipulated in a human avoidance (Martians) task, which required participants to prevent an "invasion" when a particular visual stimulus ("shield") appeared by releasing the space bar before the shield was activated. A particular symbol, 1 of up to 6, functioned as a brief warning signal. The signal-offset to shield-onset (S-S) interval varied between groups, as did the number of additional symbols acting as distractors. In Experiments 1 and 2, speed of learning declined as a linear function of both trace interval and number of distractors. Path analysis showed that the effects of the S-S interval depended primarily on the number of distractors during this interval. Experiment 2 further demonstrated that participants who failed to suppress responding were generally unable to identify which symbol was the signal, suggesting that the presence of distractors disrupted detection of the contingency rather than performance. Overall, the results indicated that learning to associate 2 temporally separated events depends mainly on the amount of interference and little on the time interval between them.