How Musical Oddballs Warp Psychological Time

Oddballs—low-probability, attention-capturing expectancy violations—are judged as longer than non-oddballs, but are temporal intervals that contain oddballs judged as longer than those that do not? In 2 experiments, we tested competing model predictions using a novel and covert measure of subjective duration—musical imagery reproduction. Participants verbally estimated and reproduced with musical imagery repeated, coherent, or incoherent familiar or unfamiliar chord sequences (3.5 s, 7 s, or 12 s) that either did or did not contain dynamic auditory oddballs. Participants verbally estimated repeated chord sequences that contained oddballs as shorter than those that did not, but reproduced with musical imagery incoherent chord sequences that contained oddballs as longer than those that did not. These findings suggest that (a) intervals that contain attention-capturing, high-priority events are judged as shorter than those that do not when people are engaged in relatively temporal information processing, but as longer than those that do not when people are engaged in relatively nontemporal information processing, and (b) temporal and nontemporal information processing are interdependent. These results support the resource allocation model of short interval time estimation. We discuss implications for attention- and memory-based models, dynamic attending theory, and the ongoing debate about the mechanisms driving the temporal oddball illusion