Do Rare Stimuli Evoke Large P3s by Being Unexpected? A Comparison of Oddball Effects Between Standard-Oddball and Prediction-Oddball Tasks.

The P3 component of event-related potentials increases when stimuli are rarely presented. It has been assumed that this oddball effect (rare-frequent difference) reflects the unexpectedness of rare stimuli. The assumption of unexpectedness and its link to P3 amplitude were tested here. A standard- oddball task requiring alternative key-press responses to frequent and rare stimuli was compared with an oddball-prediction task where stimuli had to be first predicted and then confirmed by key-pressing. Oddball effects in the prediction task depended on whether the frequent or the rare stimulus had been predicted. Oddball effects on P3 amplitudes and error rates in the standard oddball task closely resembled effects after frequent predictions. This corroborates the notion that these effects occur because frequent stimuli are expected and rare stimuli are unexpected. However, a closer look at the prediction task put this notion into doubt because the modifications of oddball effects on P3 by expectancies were entirely due to effects on frequent stimuli, whereas the large P3 amplitudes evoked by rare stimuli were insensitive to predictions (unlike response times and error rates). Therefore, rare stimuli cannot be said to evoke large P3 amplitudes because they are unexpected. We discuss these diverging effects of frequency and expectancy, as well as general differences between tasks, with respect to concepts and hypotheses about P3b's function and conclude that each discussed concept or hypothesis encounters some problems, with a conception in terms of subjective relevance assigned to stimuli offering the most consistent account of these basic effects.