Abstract: Prospective memory (PM) is the memory for actions to be performed in the future. There are several theories on the processing mechanisms underlying event-based prospective memory, such as the simple activation model, noticing-search model, preparatory attentional processes and memory processes theory, and multiprocess view. They differ from one another on whether PM uses cognitive resources at retrieval. Using behavioral experiments, previous researches have generally examined how a PM task interferes with a background task and have achieved inconsistent results. The present article aims at testing these theories, using the event-related potentials (ERPs) technique.
In this experiment, 20 junior undergraduates (10 women, 10 men) aged 19~26 years participated as paid volunteers. All participants were healthy, right-handed, and had normal or corrected to normal vision. Further, a typical paradigm for studying processing mechanisms underlying PM was used in this experiment. A triangle and a polygon were presented together in each trial. In the background task, participants were required to judge whether the polygon was on the left-hand side or on the right-hand side. In the PM task, participants were supposed to complete the background task; moreover, if the triangle was blue, they were required to determine the number of sides of the polygon and press a specified key if the number of sides was 5 or less and another key if it was more than 5. A session of the control condition involving only background trials was included at the beginning of the experiment, before participants had received any instructions concerning the PM trials. Subsequently, in the experimental condition, a PM task was administered to each participant. Brain electrical activity was recorded from 64 scalp sites.
As a function of non-PM trials, the accuracy and reaction time data in the two conditions were compared. The results indicated that the reaction time in the experimental condition was significantly longer than that in the control condition. The findings suggest that perhaps in the PM task, the participants were involved in strategic monitoring processes before the target event occurred. Data from the ERPs recorded during the non-PM trials suggest that there were significant differences in the time windows of 200~300 ms, 300~350 ms, and 350~400 ms after the onset of the stimulus. The ERP difference wave (subtracting the experimental condition from the control condition) showed the maximum amplitude over the central site (Fz) with a peak latency of 250 ms (N250). The topographic map in the time window of 200~300 ms indicated that the components mainly activate the frontal lobe.
These results might support the preparatory attentional processes and memory processes theory and partially support the multiprocess view. However, they fail to support the predictions derived from the simple activation model and the noticing-search model. Further, the possibility that PM retrieval might rely on multiprocess is discussed. The present results provide electrophysiological evidence for potential mechanisms underlying event-based prospective memory

Key words: event-based prospective memory, processing mechanisms, event-related potentials