Abstract: Prospective memory (PM) is defined as remembering to perform an intended activity. There are two types of prospective memory: time-based (TBPM) and event-based (EBPM). Interference occurs when one is processing ongoing task and prospective memory task at the same time, since processing both tasks consumes cognitive resources. The impact caused by keeping prospective intention on the accuracy and response time of processing ongoing tasks is often called prospective interference effect. This effect can be used as a way to study processing mechanisms between TBPM and EBPM. Since there is no external cue, TBPM is supposed to require more self-initiated processing. Recent studies have not found any different interference effect in TBPM and EBPM and they have concluded that the same cognitive processing mechanism exists in TBPM and EBPM. However, we believe that processing a time-based intention will cause more prospective interference to the ongoing task than processing an event-based intention. We hypothesized that manipulation of the length of the interstimulus interval (ISI) in the prospective memory tasks will provide insight on cognitive processing mechanisms in TBPM and EBPM.
Based on our hypothesis, there are 66 and 73 subjects participated in two experiments, where the ISI was defined to 2050~2250ms and 250ms respectively while each experiment includes control group, TBPM task group and EBPM task group. Participants in the control group only performed an ongoing task. In the TBPM group, participants were asked to press a specific key in a two-minute interval while they were performing the same ongoing task, and they could monitor time by pressing spacebar whenever necessary. In the EBPM task, participants were asked to press a specific key when an event cue was presented while they were performing the same ongoing task. An Analysis of Variance (ANOVA) was conducted on the average latencies and accuracy of ongoing task from the three conditions of each experiment. For TBPM, ANOVA was conducted on the number of times that participants checked the clock under different ISI.
In the first experiment when ISI was 2050~2250ms, we didn’t find different interference in terms of response time between time-based and event-based conditions, but we did find interference effect in accuracy in TBPM condition. Participants checked clock during interstimulus intervals. The frequency of clock checking within the 2-minute showed an ascending staircase curve. However, in the second experiment with 250ms ISI, we found that time-based intention had more interference in reaction time on the ongoing task than event-based intention. In addition, interference in terms of accuracy only occurred in TBPM condition, but not in EBPM condition. Participants checked clock once stimulus presented. The frequency of clock checking in the second experiment is different from the first one in that it increased gradually across the 4 time windows.
These results suggest that, when ISI is short enough, in both EBPM and TBPM conditions, participants had to monitor either event cue or time cue at the time of the stimulus presenting. At this situation, the TBPM task affected ongoing task more than EBPM task. Therefore, the TBPM task relies more on cognitive resource. Under different length of ISI, participants used different strategies to check clock. This implies TBPM has more complex cognitive processing mechanism. By using adaptive strategies based on the characteristics of TBPM tasks, one can optimize his/her cognitive resources. This study suggests different cognitive mechanisms play roles in TBPM and EBPM.

Key words: time-based prospective memory, event-based prospective memory, prospective interference effect, interstimulus interval, self-initiated process