Intention after-effects refer to failure of deactivating completed intentions, making that completed intentions impact subsequent relative information process and cause commission errors. Two potential mechanisms facilitate successful intention deactivation. Overwriting of completed intention representation hypothesized that deactivation was depending on the difficulty of new information replacing the representation of the completed intentions. If this is correct, after-effects will increase when the representation of completed intention is stronger. Attentional-dependency account posited that monitoring stimuli associated with the completed intentions was depending on the degree to which attention was allocated to. Following this reasoning, after-effects might decrease attributing to a lack of available attentional resources. This article distinguished between these two possible hypothesizes.

Modified Experiment 1 of Scullin, Einstein, & McDaniel (2009) was utilized in this research. We manipulated relative importance of the task and attention mode between participants. Experiment was divided into four phases: the training phase, the active-PM phase, the completed-PM phase and the test phase. In active-PM phase, a trial type 2 (match, non-match) × relative importance of the task 2 (emphasis PM task, emphasis ongoing task I) × attention mode 2 (full attention, divided attention) mixed design was used, and the within-subjects factor was trial type. In the test phase, a word type 2 (cues, control words) × relative importance of the task 2 (emphasis PM task, emphasis ongoing task I) × attention mode 2 (full attention, divided attention) mixed design was used, and the within-subjects factor was word type. Four groups can be defined, namely PM task important-full attention group and PM task important-divided attention group (emphasis PM task); ongoing task important-full attention group and ongoing task important-divided attention group (emphasis ongoing task I). Ongoing task I was changed to color-matching task for investigating intention interference differences between these two conditions of relative importance of the task. The PM task was to press the “Q” key whenever cues (wheaten foods or insects) occurred in the color-matching task. There were two samples in these two categories, and they were used as cues served as control words matched on frequency. Ongoing task II was changed to artifacts decision task. Participants in the divided attention condition were also instructed to perform a random number generation task during tasks performance in completed-PM phase and test phase.

For color-matching task, longer RTs and higher PM cue hit rates were under PM task important condition; Longer RTs and higher error rates were under match condition. For artifacts decision task, error rates and RTs of cues and control words were similar in the full attention condition; However, in the divided attention, RTs were shorter of cues than control words.

The above results reveal that: (1) Executive function which is induced by relative importance of task will mediate PM performance by regulating attention allocation. Higher importance induces higher hit rate yet with more interference to ongoing task. (2) Completed intentions are deactivated. (3) Divided attention facilitates to deactivation of completed intentions. (4) Results of completed intentions under different attention modes support attentional-dependency account.