| Prospective memory (PM) is remembering to complete tasks that have been planned for the future, and consists of two components: prospective (remembering when to perform the task) and retrospective (remembering the task itself). Two experiments examined the effectiveness of forming implementation intentions on PM as well as further examining the multiprocess model in PM. Forming implementation intentions, found to improve PM in healthy, elderly adults, required participants to close their eyes and picture themselves completing the task, and saying out loud that they intended to complete the task. In Experiment 1, adult participants, half of whom had multiple sclerosis, played a board game that included dozens of PM tasks. Participants were instructed to form implementation intentions or to rote-rehearse the tasks. MS patients performed worse on the prospective and retrospective components of PM, and on both time-based and event-based PM tasks. This deficit for MS patients was alleviated by forming implementation intentions, especially after practice. Healthy control participants did not benefit from forming implementation intentions. In Experiment 2, college students were instructed to form implementation intentions, rote-rehearse, or use a daily planner to aid performance on the same PM tasks. The daily planner in Experiment 2 improved retrospective performance on the game, while forming implementation intentions actually hurt performance in college students.; It was concluded that forming implementation intentions helped the MS patients, a group that did not spontaneously visualize the tasks. In groups that did spontaneously visualize the tasks, healthy, middle-aged adults and college students, the results were more mixed. Perhaps, among people who do spontaneously visualize, forming implementation intentions only helps if it makes a plan above and beyond what is already there.; Mixed support for the multiprocess model was found in both Experiments. In Experiment 1, there was support for reflexive-associative processing for both control participants and MS patients. In Experiment 2, there was support for discrepancy-plus-search processing for both non-associated and highly-associated tasks. A finding that salience increased performance on the retrospective component for highly-associated tasks indicates that perhaps the multiprocess model should be supplemented. For example, it might be possible that the prospective and retrospective components are not independent {09}of each other. |
