Purchase this article with an account.
Kirsten Adam, Edward Vogel; Trial feedback and incentive structures decrease failures of working memory. Journal of Vision 2015;15(12):663. doi: https://doi.org/10.1167/15.12.663.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Trial by trial fluctuations in attentional control can lead to failures of working memory (WM), in which the subject is no better than chance at reporting the items from a recent display. In three experiments, we used a whole-report measure of visual WM to examine the impact of pre- and post-trial information on the rate of these failures. We hypothesized that subjects’ knowledge about the upcoming task load (pre-trial) or instantaneous feedback about performance (post-trial) would reduce failures and improve average WM performance. In each experiment, subjects remembered an array of colored objects (~200 ms) across a blank delay (~1300 ms) and then reported the identity of all items. We defined performance failures as the trials in which subjects were at chance for reporting all items in the array (0 or 1 correct items). To manipulate pre-trial information, we presented subjects (N=18) with a cue before the memory array that was either informative (the set-size) or uninformative (an “X”) about the upcoming trial load (2 to 8 items). We found that failure rate and overall performance were identical for both conditions, suggesting that subjects could not take advantage of this pre-trial information. To test the impact of post-trial information, we compared no feedback to feedback about the number of items correct (N=15) after each trial. Feedback decreased failure rates from 15% to 10%. In the third experiment (N=49), we used a point-based incentive structure in the feedback condition in which failure trials resulted in lost points and consistent successful performance received “streak” points. Again feedback decreased failure rates from 11% to 6% and benefitted all subjects irrespective of average capacity. Together, these results suggest that while pre-trial knowledge of task load is ineffective, post-trial feedback is highly effective at decreasing the rate of WM failures within a session.
Meeting abstract presented at VSS 2015
This PDF is available to Subscribers Only