| Summary: | 碩士 === 國立臺灣大學 === 心理學研究所 === 101 === Internal attention reflects top-down control of representations. Through refreshing representations and recurrent processing, internal attention can protect and strengthen the representation in visual working memory. However, whether there is a minimum duration for top-down control to effectively modulate visual working representations remains unclear. Moreover, whether representation strength influences this minimum duration has not been explored. The aim of this study is to address these two issues using a change detection paradigm while varying the duration for top-down control to modulate a representation cued by a spatial indicator. In Experiment 1, a memory array of four colored shapes was presented for 180 ms, a retro cue was shown 800 ms after array offset, and the duration for retro cueing until the display of a test array was manipulated to be 100, 300, 500, or 700 ms. The retro cue could be 100% valid for indicating the item to be tested or neutral that was uninformative of the to-be-tested item. The participants’ task was to detect whether there was any change between the two arrays. The minimum duration for retro cueing to benefit memory performance was operationally defined as the threshold for top-down modulation to effectively protect visual short-term memory. The results showed cueing cost when the duration was too short (100 ms) and benefit when the duration was long (700 ms) compared with the performance in the neutral-cue condition. The threshold was 700 ms. In Experiment 2, we increased the encoding time to 300 ms for strengthening stimulus representations. Neither the cost nor the benefit was observed. In Experiment 3, the interval between the offset of memory array and retro cueing was lengthened to 1000 ms so that strong representations may decay to weaker ones. The results showed that the threshold for top-down modulation to effectively protect memory representations was 500 ms. In Experiment 4, a within-subject design was adopted to investigate the impact of encoding time on memory performance using an interval of 800 ms between the offset of the memory array and the retro cue. In Experiment 5, a within-subject design was adopted to examine the effect of cue onset time on the performance of stronger representations that resulted from 300 ms encoding time. The results of these two experiments replicated the results from Experiment 1 to Experiment 3. In sum, top-down modulation needs a minimum duration for effectively protecting visual working memory. Moreover, this threshold depends on the representation strength. It appears that the strength of visual working memory representations must be in a certain range for top-down modulation to bring forth the benefits. Future research should explore this range using more rigorous methods to manipulate representation strength.
|
|---|
