Abstract
Visual working memory (VWM) holds visual information in mind to achieve task goals. Performance varies based on factors such as stimulus modality and distractor number. Several different tasks have been used to measure VWM, each resulting in differing estimates of capacity. Our goal was to examine how task demands can influence neurocognitive measures of VWM during early childhood. We administered two tasks to 4- and 5-year-olds that were identical during visual encoding and maintenance, only differing in type of response required during recall. In the change detection task (CD), a visual stimulus was compared with a VWM representation. In the cued-recall task (CR) a VWM representation was verbally labeled. Six canonical colors were used (red, orange, yellow, green, blue, purple). Neural data were collected from bilateral frontal, temporal, and parietal regions using functional near-infrared spectroscopy (fNIRS). VWM capacity was estimated to be higher in the change detection task (k4=2.48, k5=2.54) compared to the cued recall task (k4=1.13, k5=1.84) (p < .001). A network of frontal, temporal, and parietal regions was engaged across tasks. When comparing the tasks, CR showed increased activation in both the left supramarginal gyrus, F(1,35) = 11.642, p = 0.002, and the right superior parietal cortex, F(1,35) = 7.915, p = 0.008. Further, higher activation was observed for 5-year-olds during CR in right supramarginal gyrus, F(1,34) = 8.295, p = 0.007. During the CD task, right inferior frontal gyrus showed increased activation at set size two, F(2,64) = 3.643, p = 0.032, which is around capacity for this task. Thus, the CR task was observed to be more difficult than the CD task, and 5-year-olds showed increased activation in right parietal cortex during the CR task. These data suggest that children improve at engaging compensatory neural mechanisms during more challenging tasks involving the labelling of VWM representations.