Development of mental manipulation in working memory in young children

• In Event: Poster Session 13
  In Poster Session: PS 13 Topic - Attention, Learning, Memory

Abstract

Children live in a dynamic environment, and they therefore must store and manipulate information in working memory in order to learn and make decisions. Previous work suggests that the ability to store information in working memory increases through infancy to late childhood, but less is known about the development of the ability to manipulate stored information. In one study, 25-month-olds could track identities of two moving objects during occlusion (Cheng, Kaldy, Blaser, 2019), and another study found that 7-year-olds were able to update color-location bindings while occluded objects swapped places (Pailian et al., 2020). Here, we asked whether storage and manipulation share the same resources in working memory, and how the ability to manipulate information in working memory develops in early childhood.
Sixty-one children (M=5.7; range=4.1-7.1) completed an animated ‘hide-and-seek’ working memory task online via videoconferencing. The task consisted of two blocks of trials. In the Static block (Figure 1b), children viewed sets of 3, 4 or 6 cards depicting illustrations of animals, which were then covered by animated opaque cups. Next, images of two of the hidden animals appeared over one of the cups, and children were asked to choose which of the two animals was hidden in that location. In the Manipulation block (Figure 1c), children viewed sets of 2 or 3 cards, which were then covered with the opaque cups. The cups then swapped locations either 1, 2, or 3 times, after which we probed children’s memory for the card at one of the locations. To succeed, children had to store the locations of each animal card in working memory (Static block) or update the cards’ locations as their occluders moved through space (Manipulation block).
In the Static block, children’s performance was above chance at each Set Size (all p<.001), but performance also decreased as Set Size increased (p=.04). Contrary to previous work showing development of working memory for static displays in this age range (e.g. Simmering, 2012; Applin & Kibbe, under review), we observed no main effect of Age Group (4-, 5-, 6-year-olds; p=.16), suggesting that children’s working memory capacity for the static arrays in our task (which consisted of complex semantically-rich images rather than simple features) was relatively stable across our age range. In the Manipulation block, children’s performance decreased with Set Size (p=.008) and increased with age (p=.006). Interestingly, we observed no main effect of Number of Swaps (1, 2, or 3) (p=.94) and no interactions (all p>.3) (see Figure 2). Overall, children’s performance in the Static block was significantly better than in the Manipulation block (p=.03). Together, these results suggest that manipulation of information in working memory is more demanding than storage of information, but that children’s ability to manipulate information in working memory improves with age.
Our results suggest that working memory storage and manipulation may follow different developmental trajectories in early childhood, and that storage load and manipulation load may impose different costs to working memory in young children.

Authors

• Chen Cheng, Hong Kong University of Science and Technology
  Presenting Author

• Melissa M. Kibbe, Boston University
  Non-Presenting Author