Poster #15 - Selective Offloading of Working Memory Demands in Children

• In Event: Poster Session 03

Abstract

Humans often supplement internal cognitive processing by manipulating the external environment (Clark & Chalmers, 1998). When confronted with the limits of our working memory, for example, we may choose to “offload” working memory demands into the environment by modifying the configuration of task-relevant stimuli (Risko & Gilbert, 2016). Indeed, recent evidence suggests that, among adults, manipulations of working memory task stimuli are associated with greater academic achievement (Bocanegra et al., 2019). Here we explored when and how children begin to augment internal visuospatial working memory by physically manipulating external stimuli.

Across two experimental phases, children aged 6 to 11 years (N = 81; 44 female, 37 male) were presented with sets of (i) of 2-5 constituent stimuli, together comprised of multiple black shapes and a single red shape, and (ii) a single target stimulus (see Figure 1). On each trial, children were asked whether, if combined, the constituents would look the same as or different from the target. Critically, the red shapes were designed to pop-out in vision, such that if the constituent and target red shapes were incongruent then children should have been able to easily identify that the stimuli were different rather than the same. The black shapes, however, did not pop out in vision, such that they had to be combined in working memory. In phase 1, children had to either identify the pop-out effect (when the red shapes were incongruent) or combine the constituent stimuli in working memory (when the red shapes were congruent). In phase 2, by contrast, the constituent stimuli were presented as transparent, stackable tiles, which children could physically stack to create a single percept that could be directly compared with the target.

In phase 1, children’s judgements of the constituent and target stimuli were less accurate when there were more constituent stimuli, but only when the pop-out shapes were congruent, χ2 = 18.82, p < .001, w = 0.48 (see Figure 2, top row). This pattern suggests that children were indeed attempting to combine the stimuli in working memory, while still showing the pop-out effect when applicable. In phase 2, older children were more likely to account for the number of constituent stimuli when deciding whether to stack them in the absence of a pop-out effect, χ2 = 52.33, p < .001, w = 1.38, suggesting these older children were more selective in their decisions to offload working memory demand. Intriguingly, however, older children were also more likely to account for the number of constituent stimuli when deciding whether to stack them in the presence of a pop-out out effect, χ2 = 22.41, p < .001, w = 0.89, when they did not need to offload working memory demand at all (see Figure 2, bottom row).

These findings suggest that across development, children’s cognitive offloading decisions become increasingly driven by metacognitive evaluations of task difficulty, rather than objective task difficulty. As in adults (Risko & Gilbert, 2016), these evaluations appear to drive both adaptive and redundant manipulations of external stimuli.

Authors

• Cate MacColl, University of Queensland

• Jonathan Redshaw, University of Queensland
  Presenting Author