Science with Duplo: Multilevel goal management in preschoolers’ toy house constructions

• In Event: Poster Session 07
  In Poster Session: PS 07 Topic - Cognitive Processes

Abstract

The planning of goal-directed action sequences develops over preschool years. Five-year-olds are able to plan and execute their action sequences based on a goal hierarchy, while 3- and 4-year-olds have difficulties following the highest goal in this goal hierarchy (Freier et al., 2017; Yanaoka & Saito, 2017; 2019). This study aimed to investigate whether improvements in executive functioning could underlie advancement in planning of action sequences. Furthermore, it aimed to find implicit markers of hierarchical action sequence planning using optical motion capture. Motion capture can reveal how reaching kinematics is influenced by dynamic decision processes (Song & Nakayama, 2009; Freeman et al., 2011).
Sixty-nine 3-, 4- and 5-year-olds used Duplo to build a house in an action sequence planning task with a hierarchical goal structure. The main goal of building a house consisted of multiple subgoals (walls, roof), which each in turn consisted of multiple action steps. Behavioural data of whether children were able to follow the main goal and the subgoals, and whether children disregarded distractor objects was coded afterwards. Individual difference in inhibition, working memory and updating, and set shifting were assessed.
Binary logistics regressions were used to investigate whether age and executive function task scores were predictor of the behavioural scores on the action sequence planning task. The results showed that age was a strong predictor of completing the house. 5-year-olds were more likely to be successful at finishing building the house than 3-year-olds (χ2(1) = 7.10, p = .008, Nagelkerke R2 = .114, Figure 1), indicating that they were better at keeping track of the main goal. Furthermore, inhibition skills were associated with avoidance of actions irrelevant to the goal hierarchy (χ2(1) = 7.19, p = .007, Nagelkerke R2 = .175). Lastly, improvements in updating abilities were related to executing actions relevant to the different subgoals (χ2(1) = 4.12, p = .042, Nagelkerke R2 = .086).
Motion capture was used to investigate reaching movement during which children switched from one subgoal to another. Children who followed the subgoals displayed less movement with their non-reaching hand within a subgoal (t(57) = -2.07, p = .043, Cohens d = -.548, Figure 2). This was hypothesized to be the result of an increased task focus while executing actions within a subgoal, which resulted in relative freezing of the non-reaching hand during the subgoal actions. Conversely, at a branch point the non-reaching hand relaxed reflecting planning the next subgoal actions. This effect was enhanced in children with better inhibitory skills (r(59) = .329, p = .011), suggesting that such skills allow greater focus on executing the current subgoals.
To conclude, this study shows that improvement in action sequence planning occur over preschool period and are associated with improvements in executive functions. Earlier studies provided some indirect evidence that preschoolers can represent action sequences in terms of goal hierarchies, however, we demonstrate the first direct evidence for structuring of subgoals in the action sequence planning. Again, the preschool period emerges as a critical time for the development of effective cognitive behaviour.

Authors

• Lisanne Schroer, Erasmus University Rotterdam
  Presenting Author

• Richard P. Cooper, Birkbeck, University of London
  Non-Presenting Author

• Denis Mareschal, Birkbeck, Universty of London
  Non-Presenting Author