Poster #30 - Learning from Self-Produced Actions and Gestures: The Role of Working Memory

• In Event: 3-159 - Poster Session 12
  In Poster Session: PS 12 Section - Cognitive Process

Integrative Statement

Asking learners to produce gestures, movements of the hand that can convey information, can lead to insight and improved learning outcomes (Novack & Goldin-Meadow, 2015). Yet recent work in the domain of linear measurement has demonstrated that for some students, gesture instruction is less effective than learning from actions-on-objects (Congdon, Kwon & Levine, 2018). In the current study, we ask two questions: 1) is gesture’s transient nature too challenging for some learners to understand and 2) which characteristics of the learner herself might explain some of the variance in learning outcomes?

In this ongoing study with first grade students (to date, N=143, Mage=6.99 years), all children were given a diagnostic linear measurement pre-test to determine their dominant initial strategy on test items where the object to-be-measured was shifted away from the start of a ruler. Children consistently used one of two incorrect strategies, ‘reading-off’ the number at the end of the object or ‘hatch-mark counting’ – counting the lines beneath the object. After pre-test, participants were randomly assigned to one of three between-subjects training conditions: action-on-objects (a handful of plastic unit chips to be placed on the ruler under the object); gesture (a thumb-and-forefinger pinching gesture to be moved along on the ruler under the object); or iterated action (a single unit chip, iterated along the ruler). The iterated action condition was designed to test whether a transient, though still object-centered training procedure, would prove challenging to learners like transient gesture-based instruction or be highly effective like non-transient action-on-objects instruction. To address the second research question about characteristics of the learner, children were also given two working memory measures (one spatial, one verbal).

Preliminary analyses reveal that overall, children in the hatch-mark counting group improved significantly more than those in the read-off group across all three training conditions (p’s <.01). Within groups, children who began the study using the incorrect ‘hatch-mark counting strategy’ improved markedly in each of the three training conditions (Figure 1A). By contrast, children who started the study using the incorrect and less sophisticated ‘read-off strategy’ showed more improvement in the action-on-objects and iterated action training conditions than in the gesture condition (Figure 1B).

Children in the read-off strategy group performed significantly worse on both the verbal working memory measure (p < .01) and the spatial working memory task (p < .001), but their weakness in spatial working memory was significantly larger (p < .05, Figure 2). Taken together, these preliminary findings suggest that 1) gesture’s transient nature is likely not the primary reason that some children struggle to learn from gestures. There may be other features of gesture that pose challenges, such as the fact that it is representational. And 2), it is clear that children who use a read-off strategy have relatively weak spatial working memory compared to verbal working memory. Analyses on the final sample will explore whether this working memory deficit may explain some of the variability in learning outcomes on this highly spatial linear measurement task.

Authors

• Eliza L. Congdon, Bucknell University
  Presenting Author

• Susan C. Levine, University of Chicago
  Non-Presenting Author