Using head-mounted eye-tracking to understand effects of language on children’s spatial skills

• In Event: 3-029 - Process-based approaches to understanding spatial development: the role of attention, language, and gesture

Integrative Statement

Spatial skills are essential for everyday cognition and support academic achievement. Spatial language and visual attention are two candidate processes underlying spatial development (Pruden et al., 2011; Miller & Simmering, 2018). Most research has focused on language, with limited consideration to visual attention, in part, due to the use of only outcome-based measures (i.e., accuracy). For instance, experimental studies showed that adult-provided verbal cues during spatial tasks facilitated accuracy while non-verbal visual cues had minimal effects (Dessalegn & Landau, 2008; Miller et al., 2106). By relying on outcome-based measures, it is unknown why verbal cues are more effective than visual cues and whether and how language and visual attention relate to support spatial performance.

This study used process-based measures to understand how children attend to visual information during spatial tasks and how verbal and non-verbal manipulations influence children’s attention to support performance. While wearing a head-mounted eye-tracker, 94 four- to five-year-olds participated in two blocks of a spatial recall task assessing use of an intrinsic reference frame (memory for relations among objects; see Figure 1). The task involved finding a toy under a cup on an array of cups and landmarks after a 5s delay and the array rotating. Children first completed a baseline block, receiving non-specific cues (“I am hiding the toy here, remember look at the table”) on each trial. Then children completed an experimental block in one of three randomly-assigned conditions: verbal (“… here, between the bear and cow”); visual (“… here”, shook the bear then shook the cow); or control (same as baseline). The eye-gaze measure used was the number of fixations during hiding events to the target landmarks (two landmarks closest to target cup, cf. Figure 1).

We tested whether the condition manipulations: a) influenced accuracy; b) influenced eye-gaze patterns; and c) interacted with eye-gaze to influence accuracy. First, results showed that children’s accuracy improved significantly, relative to baseline, with verbal and visual cues. However, consistent with past research, verbal cues supported accuracy more than visual cues. Second, results showed that children’s target landmark fixations increased significantly, relative to baseline, in both verbal and visual conditions, with the visual condition increasing target fixations more. Third, we found main effects of condition and eye-gaze on recall accuracy (Figure 2). This showed that, when controlling for condition, more fixations on target landmarks predicted higher accuracy. However, there was unique variance associated with the verbal cues beyond this effect of eye-gaze. Follow-up analyses revealed that eye-gaze mediated effects of both the verbal and visual conditions, but the direct effect of condition on accuracy remained significant in the verbal condition, indicating that eye-gaze did not full account for this effect. Thus, visual attention can be directed to support spatial performance, and adult-provided language influences performance not only through directing visual attention. Overall, the results highlight the need to use process-based measures to understand spatial cognition and show that both language and visual attention support spatial performance.

Authors

• Hilary Miller, Emory University
  Presenting Author

• Heather Kirkorian, University of Wisconsin-Madison
  Non-Presenting Author

• Vanessa Simmering, ACTNext by ACT, Inc.
  Non-Presenting Author