Poster #42 - Adult-provided versus children’s spontaneous use of spatial words on spatial performance

• In Event: 1-005 - Poster Session 01
  In Poster Session: PS 01 Section - Cognitive Processes

Integrative Statement

Relations between language and spatial skills over development are well-documented but questions remain as to how such relations emerge. Children’s spatial task performance correlates with their production of task-relevant words (e.g., Hermer-Vasquez et al., 2001) and can be facilitated by experimenter-provided language (e.g., Miller et al., 2016). One potential explanation of such findings is that spatial word acquisition enables verbal encoding of relevant spatial features, enhancing performance. However, Farran & O’Leary (2016) showed that 4-year-olds only benefitted from spatial language in a feature-binding task if they knew the relevant words (measured by production and comprehension). This suggested that children did not spontaneously use their spatial word knowledge to support their performance. The current study extends these findings by testing this effect in a different spatial task and a wider age range. Our measure of spatial word knowledge is only production to test whether children who produce task-relevant language in one context may fail to use that language to support spatial performance.

We tested 117 children (Mage=4.54 years, range=3.32-6.56) in two tasks measuring use of objects’ relative positions to recall versus describe spatial locations (following Miller et al., 2016, where 4-year-olds’ performance was previously reported). In the recall task, children searched for a toy within an array of cups and landmarks after a 10s delay. Across trials, the child and/or array could be rotated during the delay (Figure 1), misaligning egocentric and/or room-centered reference frames between hiding and search to isolate children’s reliance on an intrinsic reference frame (defined by objects’ relative positions). In this task, children perform best in neither-move (all reference frames aligned) and worst in table-move (only intrinsic reference frame aligned; Miller et al., 2016; Nardini et al., 2006). Children were assigned to the verbal or control condition, which differed in the experimenter’s cues during hiding (“I’m hiding the toy [by the cow]/[here]”). Before the recall task, children were asked to describe the cup locations within the array, and we grouped children for analysis based on their use of relational terms (presence/absence).

We compared whether search accuracy across conditions and rotation types differed between children who did (n=73) versus did not (n=44) produce relational terms, controlling for age. As in Miller et al. (2016), we found effects of rotation (neither-move highest, table-move lowest) and condition (verbal>control) on recall accuracy (Figure 2). We also found differences between production groups (present>absent) and an interaction between rotation and production (equal in neither-move, present better otherwise). Interestingly, condition differences were present in both production groups, indicating that some children produced relational terms in the description task, but such knowledge was insufficient to support mature recall performance, as experimenter-provided language augmented recall. We are currently developing a probabilistic model to further understand how children’s strategies change based on rotation, condition, and production ability. Our results demonstrate that children do not necessarily use verbal encoding for recall even if they have acquired relevant language. Future research should consider other processes involved in spatial development and verbal strategy use.

Authors

• Hilary Miller, Emory University
  Presenting Author

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