Poster #31 - Longitudinal Effects of Spatial Training on Preschoolers' Spatial and Math Outcomes

• In Event: 3-206 - Poster Session 13
  In Poster Session: PS 13 Section - Cognitive Processes

Integrative Statement

Spatial skills, or the ability to mentally and physically manipulate objects or spaces, are ubiquitous and predict later achievement in STEM disciplines (Mix & Cheng, 2012; Wai et al., 2010). However, prior research finds that males outperform females on spatial tasks (Uttal et al., 2013) and children from high-income households outperform their middle-income peers (Verdine et al., 2014). Fortunately, spatial skills are malleable (Uttal et al., 2013), but little is known about which interventions optimize spatial and math skills. Thus, this study examines longitudinal effects of a spatial training program on spatial and math outcomes and to see if these effects vary by sex or SES.
Preschoolers were trained on a 2D-puzzle task comprised of different colored geometric forms (see Figure 1) and randomly assigned to one of three training conditions, given once per week over 5 weeks, or a no-training control. Gesture (Ping et al., 2011) and spatial language (Feist et al., 2007) are particularly useful in spatial learning and were further explored as malleable spatial training factors. Feedback-only corrected wrongly placed puzzle pieces. Gesture highlighted puzzle piece shapes and orientations as the trainer used gesture. Spatial language emphasized shape names and spatial relations between pieces. Three-year-olds (Mage=42.65 months, SD=3.37; 50% high-SES) were pre-, post-, and follow-up tested (9-12 mo after pre-test, M=9.58mo, SD=0.70) on a 2D-spatial assembly task (N=40; 21 females) and math assessment (Woodcock Johnson-Applied Problems) (N=79; 41 females).
We hypothesized that children in the three treatment conditions would outperform control group children on the follow-up spatial assembly task. We also hypothesized that girls’ and low-income children’s spatial skills would be impacted the most with training. Second, we hypothesized that children in the spatial language condition would outperform children in the other three conditions on follow-up math outcomes due to the additional puzzle piece shape descriptions (e.g., “a triangle has three sides”).
Preliminary results analyzing the longitudinal follow-up indicate there was no effect of individual training conditions on spatial or math outcomes. Therefore, we collapsed the individual training groups to compare children who received ‘any training’ to the control group. First, there was no effect of spatial training on children’s, including girls and low-income children, long-term spatial skills. Second, there was a significant interaction on children’s math scores, p=.001, such that males’ math scores benefited from any training, p=.033 (see Figure 2). There was no SES moderation.
The finding that males with any spatial training increased their math skills is interesting given that there was no long-term training effect on spatial assembly skills—possibly due to the small sample size at follow-up. Thus, this spatial training could have improved males’ immediate spatial skills or other cognitive processing skills (e.g., executive function) during the intervention to allow cascading effects on math performance. It is unclear why females’ math scores did not benefit from training. Nevertheless, these findings provide evidence of an earlier causal link between spatial and math skills than prior studies have documented.

Authors

• Cori Bower, Temple University
  Presenting Author

• Laura Zimmermann, SRI International
  Non-Presenting Author

• Brian N. Verdine, University of Delaware
  Non-Presenting Author

• Lindsey Foster, University of Delaware
  Non-Presenting Author

• Siffat Islam, Temple University
  Non-Presenting Author

• Roberta Michnick Golinkoff, Univ of DE
  Non-Presenting Author

• Kathy Hirsh-Pasek, Temple University
  Non-Presenting Author