Poster #7 - An Exploratory Analysis of Preschoolers’ Structure of Spatial Language Across Socioeconomic Status

• In Event: Poster Session 07

Abstract

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). Though these spatial skills are malleable (Uttal et al., 2013), high socio-economic status (SES) children consistently score higher than lower SES children on spatial assessments (Verdine et al., 2017), including a spatial language task (Bower et al., 2020). Spatial language consists of words and phrases used to describe spatial relations between objects (e.g., over, under) and properties of objects (e.g., circle, tall).
Here we have taken a look at how spatial language, influential in spatial skill problem solving (Feist & Gentner, 2007), develops between higher and lower SES children. Specifically, we examined how the factor structure of spatial language may differ between low and high SES children and how this organization may differentially be associated with task performance. We have used a sample of preschool-aged children in particular to view the importance of early home environments, which play a key role in language development and academic skills (Son & Morrison, 2010), and how spatial language develops in a pre-formal education context.
Three-year-olds (N=331; Mage=42.98 months, SD=3.24; 49% high-SES) were tested on a spatial language comprehension task in which they selected one of three photographs that represents the target spatial relation (e.g., “the bear is in front of the bucket”). Higher SES (measured as parents’ education) was correlated with greater success on the task (r = .24, p = .010). We then conducted an exploratory factor analysis (EFA) to assess how children’s knowledge of specific spatial relations clustered between SES groups. Within the EFA we found a tighter organization with higher SES children (3 factors) than with lower SES children (6 factors). Moreover, among high-SES preschoolers, three items loaded highly across several factors (factor loadings >= .40), whereas for low-SES preschoolers, two items loaded highly across several factors.
The results suggest that a smaller number of clusters represents a more interconnected network, which in turn is associated with more efficient and accurate performance on the spatial language comprehension task. As “exploratory” in EFA suggests, it is but a preliminary examination of the data; in the future we will conduct more thorough data analyses using network analysis techniques. We seek to elucidate these results, especially as they may help inform future work on how spatial language associates with strategies to solve spatial problems. Still, these preliminary results indicate 1) the structural organization of spatial language between high and low SES children and, 2) how early home environments may be influential in not only spatial language task accuracy, but the factor structure of spatial language.

Authors

• Wilder VonSchonfeldt, California State University, Los Angeles
  Presenting Author

• Corinne A. Bower, California State University - Los Angeles

• Kathy A Hirsh-Pasek, Temple University

• Roberta M Golinkoff, University of Delaware