Talk Spatial to Me: Developing a Broad Measure of Receptive Spatial Vocabulary

• In Event: Poster Session 07
  In Poster Session: PS 07 Topic - Cognitive Processes

Abstract

Preschoolers’ spatial development has recently received much empirical attention (Casasola et al., 2020; Geer et al., 2019; Rittle-Johnson et al., 2019; Verdine et al., 2017). Early spatial skills predict and promote later spatial skills and mathematics knowledge (Bower et al., 2020; Casasola et al., 2020; Cheng & Mix, 2014; Geer et al., 2019; Mix & Cheng, 2012; Rittle-Johnson et al., 2019; Verdine et al., 2017). Variation in spatial skills exists partly due to exposure to spatial language from parents (Hart et al., 2016; Polinsky et al., 2017; Pruden et al., 2011; Szechter & Liben, 2004; Verdine et al., 2014); thus, efforts have been made to measure the spatial vocabulary of preschoolers (Bower, Foster, et al., 2020; Casasola et al., 2020). However, spatial language is conceptualized differently (e.g., sub-types such as locations/directions, features/properties, shapes, and dimensions and “what” vs. “where” words, or locations/directions vs. other subtypes). The current study aimed to determine the best fitting dimensional structure of early spatial vocabulary using existing and novel receptive spatial vocabulary items, and to examine relations between spatial vocabulary with measures of math language and skills. A Head Start sample was recruited because children from low-income backgrounds often hear fewer spatial words from parents and perform worse on spatial tasks than children from middle-income backgrounds (Bower et al., 2020; Levine et al., 2005; Verdine et al., 2014; 2017).

Participants (n = 104) were between 3.17 and 5.26 years old (M = 3.59, SD = 0.55; 51% female) and assessed on various math/spatial language and numeracy measures. Their parents reported demographic information and the home numeracy and spatial environments (See Table 1 for measure details/descriptive statistics).

Three theoretical confirmatory factor analyses were performed: a unitary factor model (Model 1), a two-factor “where” and “what” word model (Model 2), and a third 5-factor model (i.e., dimension, feature/property, location/direction, orientation/transformation, shape). The first two models demonstrated good fit (Model 1: χ2= 3157.45, df= 3002, CFI= .92, TLI= .92, RMSEA= .02; Model 2: χ2= 3140.64, df= 3001, CFI= .93, TLI= .93, RMSEA= .02), while the third did not converge. A χ2 difference test determined that the two-factor model (Model 2) fit better than the one-factor model (χ2 difference= 16.813, df= 1, p= .000041). We then correlated both spatial language factor scores with math/spatial language and numeracy measures (see Table 1). Both the “where” and “what” words correlated moderately-strongly with similar constructs. “Where” words significantly correlated with all measures except the home spatial environment, number line estimation, and measures of quantitative math language and spontaneous focus on number/counting. “What” words significantly correlated with all measures except number line estimation.

Preschoolers’ receptive spatial vocabulary is best categorized as “what” vs. “where” words, though future work should further evaluate the comparative fit of the 5-factor model. Validity evidence was found for both spatial vocabulary types. Findings will be discussed in terms of implications for measuring early spatial thinking via space-specific receptive vocabulary, as well as the role of spatial vocabulary type in the development of specific aspects of mathematics knowledge.

Authors

• Lauren Westerberg, Purdue University, West Lafayette
  Presenting Author

• Erica L. Zippert, Purdue University, West Lafayette
  Non-Presenting Author

• Alexa Ellis, University of Alabama, Tuscaloosa
  Non-Presenting Author

• Jimena Cosso, Pennsylvania State University, University Park
  Non-Presenting Author

• Yemimah King, Purdue University, West Lafayette
  Non-Presenting Author

• Connor O'Rear, Purdue University, West Lafayette
  Non-Presenting Author

• David James Purpura, Purdue University, West Lafayette
  Non-Presenting Author