Delineating the Linguistic Pathway of Early Mathematics Development: The Role of Mathematical Language

• In Event: More Than Just Numbers: Varied Predictors of Mathematics Knowledge

Abstract

Introduction: A considerable body of evidence indicates development of early mathematics skills and language/literacy skills is tightly linked. In fact, LeFevre and colleagues (2010) indicated that, of the three pathways of numeracy development (linguistic, spatial, and quantitative), the linguistic pathway was the strongest and most stable. In most work examining the relation between these domains, measures of general vocabulary are used. However, recent evidence indicates that mathematical language (e.g., knowledge of content-specific terms and concepts such as more, few, near, far) is more proximal to mathematics performance than general vocabulary. Further, other language and literacy skills such as syntactical awareness (Nelson et al., 2011), print knowledge (Purpura et al., 2011), and phonological awareness (Neumann et al., 2013) are also linked to mathematics development. Delineating the aspects of language that are uniquely related to numeracy skills in a comprehensive model is needed. Based on prior evidence, we hypothesize that quantitative language, syntactical awareness, and print knowledge would be significant predictors of numeracy performance.

Methods: Participants included 124 3- to 5-year old (M = 4.8 years) preschool children representative of local demographics (76.6% Caucasian; 54% male; 52.1% parents with less than a 4-year degree). Children were assessed on numeracy, quantitative language, spatial language, receptive vocabulary, expressive language, syntactical awareness, phonological awareness, and print knowledge. Covariates included, RAN, IQ, age, gender, and parent education. A four-step hierarchical regression was conducted: (Step 1) Covariates were included. (Step 2) Quantitative and spatial language were added to the model. (Step 3) Expressive vocabulary, receptive vocabulary, and syntactical awareness were added to the model. (Step 4) Phonological awareness and print knowledge were added to the model.

Results: (Step 1) The covariates accounted for 52% of the variance in numeracy performance. (Step 2) When quantitative and spatial mathematical language were added into the model, quantitative language (β = .31, p < .001), but not spatial language (β = .04, p = .673) significantly predicted numeracy performance. (Step 3) When vocabulary and syntax were added, quantitative language remained a significant predictor. However, neither expressive (β = .01, p = .904) nor receptive (β = .08, p = .414) vocabulary were significant. Syntax was a significant predictor (β = .16, p = .044). (Step 4) In the final step of the model, inclusion of the phonological awareness and letter sound measures, the model accounted for 68% of the variance in numeracy performance. Quantitative language remained a significant predictor (β = .20, p = .006). Both phonological awareness (β = .18, p = .007) and letter sound knowledge (β = .25, p < .001) were significant predictors of numeracy performance. However, syntax was no longer a significant predictor (β = .08, p = .303).

Discussion: As expected, quantitative language and print knowledge were related to numeracy performance. However, contrary to expectations phonological awareness was related to performance and syntactical awareness was not. These findings may be due to the nature of tasks assessed and the ages and developmental abilities of the participating children. Implications, limitations, and future directions will be discussed.

Authors

• David J. Purpura, Purdue University

• Sarah Schmitt, Purdue University

• Amy Napoli, Purdue University