Number-concept acquisition and general vocabulary development

Abstract

How is number-concept acquisition related to overall language development? Experiments 1 and 2 measured number-word knowledge and general vocabulary in a total of 59 children, ages 30-60 months. A strong correlation was found between number-word knowledge and vocabulary, independent of the child's age, contrary to previous results (D. Ansari et al., 2003). This result calls into question arguments that (a) the number-concept creation process is scaffolded mainly by visuo-spatial development and (b) that language only becomes integrated after the concepts are created (D. Ansari et al., 2003). Instead, this may suggest that having a larger nominal vocabulary helps children learn number words. Experiment 3 shows that the differences with previous results are likely due to changes in how the data were analyzed.

Figure 1

Relations among expressive vocabulary, receptive vocabulary, and knower-level variables. Concentric markers are used when data points overlap.

Figure 2

Examples of the samples drawn in Experiment 3. (A) A sample of 28 knower-levels drawn from a uniform distribution. (B) A sample of data from a simulated 2-knower from the Lee & Sarnecka (2010, model. (C) A sample of knower-level versus age and vocabulary. A total of 10,000 of these sets of samples were drawn, analyzed using both our method from Experiments 1 and 2, and also using the method described by Ansari et al. (2003). The results suggest that analysis of Give-N data by knower level has greater statistical power than analysis of the same data by percent correct. This difference appears large enough to explain why our Experiments 1 and 2 reached different conclusions than Ansari et al. (2003) did.