Abstract:

Home literacy environment (HLE) can be more influential than reading intervention programs because HLE affects children’s language and literacy development from the very beginning. A better understanding of the developmental cascades of children’s HLE, character recognition, oral vocabulary knowledge, and reading comprehension is of great value to unpack and promote children’s reading development. The present study aimed to test these developmental cascades among Chinese children during their lower elementary grades.

This work followed 149 children from Grades 1 to 3. Their HLE was estimated based on information provided by their parents in Grade 1. Their character recognition, oral vocabulary knowledge, and reading comprehension abilities were assessed with age-appropriate measures once per school year, three times in total. A structural equation model was carried out to examine the developmental cascades of HLE, character recognition, oral vocabulary knowledge, and reading comprehension over the three testing time points.

Table 1 presented means and standard deviations for all variables. Obvious increases with grade were observed in character recognition and vocabulary knowledge. Table 2 presented the Pearson correlation matrix among all measures over time.

Structural Equation Model was conducted to examine the developmental cascades of HLE, character recognition, oral vocabulary knowledge, and reading comprehension (Figure 1). Results suggested the excellent model fit of the developmental cascade model, χ²(44) = 58.596, p= 0.069, CFI = 0.984, TLI = 0.972, RMSEA = 0.047 (90% CI = 0.000~0.077), SRMR = 0.069.

However, the path coefficients from Grade 1 vocabulary knowledge to Grade 2 character recognition (β = 0.09, p= 0.113), from Grade 1 reading comprehension to Grade 2 vocabulary knowledge (β = −0.07, p= 0.415), from Grade 2 character recognition to Grade 3 vocabulary knowledge (β = 0.12, p= 0.147), from Grade 2 reading comprehension to Grade 3 character recognition (β = 0.06, p= 0.216) were non-significant. To develop a more parsimonious model, the non-significant path was constrained to zero. We provided all the fit indices of the competing models in Table 3. Chi square difference testing was also conducted to compare the model fits.

As shown in Table 4, no significant Chi square difference was found when comparing the competing models. Based on the parsimony principle, Model 6 was selected as the final model.

Figure 2 showed this model with the estimated of the standardized path coefficients. Results showed that, (1) HLE spread through reading comprehension in Grade 1 (β = 0.19, p= 0.030) to character recognition in Grade 2 (β = 0.28, p< 0.001), then to reading comprehension in Grade 3 (β = 0.26, p= 0.002). (2) HLE spread through oral vocabulary knowledge in Grade 1 (β = 0.45, p< 0.001) to reading comprehension in Grade 2 (β = 0.20, p= 0.003), which in turn predicted oral vocabulary knowledge (β = 0.27, p< 0.001) and reading comprehension (β = 0.34, p< 0.001) in Grade 3. (3) The direct predictive effect of HLE on character recognition in Grade 1 (β = 0.16, p= 0.078) was marginally significant, but character recognition in Grade 1 predicted reading comprehension (β = 0.41, p< 0.001) and oral vocabulary knowledge (β = 0.14, p= 0.051) in Grade 2, which in turn predicted character recognition (β = 0.15, p< 0.001) and reading comprehension (β = 0.27, p< 0.001) in Grade 3.

These results together demonstrated the strong predictive power of HLE to children’s early reading development, and clarified its complex direct and indirect impacts on children’s character recognition, vocabulary knowledge, and reading comprehension over time. These findings help elucidate the way in which HLE may affect children’s reading development, which leads to the theoretical advancement and practical implications for HLE and children’s reading development.

Key words: Home Literacy Environment, vocabulary knowledge, character recognition, reading comprehension, developmental cascades