Which Variables Better Predict Phonemic Awareness of First Grade Primary School Students?

Abstract

The purpose of the research was to discover which variables better predict phonemic awareness. Socioeconomic status (SES), quality of parent-child interaction (PCI), screen time (DST), visual-spatial ability (VSA), and mathematical reasoning (MR) were included as independent variables in the model, while phonemic awareness (PA) was the dependent (outcome) variable. The research was designed as correlational research. A total of 556 first grade primary school students were recruited into the research sample upon approval by their parents. In the analytic procedures, supervised machine learning was adopted and data were analyzed through classification and regression trees (CART) by using rprart, rpart.plot, tidyverse, dplyr, ggplot2, and corrplot packages in R. Results of data analysis indicate that MR, PCI, and VSA can predict PA, while SES and DST are not predictors. Findings of the research were discussed along with intelligence theories and practical implications were noted for teachers and researchers.