Using Hierarchy Linear Modeling to Examine Factors Predicting Students' Reading Achievement by

Using Hierarchy Linear Modeling (HLM), this study identified factors, including students’ language background and exposure for practicing English language, which predict students’ reading achievement. Using data from the 2007 administration of the Pan-Canadian Assessment Program (PCAP) and its accompanying surveys for students and the schools, a twolevel (student level and school level) HLM model was analyzed for predictive relationships. Results showed that at the student level, predictors such as students' participation in class discussions, language spoken at home, parents' encouragement to read at young age, and the amount of individual projects requiring students to work outside of class contributed significantly to the students' reading scores. However, none of the school level predictors were found to be significant. All the significant predictors contributed to only 12% of the variability in this HLM model. Identification of more significant variables is needed in order for a full picture to be seen. This research shed lights for educators regarding how students' language history, and their amount of exposure to practice English contribute to predicting students’ reading achievement, thus helping English learning students and students of poor reading skills.