Open AccessExploring the Predictors of College Readiness for Low Achieving High School Graduates Through Multilevel Modeling
Author(s) -
Bidya Raj Subedi,
Mark Howard
Publication year - 2018
Publication title -
journal of studies in education
Language(s) - English
Resource type - Journals
ISSN - 2162-6952
DOI - 10.5296/jse.v8i4.13859
Subject(s) - multilevel model , english language learner , certification , mathematics education , reading (process) , psychology , hierarchical generalized linear model , random effects model , grade level , student achievement , academic achievement , medical education , mathematics , english language , medicine , generalized linear mixed model , statistics , political science , meta analysis , law
For low achieving (at-risk) high school graduates, this article identified significant student and school level predictors of college readiness in reading and mathematics. This study employed a two-level hierarchical generalized linear model (HGLM) to explore the fixed and random effects. The study included 36 high schools where 3,784 students in reading and 2,903 students in mathematics with achievement levels 1 and 2 in both subjects were selected from one of the largest school districts in the United States. At the student level, grade point average (GPA), exceptional student education (ESE), English language learner (ELL), and Hispanic status of students were significant. At the school level, percentage of teachers with National Board certification, percentage of teacher effectiveness and advance degrees as well as average years of teaching experience were significant in predicting college readiness. The effect sizes, which ranged from .29 to .37, were determined to be small.