Impact of selected noncurricular variables on regular education student achievement as measured by the 2001--2002 reading and mathematics PSSA scores

The purpose of this study was to determine if it is possible to predict the math and reading scaled scores of regular education students on the Pennsylvania System of School Assessment (PSSA) exams based on an analysis of selected variables obtained from 899 student survey responses; and to determine the magnitude of explained variance attributed to those predictor variables under the category of the Home Environment.;This study addressed the following two research questions: To what extent can the selected independent variables predict regular education student scores on the PSSA Math and PSSA Reading Grade 8 exams and thus, identify students at-risk of failing to achieve a proficient level? And to what extent does the home environment affect student performance?;Multivariate correlation analyses were employed to determine the relationship between the independent (student survey responses) and dependent variables (student PSSA Math and Reading scaled scores). This study also used a cross validation procedure to determine if the regression equations could be applied successfully to a new group of students.;An analysis of the data provides evidence that student's perceptions of their ability to plan and carry out schoolwork was the best single predictor of regular education students' performance on the Grade 8 PSSA Math and Reading exams. The Home Environment variables accounted for the largest percentage of the explained variance in both predictive equation models (50% in math and 40% in reading). The most significant Home Environment variables in both models were: the highest level of formal education of the male parent/guardian, student absenteeism and its effect on learning, and parent expectation of their child in class. Those significant variables found only in the math prediction model were number of books in the home and the transience of the students. The reading prediction model included the number of hours a student spends reading for pleasure. The coefficients for the cross validation were also significant.;This study adds to the body of research related to student performance on the PSSA Math and PSSA Reading exams.