Date of Award

2001

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Education

Abstract

A study was completed to determine what relationships existed among spending per student for instruction, student per counselor ratio, students' cognitive ability, socioeconomic status, and racial homogeneity as compared to college attendance rates for Indiana school districts in 1999. Data was collected from the Indiana Department ofEducation and a general linear model was utilized for data analysis. In simple linear regression analysis, results indicated significant relationships existed among cognitive ability and socioeconomic status when compared to college attendance rates of the districts. However, no significant relationship existed between spending per student, students per counselor, or racial homogeneity as compared to college attendance rates. Multivariate analysis discovered much different findings. In these analyses, three models were selected for the study. Analysis A showed a significant relationship existed between the linear combination of spending per student for instruction, racial homogeneity, cognitive skills, and socioeconomic status as compared to college attendance rates in the Indiana school corporations. Analysis B indicated a similarly significant relationship between students per counselor, racial homogeneity, cognitive skills, and socioeconomic status as compared to college attendance rates in the Indiana school districts. The third multiple regression presented findings showing a significant but weaker relationship existed between college attendance rates and spending per student combined with only cognitive ability and socioeconomic status therein eliminating race from the regression analysis. Conclusions suggest when examining institutional influences upon college attendance rates such as spending per pupil for instruction or students per counselor, one must evaluate these factors within the frame of reference of demographic characteristics of the students. Institutional influences studied in isolation may not provide enough information to researchers to draw appropriate inferences.

Share

COinS