Electronic Theses and Dissertations


Open Access Dissertation


The American College Test (ACT) is a nationally normed test developed by the American College Testing program in Iowa City, Iowa. The test battery assesses the aptitude of high school students and is especially important for those students who wish to attend a post-secondary educational institution. The ACT assessment is a comprehensive evaluation, guidance, and placement service for educators and students in the transition from high school to college. This study examined the relationship between ACT test scores and certain selected student, school and family variables. The sample for this study was the 1986-1987 ACT test taking population of senior public high school students in the state of Iowa. The data tapes of 21,131 public high school students were provided by the American College Testing program in Iowa City, Iowa. The data tape provided information about each public high school student in regard to his/her: (a) ACT composite test score; (b) academic preparation; (c) graduating class size; (d) school district's cost per pupil; (e) ethnicity; (f) gender; (g) class rank; (h) grade point average; and (i) family income level. Descriptive statistics and frequency distributions portrayed the sample. A multiple regression analysis was used to test the relationship between the independent and the dependent variables. The concept of effect size was used to interpret the size of the correlation coefficients. The conventional definitions proposed by Cohen were used: "small" effect size, r =.10 and r$\sp2$ =.01; "medium" effect size, r =.3 and r$\sp2$ =.09; and "large" effect size, r =.5 and r$\sp2$ =.25. The analysis of data of this study revealed three variables that were classified as having a "large" effect size. Those three variables were: (a) high school grade point average; (b) high school class rank; and (c) ethnicity. Two variables, academic preparation and gender, produced a "medium" effect size. Family income level and graduating class size were calculated as having a "small" effect size and a school district's cost per pupil was noted as having less than a "small" effect size. High school grade point average, high school class rank, and ethnicity of Iowa public high school students had the greatest relationship to ACT test scores. Family income level, graduating class size and a school district's cost per pupil were found to have very little relationship to ACT test scores of Iowa public high school students.

Year of Submission


Degree Name

Doctor of Education


Department of Educational Administration and Counseling

First Advisor

Ann Vernon

Date Original


Object Description

1 PDF file (viii, 98 pages)



File Format