Theses and Dissertations

Issuing Body

Mississippi State University


Yu, Chien

Committee Member

Mathews, Jerry

Committee Member

Prince, Debra

Committee Member

Davis, James

Date of Degree


Document Type

Dissertation - Open Access


Instructional Systems and Workforce Development

Degree Name

Doctor of Philosophy


College of Education


Department of Instructional Systems, Leadership and Workforce Development


The purpose of this study was to examine the effect of selected student profile variables in the 2005 American College Test (ACT) on academic performance of Mississippi high school graduates as measured by the subscales in the ACT. The sample consisted of 16,779 high school graduates that completed the ACT in Mississippi in 2005. The variables that were studied were the ACT subscores: English, mathematics, reading, and science reasoning subscales; race/ethnicity, size of graduating high school class, high school curriculum, and gender, which were completed on the student profile section when the student registered for the ACT. Analysis of variance (ANOVA) and multiple regression analysis at the .05 alpha level, and Means were used to test the statistical significance of the effect of each variable on the ACT subscores. The findings resulted in a statistically significant difference occurring in all four predictor variables: race/ethnicity, size of graduating high school class, high school curriculum, and gender; except no statistically significant difference in gender in the ACT subscore of English and reading. Only two variables explained 40% of the variance of the English ACT subscore: race and ethnicity, and class size. Three variables explained 52% of the variance of the math ACT subscore: (a) race and ethnicity, (b) class size, and (c) gender. All four variables explained 36 % of the variance of the science reasoning ACT subscore. The conclusions drawn from this study were that there were statistically significant differences in the ACT subscore means among the different racial and ethnic groups of students. The mean scores were the lowest for the race and ethnicity variable in the African American background. The lowest ACT subscore mean was in the 399 or below class size, and the highest ACT subscore mean was in the 900 or more class size. The college preparatory program of study scored higher mean scores than the lower ranking other or general and business/vocational program of study, respectively. Overall, males scored higher on the ACT subscores than the females with females outranking males in English. These were found to be probable predictors of success on the ACT.