A longitudinal study of the interaction between gender, computer anxiety, math anxiety, and test anxiety in a college-level computerized testing situation

The three major goals of this research were to investigate gender differences in computer, math, and test anxiety in a computerized testing situation, how the levels of anxieties change following repeated exposures to computerized testing, and to determine which factors best predict the levels of the three anxieties.

Subjects were the students enrolled in Business Statistics during Fall semester, 1990 at Texas Tech University. The scores of three anxiety measurements on the pre and post or intermediate survey were dependent variables, while demographic variables and anxieties were used as independent variables. T-tests, repeated measures ANOVA, and multiple regression analyses were used for statistic analyses at p = 0.05 level.

The results of t-tests showed that no gender differences for computer anxiety, computer confidence, and computer like were found in the pre-survey, while females reported significantly higher levels of math anxiety than males.

Time effect (main effect) was significant for the Computer Attitude Scale in the repeated measures ANOVA. Repeated exposures to computerized testing decreased significantly the levels of computer confidence and compute like in the post-survey for both male and female students, while the levels of computer anxiety increased significantly. Time effect also was significant for math anxiety. The students' levels of math anxiety in the post-survey were significantly higher than those of the pre-survey. Gender and interaction effects between gender and time were not significant for either computer anxiety or math anxiety study.

The results of the repeated measures ANOVA for test anxiety indicated that gender and gender x time interaction effects were not significant. However, significant main effects for time were identified in the analyses. The difficulty levels of exams and repeated exposures to computerized testing in the statistics course may have escalated the levels of test anxiety.

In the regression analyses of computer anxiety, computer related variables and test anxiety were the best predictors. Test anxiety was the most important predictor of math anxiety. Number of math courses taken and computer anxiety were significant predictors for females. Math anxiety and computer anxiety contributed significantly to the regression of test anxiety for both males and females. Additionally, number of math courses taken and age were significant predictors for males.