THE EFFECTS OF COMPUTER ANXIETY AND PAST COMPUTER EXPERIENCE ON COMPUTERIZED TEST-PERFORMANCE

This study investigated the effects of computer anxiety and past compu ter experience on the computerized intelligence test performance of co llege students. It also explored the relation between computer anxiety and Stare and Trait anxiety. 406 undergraduate psychology students we re administered a Computer Anxiety Scale (CAS) developed for this stud y prior to being assigned to one of three level anxiety groups: High a nxious (N = 40), Low anxious (N = 42) and Medium (N = 79). These 161 s ubjects then completed a Questionnaire which determined their Computer Experience, the State-Trait Anxiety Inventory (STA1; Spielberger, Gor such, & Lushene, 1970), the B.53 (Bonnardel, 1977), paper and pencil i ntelligence test end a computerized version of the Advanced Progressiv e Matrices (PMA; Raven, 1990). Factor analysis is used to identify fac tors of computer anxiety underlying the GAS. In general the CAS is sai d to be a valid instrument. High correlation is found between CAS and past computer experience (r = -.59) and between CAS and Trait anxiety (r = .46). Results indicated no differences in test performance betwee n computer anxiety groups, either when tested on a paper and pencil te st or on a computerized test. They also indicated a significant differ ence in State anxiety level compared to those with high computer anxie ty level when tested with computer. Further analyses revealed that the differences between State anxiety after the B.53 and the PMA were sig nificantly greater in the high computer anxiety versus low computer an xiety group This suggests that computer anxiety result in an increase in negative feelings during computer testing, but it does not affect t est performance. It may also provide some insight into the reasons for mixed results in past research investigating the effects of test admi nistrations on State anxiety. Specifically, the different samples used throughout the various studies may have varied in terms of computer a nxiety. Tie results also indicated that past computer experience was a significant factor affecting a computerized rest performance. Compute rized testing discriminated against those who had worked with computer s prior to testing. Further analyses revealed significant differences in test duration. These findings suggested that there was a strong ten dency far individuals with high past computer experience to complete c omputerized tests faster. Finally, considerations regarding the admini stration of computerized tests are presented along with an assessment of future research directions on the effects of computer anxiety and o f past computer experience on tea performance.