An Exploratory Validation Study of the Computer Attitude Scale for ESL Students

Abstract

Awareness of test takers' attitudes toward computers is critical in the development of computer-based testing. We need a reliable and valid scale to measure test takers' attitude toward computers. The present study investigates whether the Computer Attitude Scale (CAS) is a valid measure of computer attitude and whether it can be continuously used for ESL students. This study examines factorial validity of the CAS for ESL students. A Confirmatory Factor Analysis (CFA) followed by an Exploratory Factor Analysis (EFA) was conducted using SAS proc calis and SAS proc factor, respectively. The CFA examined the fit of a three-factor solution suggested by Loyd & Loyd (1985). All of fit criteria and the pattern of discrepancies suggest unacceptable fit for the hypothesized three-factor model. The EFA revealed that the three-factor model best represented the data where computer confidence scale was collapsed into computer liking scale. The results of the CFA in conjunction with the EFA suggest that the structure of the CFA is best represented by three underlying constructs such as a combination of computer confidence and liking, anxiety, and usefulness for ESL students. The CAS is generally a good measure of general attitude toward computers for ESL students. However, we need to develop a new scale to adequately assess their computer confidence construct.