The relationships among test-takers' variables and English reading comprehension ability of Thai university students using a computer-based test

Abstract

This study aimed to investigate (1) the relationships among three selected variables, namely computer attitudes, computer anxiety, and computer familiarity, (2) the relationships between those three variables and reading comprehension computer-based test performance, and (3) the extent to which the three variables individually or in combination predict the reading comprehension computer-based test scores. Two research instruments were used in this study. A Computer Attitudes, Familiarity, and Anxiety Rating Scale (CAFARS) was used to collect data on the three variables while a Reading Comprehension Computer-Based Test (RC-CBT) was used to evaluate students' performance. Both instruments were administered to 90 students at Dhurakij Pundit University. These students were randomly selected and divided equally into high, average, and low English language ability groups according to their average achievement scores in Foundation English courses. Pearson correlation coefficients were calculated to estimate the strength and the significance of the relationships among the three variables and the RC-CBT scores. Multiple regression analysis was conducted to predict the influences of the three independent variables on the RC-CBT performance. The results of the analyses revealed that : (1) there are significant relationships among the three variables of students of high, low, and combined language ability groups at .05 level, (2) computer attitudes and computer anxiety werefound to be significantly correlated with RC-CBT scores of students with all levels of language ability at .05 level while computer familiarity was significantly correlated with RC-CBT scores of students of average, low, and combined language ability groups at .05 level, and (3) computer familiarity was found to be the only significant predictor of RC-CBT scores of students of the average and combined language ability groups with R²of .384 (p < .05) and R² of .190 (p < .05) respectively while computer attitudes were found to be a significant predictor of RC-CBT scores of students with high language ability with R² of .409 (p < .05). The findings also provided further evidence that the use of computer-based tests was much preferred by the sample.