Effect of Computer Attitude and Test Anxiety on the Reliability of Computer Based Science Test in Nairobi City County, Kenya

Abstract

Computer based testing is increasingly being adopted in assessment systems the world over because of the many advantages that it has over the traditional paper based test format. In spite of its advantages, there has been concern that this mode of administration should be of good quality. There have also been concerns about the effect of some pupil psychological and personal characteristics on computer based test quality. The indicator of test quality that was considered in this study was reliability. While a number of studies on computer based test reliability have been done in other countries, none are readily available in Kenya. The purpose of this study was to determine the effect of computer attitude and test anxiety on the reliability of a computer based Science test in Nairobi City County. The study also explored the intervening effect of gender. The theoretical framework that guided this study consisted of the Cognitive Load Theory, the Theory of Reasoned Action and the Processing Efficiency Theory. Multi-stage sampling was used to select 424 standard seven pupils in Nairobi City County that participated in this study. A 2x2 factorial between subjects experimental research design was utilized, whereby sampled pupils were assigned at random to the computer based test or the treatment group and the paper based test group which was also the control group. The main tool used in this study was a Science test that was administered via either the computer or the paper based test modes. All sampled pupils also responded to the Computer Attitude Scale and the Westside Test Anxiety Scale. A pilot study was used to refine and calibrate these instruments. Data was analyzed using the independent samples t-test, two-way and three-way analysis of variance. Significance of the results was tested at an alpha value of 0.5. Results of analysis indicated that the computer based Science test was reliable and therefore fulfilled the quality imperative. It was also found that computer attitude did not have an effect on the reliability of the computer based Science test. Regarding test anxiety, it was found that it had a non-significant effect on the reliability of the computer based Science test. In combination, computer attitude and test anxiety did not have an effect on reliability. Gender did not have an effect on the reliability of the computer based Science test. It was however found that there was a non-significant interaction of computer attitude and test anxiety among girls. It was concluded that computer based testing provides reliable test scores and therefore its use in schools as well as in national assessments can be upscaled. Future research can consider samples from other counties and class levels as well as other school subjects. It was also recommended that future research considers the reliability of the computer adaptive test mode.