Computerized versus traditional administration of the multidimensional aptitude battery-verbal scale: An examination of reliability and validity

摘要

This study compared the effects of a computerized administration procedure on the reliability and validity of the Verbal scale of the Multidimensional Aptitude Battery (MAB-V), and objective item measure of intelligence, to the standard paper-pencil administration procedure. Using a counterbalanced repeated-measures design, the MAB-V was administered to 80 undergraduates, along with a pretest anxiety measure and a posttest evaluation of attitudes toward the assessment format. Results indicated no significant differences in mean MAB-V subtest and Verbal IQ scores between the two modes of administration. Mean Verbal IQs obtained on the second testing were significantly higher than initial scores across all groups. However, retest IQs fell within the 95% confidence interval for predicted retest scores in 91% of the cases, reducing the psychometric significance of this result. The MAB-V demonstrated good test-retest reliability and subtest internal consistency across both administration modes. No differences in anxiety were obtained across administration format or time, but participants indicated some preference for the computerized procedure, rating it less difficult, more useful, and as commanding greater attentiveness. Findings indicate computer-specific factors have no appreciable effect on MAB-V scores and that normative and validity data available for the standard MAB-V are applicable to the computerized version when administered to population samples of average or higher intellectual ability.