The objective of this study was to assess the degree to which the mode of administration of the computer-based Academic Skills Assessments of the "Praxis Series: Professional Assessments for Beginning Teachers" contributes to performance differences among test takers. Inexperienced or anxious computer users (Praxis pilot sample of 446 college students and graduates and study sample of 145) were recruited to take the assessments. The degree to which test design and test familiarization procedures effectively minimized variation due to comfort and familiarity with computers was examined from three perspectives: (1) the extent to which the availability of a personal, information-providing test center supervisor influenced test performances, beyond the help provided by a computerized test familiarization tutorial; (2) the effect of within-test practice on later performance on a subsequent section of the test; and (3) the relationship of computer-based test performance to attitudes toward computers and experience in using them. The conclusion was that performance on the tests is not unduly affected by computer administration. Eleven tables present study results. Two appendixes, with five more tables, present pilot and study samples' responses to computer attitude scale items and supplemental information. (Contains 38 references.) (Author/SLD)