This investigation compares the effects of test mode, gender and race on paper-based versus computer-based delivery of two high stakes multiple-choice course examinations, Midterm and Final. Computer-literate students in upper-level business courses (n = 144) were randomly assigned to receive both tests either on paper or on computer. There were no significant gender effects, though males scored slightly higher than females on both tests. However, participants who received the tests on paper significantly outscored those who received the tests on computers, but this difference occurred only on the Midterm examination. Most striking, non-white females receiving the computer-based test mode scored lowest on the Midterm examination but then scored highest on the Final; all other groups maintained their relative positions from Midterm to Final. It was concluded that test mode familiarity does impact test performance. The results suggest that even computer-literate students in advanced Information Systems classes should practice using mock computer exams before taking high stakes computer-based tests, and that test mode familiarity affected non-white females most.
Wallace, Patricia and Clariana, Roy B.
"Test Mode Familiarity and Performance - Gender and Race Comparisons of Test Scores among Computer-Literate Students in Advanced Information Systems Courses,"
Journal of Information Systems Education: Vol. 16
Available at: https://aisel.aisnet.org/jise/vol16/iss2/7