A Visual Test Battery: Tale of Two Computers

Concerted efforts are made to make performance tests comparable across media. When comparing computerized test batteries, however, there may be differences in hardware and software that can produce less than optimum psychometric properties. The present study examined a battery of seven tests implemented on two computers, a Gateway 2000 P4D-33 and a 286 Zenith PC. Results showed that there were differences in performance levels as a function of trials which may be attributed to individual differences. The means and standard deviations were more stable for the Zenith than for the Gateway 2000 P4D-33. It was equivocal whether this was a function of computer or methodology. Additionally, there were no differences in cross-correlations across computerized test batteries. This finding implied that the constructs were consistent across both batteries although mean differences will need to be reckoned with. Implications and limitations for computerized testing are discussed.