A cross-sectional study was performed to assess changes in computerized neurobehavioral performance across multiple test sessions for developing a valid performance level and reducing test time.

The study was conducted on 118 male car painters exposed to low-dose, mixed organic solvents. The study controls were 113 unexposed reference workers matched for age and educational year from different sections of the same factory. Each worker completed a medical and occupational exam and questionnaire. Three tests of the Swedish Performance Evaluation System (5 minutes of Simple Reaction Time, 9 sessions of Symbol Digit, and 4 sessions of Finger Tapping Speed) were administered to each group.

The mean age, mean duration of employment, and mean years of education were 33 years, 6.7 years, and 12 years in both groups, respectively. The mean reaction times of the Simple Reaction Time increased with the lapse of time. The known group difference between the exposed and control groups was larger for the first minute than for the successive 3 minutes. Both groups showed improved performance across the first to the seventh sessions on Symbol Digit. Following the seventh session, the performance levels of Symbol Digit had worsened. The t statistic of the two groups was the largest when the two mean reaction times were calculated from the sessions of the fifth to the seventh. In the dominant and non-dominant hands, the number of taps increased and decreased, respectively, across the sessions on Finger Tapping Speed. The known group difference was the largest when comparing the performances from the first to the second sessions.

For valid interpretation and timesaving in computerized neurobehavioral tests, the following is recommended: one practice and one test session for Simple Reaction Time, four practice and three test sessions for Symbol Digit, and one practice and two test sessions for Finger Tapping Speed.