It is difficult to carry and set up a large numbers of computers when computerized neurobehavioral tests are performed to many subjects. There are difficulties in mobilizing and resetting desktop computers and so, the use of desktop computers will decrease due to the recent increasing popularity of laptop computers. However, it is essential that there is no difference between the tests performed on a laptop and on a desktop computer. This study was conducted to estimate the effect that the type of computer has on computerized neurobehavioral performance tests.

Three types of computers were applied. Type 1 was a normal laptop computer. Type 2 was a laptop computer that was remodeled and equipped with a modified keyboard. A common desktop computer with a remodeled keyboard as type 3. The computerized tests included Simple Reaction Time, Choice Reaction Time, Symbol Digit, and Finger Tapping Speed. A total of 60 subjects were examined by the identical tester in the same testing conditions. The computers were randomly allocated, which were used in the test and let examinees take tests with the separate computers.

There were no significant differences among the computers for the results of the Simple Reaction Time and Finger Tapping Speed tests. The mean reaction time, error number of Choice Reaction Time and mean reaction time, correct mean reaction time, error number of Symbol Digit produced significantly different results among the three types of computers.

This study indicates that with only a laptop computer the Simple Reaction Time test using the space bar was useful. But with the laptop computer, the Choice Reaction Time test using the four-direction key and the Symbol Digit test using the upper end number key were unsuitable. The keyboard should be developed further for the Choice Reaction Time and Symbol Digit tests. Among the three computers, we recommend the use of the desktop computer with the remodeled keyboard which the examinees are familiar with and the laptop computer with the remodeled keyboard.