times, in interacfive computer tasks
Physiological Psychology, University of Wuppertal, Germany
For each computer task and sub-task, a system response time (SRT) of optimal duration can be determined. SRTs being longer or shorter than a once identified optimal duration have been demonstrated to have a negative impact on the users' work motivation and related attitudes ( Thum, Boucsein, Kuhmann and Ray 1995), on error rates, task completion times, and the general level of physiological arousal ( Kohlisch and Kuhmann 1997), and on electrocortical correlates of cognitive processes during computer work ( Schaefer and Kohlisch 1995). Particularly the guideline 'the faster the better' does not seem to be appropriate for setting SRTs in interactive computer tasks. However, software developers usually ignore the negative impact of sub-optimal SRTs concerning the computer user, although all modern computer systems meet the technical requirements for optimizing SRTs. One reason may be, that optimal SRTs for particular tasks cannot be inferred using a rule of thumb. Therefore, it was asked in the present experiment whether optimal SRTs can be determined by psychophysical methods.
A comparison was made among four psychophysical methods for determining optimal SRTs (see Luce and Krumhansl 1988): (1) The method of adjustment, that is, the subject terminates each SRT when having the desired duration. (2) The up-method of thresholds and (3) the down-method of thresholds. Here, the experimenter varies SRT duration upward or downward and the subject signals if it meets the desired duration. (4) A procedure using graphic representations of SRTs being adjusted by the subject. This method can be regarded as a variant of cross-modal matching.