The effect of training and method of stimulus presentation on estimates of task-performance times

The present study was designed to determine: 1) if subjects can be trained to estimate task performance times; 2) if training by a paired-comparison paradigm would produce more accurate estimates than training through performing the actual task; and 3) if viewing a video tape of the tasks being performed would provide more accurate estimates than reading the names of the tasks. Subjects were presented with instructions and with seventeen tasks. The seventeen tasks were presented either by video or written stimuli, and the subjects were then required to make estimates of the 50th and 95th percentile performance times. That is, respectively, the time above which 50% and 5% of trained operators would need more time to complete the task correctly. These tasks were selected from a list of one-hundred on which estimates of performance times had been published. They were selected on the basis of familiarity to the general public. Estimates of performance times of five of the tasks had been validated by obtaining actual performance times. These five tasks were embedded within the remaining twelve tasks. Following an introduction to the concept of estimating task performance times, twelve subjects were trained to make estimates by viewing as many pairs of the tasks as time permitted, and indicating which took more time to complete. The other twelve subjects were trained by performing twelve of the tasks, including the five validated tasks on a device which simulated equipment designed to test electronic components. The remaining five tasks could not be performed on the equipment. Finally, all subjects were again presented the seventeen tasks by one of the two methods of stimulus presentation, and were asked to estimate their performance times of the tasks. Estimates of performance times for the five validated tasks comprised the dependent variable. The study utilized eight experimental groups to include all combinations of the two pretraining stimulus presentation methods, two training methods, and two posttraining stimulus presentation methods. (See more in text.)