The Effect of Feedback on Subjective and Objective Measures of Workload and Performance

Abstract

Thirty subjects were employed in a mixed experimental design that examined five levels of feedback and two levels of difficulty for two tasks, with repeated measures on the difficulty and task variables. The amount and type of feedback was varied so that it provided information about performance of the task on the objective measures within a block of trials, or provided the same information at the end of the task (simply providing knowledge of results), and was also varied in quality, either as a comparison to the subject's own average performance of the task, or in comparison to an experimentally determined ‘figure of merit’. Two tasks, each with two levels of difficulty, were used: (1) a task that primarily imposed cognitive demands, a version of the Sternberg memory task, and (2) a task that primarily imposed psychomotor demands, a target acquisition task modelled on the Fitts' Law paradigm. Both objective and subjective measures demonstrated reliable and predictable effects for the difficulty levels of the two tasks, however the tasks were differentially affected by the feedback conditions, but differences between and within tasks were generally small. The relationships between objective measures, and subjective ratings of workload and performance rarely reached significant levels.