Computer-Mediated Task Performance Under Stress and Non-stress Conditions: Emphasis on Physiological Approaches

Abstract

AbstractAlthough there is an influx of studies about the relationship between stress and performance, those of the performance in doing computer-mediated tasks under real-time stress conditions are scarce. In this research, we examined the performance of playing a computer game as is considered to share common features with computer-mediated tasks under manipulated job-related stress conditions. To measure computer-mediated task performance, we examined subjects playing the Minesweeper game and their performances in the experiments after we divided the subjects into two groups: the stress-manipulated group and non-stress group. The stresses are manipulated by an assistant who gives challenging comments to the subjects. Adapting physiological approaches, we measure subjects’ galvanic skin response (GSR) and electrocardiogram (ECG) to measure the level of stress during the performance of the task. Additionally, a questionnaire is given to the subjects in order to ask the level of their perceived stress. Then, we investigated how much stress each group is taken by recording physiological signals and the surveyed perceived stress scale (PSS). The results show insignificant differences between the controlled group (non-stress) and manipulated group (stress) in performing the computer-mediated tasks. The findings underscore that there is a better effective physiological signal to measure the level of stress. In addition, to understand the relationship between tasks performance and stress condition, more mediating or influencing factors on the task performance should be explored. For future studies, subjects’ personalities and their skill and knowledge also will be considered to find not only direct casualties but also combined of mediated influencing variables.KeywordsHeart Rate VariabilityTask PerformancePhysiological SignalSkin ConductancePerformance FeedbackThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.