Social stress in human-machine systems: opportunities and challenges of an experimental research approach

Abstract This article presents some deliberations on methodological approaches to researching the effects of work-related social stress on performance, with particular consideration being given to machine-induced social stress. The article proposes a broad methodological approach to examine such effects. A particular focus is placed on performance after-effects (e.g. unscheduled probe tasks), extra-role behaviour, and task management behaviour because of conventional performance measures (i.e. scheduled tasks) often being unimpaired by social stressors. The role of the ‘performance protection mode’ as an important concept is discussed. A distinction is made between three facets of after-effects: performance-related, behavioural, and emotional. Unscheduled probe tasks and voluntary tasks are proposed to measure performance-related and behavioural after-effects. Propositions for specific experimental scenarios are made, allowing for sufficiently realistic simulations of social stress at work. The availability of such lab-based simulations of work environments offers good opportunities for this line of experimental research, which is expected to gain in importance since highly automated systems may modify the impact of human-induced social stress or may even represent a social stressor themselves. Finally, the considerations presented in this article are not only of relevance to the domain of social stress but to experimental stress research in general.