Random and Systematic Error Effects of Insomnia on Survey Behavior

Abstract

Insomnia is a prevalent experience among employees and survey respondents. Drawing from research on sleep and self-regulation, we examine both random (survey errors) and systematic (social desirability) effects of research participant insomnia on survey responses. With respect to random effects, we find that insomnia leads to increased survey errors, and that this effect is mediated by a lack of self-control and a lack of effort. However, insomnia also has a positive systematic effect, leading to lower levels of social desirability. This effect is also mediated by self-control depletion and a lack of effort. In supplemental analyses, we find that psychometric side effects of random and systematic error introduced by individuals high in insomnia negatively affect internal consistency estimates and measurement invariance on various organizational measures. Results were replicated across two studies, with alternative operationalizations of survey errors and social desirability and some alternative explanations examined. These findings suggest sleep may be a key methodological issue for conducting survey research. Recommendations from the sleep and self-regulation literature regarding potential strategies for counteracting the effect of insomnia on survey responses are discussed.