The Differential Impacts of Two Forms of Insufficient Effort Responding

Recent years have seen a renewed interest in insufficient effort responding (IER). Previous research has demonstrated that IER can have detrimental effects on survey research ranging from introducing untrustworthy data to influencing psychometric and statistical results. The present simulations examine two forms of IER, straightlining (SL) and random responding (RR), in an attempt to determine whether the presence of these response patterns have differential impacts on data. In three studies, we explore the combined effects of extreme SL and RR, the effects of full and partial RR, and the effects of full and partial SL on scale characteristics such as inter-item correlations, alpha, and component structure. We also explore how various IER response distributions may influence these statistics. Empirical results demonstrate a tendency for SL to increase and RR to decrease the magnitude of inter-item correlations, alpha, and the first component eigenvalue. Results also indicate that the impact of SL may be more pronounced than the impact of RR in the organisational sciences. It is important for researchers to consider the type of IER in addition to the prevalence of IER in a sample prior to conducting statistical analyses.