The quality of data collected online: An investigation of careless responding in a crowdsourced sample

Abstract

Despite recent concerns about data quality, various academic fields rely increasingly on crowdsourced samples. Thus, the goal of this study was to systematically assess carelessness in a crowdsourced sample (N = 394) by applying various measures and detection methods. A Latent Profile Analysis revealed that 45.9% of the participants showed some form of careless behavior. Excluding these participants increased the effect size in an experiment included in the survey. Based on our findings, several recommendations of easy to apply measures for assessing data quality are given.