Careless Responding: Rates and Reactions in a Quota Sample and a Voluntary Opt-in Sample

In recent years, internet based surveys have become a popular survey mode in social science. However, this survey mode is prone to careless responding (CR), the tendency to respond to a survey item by ignoring the item’s content. CR might lead to incorrect factor structures as well as to attenuate associations between survey items such that type II errors may occur. Therefore, several detection methods for CR have been proposed and examined. However, the issue of CR has mainly been discussed among psychologists concerned about serious threats to data quality, when questionnaires are internet based and when the sample consists of college students who participate in exchange for a course credit or for remuneration. In this paper, we examine rates of CR based on two different samples utilizing different detection methods. First, we use a quota sample (with remuneration), a central data source for internet based surveys in sociological and political research. Second, we include a voluntary opt-in panel (without remuneration). The experimental setting in both samples allows us examining respondents’ reactions to the detection methods.