Adjusting Self-Reported Attitudinal Data for Mischievous Respondents

Abstract

For various reasons, survey participants may submit phoney attitudinal self-reports meant to bypass researcher scepticism. After suggesting reasons for this new category of problematic survey participant - the mischievous respondent (MR) - and reviewing related response bias, faking, inattentive respondent and outlier literatures, an initial algorithm for removing such respondents from polychotomous attitudinal data sets is posited. Applied to four data sets, this algorithm marginally reduced EFA cross loadings and improved CFA model fit. Although purging subtly suspicious cases is not standard practice, the extant literature indicates that such algorithms can reduce artifactual statistical findings substantially.