Exploring a proactive measure of making items of a personality questionnaire resistant to faking: An employee selection setting

Abstract

This study examined resistance of item parameters of a personality questionnaire to faking. We proposed seven construct-irrelevant item attributes (CIIAs), each composing of two opposite values, representing degrees of adherence to the respective CIIAs (e.g., based on the time frame embedded in the stem, items were divided into past and present). We developed a personality inventory that measured conscientiousness. Items of the questionnaire were written to embody all the seven CIIAs. Participants of the study (N = 543) were divided into two groups: honest and faking groups. Those in the honest group were instructed to respond honestly whereas those in the faking groups were asked to respond as if they were job applicants. Multi-group latent variable modeling was used to examine the score differences of the participants' responses to the items of the conscientiousness section of the NEO FFI. It was found that items on two of the CIIAs were resistant to faking. Applying a Wald test of equality of item parameters, we found that the items did not show differential item functioning in either of the attributes.