Do People Answer Honestly When Asked about a No Honest Behavior?: Testing the Social Desirability Bias When Measuring Bullshitting

Recently assessment of bullshitting has gained considerable interest. However, self-report measures of bullshitting are at risk of socially desirability (SD) responding because people are questioned about a sensitive and personal topic (i.e., their tendency to embellish or modify the truth to portray themselves positively). This study aims to identify the impact of SD on the Bullshitting Frequency Scale (BSF) to ensure that the scale can provide sound measures. Italian (N = 298, 69.80% female) and English participants (N = 301, 72.76% female) were recruited. To overcome some limitations of the classical procedures used to deal with SD, factor analysis was employed to identify and control this response bias. At the same time, the validation of the Italian version of the BSF (BSF-I) was conducted to provide evidence that the scale maintains its psychometric properties once translated into Italian. Factor analysis confirmed the two-factor structure of the scale and factor loadings indicated a minimal impact of SD. Multidimensional IRT showed moderate to excellent MDISC values and strong conditional reliability. Multigroup CFA indicated strict invariance across genders and metric invariance across Italian and English versions. The BSF-I showed sound psychometric properties and these findings suggest that people answer honestly when asked about bullshitting.