Revisiting itemmetrics: Do psychologists need to watch their language?

Despite a long tradition of studying psychometric properties of self-report questionnaires in psychology, the literature identifying specific linguistic features of questionnaire items is sparse. Moreover, it is unclear whether linguistic features affect all individuals similarly, or interact with individual characteristics. The present study offers a novel methodological contribution whereby a variety of linguistic features, based on the domains typically studied by linguists (i.e., morphology, syntax, semantics), are proposed. To demonstrate how these itemmetrics can be used empirically, we analyzed data from the Center for Epidemiological Studies-Depression (CES-D) scale. Additionally, we probed interactions between sex and English fluency and each of the features to examine whether there were differential effects depending on the individual. Our results suggest that certain features may impact responding and interact with individual characteristics. We argue that our findings necessitate a stronger focus on this area of research.