Understanding response styles in self-report data: consequences, remedies and sources

Self-report data are now ubiquitously collected in all kinds of research as well as practical occasions worldwide. Important research, politics or commercial decisions are taken daily on the basis of such evidence. However, the use of self-report does not come without problems. Assumptions that respondents use and interpret the given response categories in the same way (comparability assumption) and give unbiased, and honest responses are not always held. One of the main reasons for this are response styles, defined as “differential use of the response options independent of the items’ content” (Wetzel, Böhnke & Brown, 2016, https://kar.kent.ac.uk/49093/)“ being one of the many possible response biases (“systematic tendency to respond to a range of questionnaire items on some basis other than the specific item content” (Paulhus, 1991, doi: 10.1016/B978-0-12-590241-0.50006-X)).

Despite growing amount of research evidence, the consequences, remedies and sources of response styles are still elusive. Researchers still do not know, e.g. what are the consequences of response styles presence for data quality? What methods are best to identify and control for response styles in analytical models? Why response styles are even present? What are their individual and contextual covariates?

This project aims to answer these and similar questions. To this end we will employ simulation studies, use secondary data from large, international databases, apply novel data analytic techniques, e.g. machine learning tools. We also plan a series of experimental studies that will hopefully broaden understanding of response styles predictors such as personality, motivation or cognitive abilities. In this respect our research programme responds to the call of more comparative, validation and experimental studies voiced by researchers in the response biases field.

The project started in October 2020 and will last until the end of year 2023. It is financed by the National Science Centre (NCN) research grant (2019/33/B/HS6/00937)