Search results

In a report of research on extreme response style (ERS) in rating scale responses, it is proposed that, when ERS is defined as a proportion of extreme responses, an ERS measure will be more accurate if the items are uncorrelated & have equal extreme response proportions. Further, appropriate stochastic models should be used to assess the internal reliability & convergent validity of these measures. An ERS measure is created & validated with this method, using data from 1975 & 1987 surveys of large samples of US adults serving on a consumer panel. Findings show that ERS is stable over a lengthy survey compared to a benchmark stability for a "perfect" measure. Further, the distribution of ERS over this population is stable over time. Respondents' ERS is related to their age, education level, & household income, but not to their gender. 4 Tables, 1 Appendix, 46 References. Adapted from the source document.