Does Adjusting for Social Desirability Reduce Ceiling Effects and Increase Variation of Patient-Reported Experience Measures?

Abstract

Social desirability bias (a tendency to underreport undesirable attitudes and behaviors) may account, in part, for the notable ceiling effects and limited variability of patient-reported experience measures (PREMs) such as satisfaction, communication effectiveness, and perceived empathy. Given that there is always room for improvement for both clinicians and the care environment, ceiling effects can hinder improvement efforts. This study tested whether weighting of satisfaction scales according to the extent of social desirability can create a more normal distribution of scores and less ceiling effect. In a cross-sectional study 118 English-speaking adults seeking musculoskeletal specialty care completed 2 measures of satisfaction with care (one iterative scale and one 11-point ordinal scale), a measure of social desirability, and basic demographics. Normality of satisfaction scores was assessed using Shapiro-Wilk tests. After weighting for social desirability, scores on the iterative satisfaction scale had a more normal distribution while scores on the 11-point ordinal satisfaction scale did not. The ceiling effects in satisfaction decreased from 47% (n = 56) to 2.5% (n = 3) for the iterative scale, and from 81% (n = 95) to 2.5% (n = 3) for the ordinal scale. There were no differences in mean satisfaction when the social desirability was measured prior to completion of the satisfaction surveys compared to after. The observation that adjustment for levels of social desirability bias can reduce ceiling effects suggests that accounting for personal factors could help us develop PREMs with greater variability in scores, which may prove useful for quality improvement efforts.