Removing Stigmatizing Language in Self-Reports: Effects on Psychometric Properties and Respondent Beliefs

Abstract

Stigma facilitates negative outcomes for those with substance use disorders (SUDs), such as restricted healthcare access, job loss, and social ostracization. Many professionals advocate for reducing stigmatizing language in their conduct. However, there remains a notable degree of stigmatizing language in self-report measures. Thus, I aimed to examine psychometric changes from replacing stigmatizing language in self-report measures and test if these changes affect the respondents’ stigmatizing beliefs. I conducted a randomly assigned double-masked experiment with undergraduates (n = 125; M_age = 19.18, SD = 1.05; 62% female; 58% White), Reddit users (n = 224; M_age = 31.28, SD = 9.21; 58% male; 82% White), and Buddhist Recovery Network members (n = 136; M_age = 50.48, SD = 11.50; 61% male; 94% White). The participants completed either the original version of the Addictions Belief Inventory consisting of stigmatizing terms (e.g., drug abuser) or a modified version consisting of non-stigmatizing terms (e.g., those with SUDs). I found no changes to the psychometric properties when I compared the factor structures, distributions, and construct validities. There was a minor reduction in one subscale’s internal consistency and a large change in the mean scores, which varied by sample type. Removing stigmatizing language did not change the measure’s psychometric properties notably. However, I found that removing stigmatizing language reduces stigmatizing beliefs and that this reduction was substantial, especially for those who have no history of SUDs, such as some undergraduates or Reddit users. These findings support an ongoing effort to reduce stigmatizing language among professionals and academics.