Initial evidence of reliability and validity of an implicit association test assessing attitudes toward individuals who use substances

ABSTRACT

Background: Substance use disorders (SUDs) are stigmatized conditions, with individual biases driving poor health outcomes. There are surprisingly few validated measures of bias or stigma toward individuals who use substances. Bias can be classified as explicit (self-report) or implicit (behaviorally based).

Objectives: The goal of the present study was to establish preliminary indices of reliability and validity of an implicit association test (IAT) designed to measure implicit bias toward individuals who use substances.

Methods: A large United States-based, crowd-sourced sample (n = 394, 51.5% male, 45.4% female, 2.5% nonbinary) completed the IAT and a small battery of survey instruments that assessed social distance to mental illness (including heroin use), attitude toward and perceived controllability of injection drug use, perception of public stigma, and social desirability.

Results: Nearly all (92%; n = 363) scores on the IAT indicated greater negative than positive attitudes toward those who use substances. Spearman-Brown corrected split-half reliability on the IAT scores was excellent, r = .953. Controlling for social desirability, IAT scores positively correlated with all included measures pertaining to substance use as well as social distance for heroin and schizophrenia (but not diabetes). A principal component analysis resulted in two interpretable components representing disapproval (perceived controllability and negative attitudes) and perceived stigma (social stigma and social distance). Scores on the IAT positively correlated to scores on both components, again, controlling for social desirability.

Conclusion: These results provide compelling preliminary evidence of validity of an IAT designed to measure bias toward individuals who use substances.

KEYWORDS:

• Stigma
• implicit bias
• explicit bias
• substance use disorder
• addiction

Acknowledgments

The authors would like to thank Nicole Frost, Amy Johnson, Lauren Simon, Jack White, and Dr. Colin Smith of Project Implicit.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was partially supported by an Alliance for Academic Internal Medicine Innovation Grant 2020.