Prejudicial essentialist bias in cognitive neuroscience studies with people who use drugs

Professional judgement can be biased (or prejudiced) by the incidental social characteristics of marginalised clients. Perhaps unexpectedly, psychologists have been hypothesised to express prejudicial essentialist bias in their writing; disproportionately invoking pathological constitutional deficits to explain observed differences in study participants with marginalised social characteristics.

To test whether this bias exists in cognitive neuroscience publications, 67 case-controlled studies reporting lower cognitive task performance in people who use drugs (PWUD) versus controls were collated. The percentage of explanations of PWUD's performance using essentialist terms ('deficits', 'dysfunction', 'impairments') versus neutral-agnostic terms ('difference', 'change', 'alteration' etc.) were tested for associations with PWUD's social characteristics, study effect size and publication year.

Deficit descriptions were used on 64% of occasions and were significantly associated with PWUD's social characteristics including their mean age (r = 0.40), percent male (r = 0.24), primary drug (η² = 0.24), and dependence severity indexed by their recruitment from clinical sources (η² = 0.12) and characterisation using clinical diagnostics (η² = 0.15); but not with the effect size of the task performance difference (r = 0.029) or publication year (r = -0.18).

Psychologists' unwarranted causal inference that task performance differences arise from constitutional pathogenic origins may be prejudiced by the incidental social characteristics of their PWUD sample. This bias may be explained by professional incentives, journal remit and author training influencing narrative construction. Solutions for the bias may involve meta-scientific auditing, strengthening of language guidelines, and critical reflective practice implemented through positionality statements in quantitative research.