Conviction Is Not Commission: A Methodological, Conceptual, and Ethical Critique of Khoshnood et al

Khoshnood et al. (2026) report a strong association between immigrant background and rape convictions in Sweden using nationwide register data and a matched case-control design. This commentary argues that the study suggests important descriptive evidence on conviction patterns, but that its interpretation exceeds what the data can support. We believe that the central limitation in the published study is that the authors investigate rape conviction as the dependent variable, not actual rape offending. It must be noted that conviction is the filtered outcome of selective criminal justice processes consisting of a series of events, including reporting, policing, prosecution, evidentiary thresholds, legal representation, interpretation, and judicial discretion, which could influence the outcome variable. Additionally, we also argue that omitted confounding factors, including education, language proficiency, legal literacy, legal status, and access to effective defense, may partly explain the observed associations. Finally, we highlight concerns about overreaching interpretation and biased language use in a socially sensitive area, particularly in light of APA guidance on bias-free language use and inclusive scholarly communication. In conclusion, we invite readers to evaluate the findings of the original study more carefully. Simultaneously, we argue that these findings may also be interpreted as evidence of unequal conviction patterns, rather than as evidence of unequal offending propensities.