L’impact des minorités bruyantes de chercheurs sur l’état du savoir scientifique. L’exemple de la racial invariance thesis dans la criminologie états-unienne

Abstract

This paper investigates the influence of the research produced by a « vocal minority » of scientists in a corpus of 48 articles regarding the differential effect of structural factors on racially disaggregated crime rates in the US. The computational social science analysis underlines the limited influence of methodological choices on the scientists' conclusions.However, we show that a particularly productive research team distorts the state of the art by systematically concluding that the racial invariance thesis is refuted, even though their raw quantitative findings differ from one study to another. In the absence of constraining scientific standards, this vocal minority is free to adopt interpretations that are consistent with its prior theoretical interests.