Control variable selection in applied quantitative sociology: a critical review.

The article critically reviews 95 quantitative research papers from the European Sociological Review (ESR) published in 2016 and 2017 that interpret statistical associations adjusted for control variables. It finds that 61% of these papers do not clearly define their parameter of interest, 75% fail to provide sufficient theoretical justification for their selection of control variables, and 59% incorrectly interpret coefficients of control variables. The authors emphasize that proper selection and justification of control variables, grounded in explicit causal models—such as those represented by directed acyclic graphs (DAGs)—are essential for meaningful interpretation of adjusted associations, particularly when estimating causal effects. The review highlights that these methodological shortcomings persist even in a leading journal known for methodological rigor, suggesting broader implications for sociological research relying on observational data. The authors recommend greater transparency in defining research parameters and causal assumptions to improve research design and interpretation.