Which variables should I control for in regressions analysis?

Control variables are often used when randomization is not feasible and natural experiments are not available or hard to find. As such, “non‐experimental designs are sometimes presented as the only feasible way to conduct research in social science” (Narita et al., 2023, p. 1). Thus, whereas there are often better strategies for causal identification—as exemplified by the increase of articles using experiments, natural experiments, and instrumental variables in the journal Leadership Quarterly – control variables remain a fallback option for many researchers.

Link to the paper 'The choice of control variables in empirical management research: How causal diagrams can inform the decision'

No registration is needed for this seminar.

Organiser

Robert Klemmensen, Professor at the Department of Political Science 

Welcome!

Arranged by: Lund Social Science Methods Centre