Evidence-based policy making and evaluating programmatic interventions require asking whether a change in the outcome of interest can be causally attributed to the key explanatory variable or intervention. However, many interventions or explanatory variables whose effect researchers are trying to assess are not assigned randomly. Strong correlation or statistically significant coefficients in regression do not imply causation. Researchers should also be able to address the possibility of reverse causality or selection bias in their results. This workshop will cover different research methods that allow researchers to make causal inferences in the analysis of cross-sectional or longitudinal data, even in non-experimental settings.

Scan the QR code in the poster below to register.