Because observational studies rely on data collected in a nonexperimental setting, it is difficult to ensure that the control group forms a valid counterfactual. Even though the estimated regression coefficient provides a measure of the correlation between the independent and dependent variables, one cannot assume a causal relationship because outside factors could affect both of these variables Our regression analysis Consider the previous
hypothetical labor supply example