The article presents a discussion on the difference-in-differences (DiD) estimator. Also, it briefly describes the DiD model and its underlying assumptions. The DiD estimator is a useful tool for applied quantitative education and public policy researchers confronted with observational data in which self-selection to treatment may be confounded with the outcome measure. The model is extremely flexible, and allows for the inclusion of additional covariates that are hypothesized to influence either the baseline change common to all units of observation or the amount of change predicted by the treatment. Moreover, the models are simple to estimate with off-the-shelf technology and are reasonably easy to interpret.