This document discusses different methods for conducting impact evaluations to assess the effects of projects, programs, or policies. It describes what impact evaluations are and why they are important. It then explains the challenge of finding an appropriate comparison group to account for selection bias. The document proceeds to discuss specific methods for constructing comparison groups, including difference-in-differences, randomization, instrumental variables, matching, and regression discontinuity design. For each method, it provides details on how the method works and its advantages and limitations.