One of my favorite tools of program evaluation is the logic model. First introduced to them through my work at the Covenant Foundation, I learned more about how to construct and evaluate them in a Program Evaluation and Analysis course that I took at NYU Wagner.
The process of constructing a logic model helps one focus the overall program evaluation. For example, the metrics that one chooses to measure depend largely on what one is trying to ascertain. If one wants to know if one’s processes or programs are being implemented as planned, one conducts a process or implementation evaluation and measures resources and activities, and then the outputs they produce. If one wants to know if one’s programs are having the hoped-for impact, though, an impact evaluation should be conducted, and one has to measure outcomes and impacts. Outcomes and impacts are much trickier to measure than outputs, and many institutions measure outputs when they are actually trying to evaluate their impact.
Below is one example of a logic model: