In experimental studies, researchers are often interested in secondary research questions that explore important aspects of main findings, such as whether or not program effects vary according to the level of fidelity in which the program has been implemented; or according to the dosage received by individuals participated in the program. In this evaluation, stakeholders questioned whether higher fidelity of implementation was related to greater impact of the intervention on student outcomes. This paper addresses this question and compares the methodology and results of three approaches for linking implementation fidelity to impacts. Exhibits are appended.