“…Random selection and random assignment of participants, although not always feasible, increase the generalizability of findings to the target population, and possibly to other populations. Threats such as interactions of treatment and setting, or of treatment and history, can be addressed methodologically by introducing multisite aspects to the study, by addressing timing of studies such that history varies, or through replication (Fahs, Morgan, & Kalman, 2003; Hayes, 1998). Researchers could acknowledge historical events occurring during the course of the study, and, if possible, address these threats through data collection in a variety of circumstances.…”