Surfacing, as well as testing, "elliptical assumptions" in a theory of change: Principled discovery.

While the testing of known assumptions in a theory of change is important, so too is the discovery or surfacing of previously unrecognized assumptions. This paper describes and illustrates the surfacing of "elliptical assumptions," which involve the unknown ingredients that are necessary for a program to be effective. Identifying the ingredients of program success is important for several reasons, which include (a) guiding the development of an improved theory of change, which in turn can guide program improvement, and (b) informing efforts to transfer the program to other settings and populations. However, when an observed pattern, such as differential program effects, points to the existence of a previously unidentified but important ingredient, this might be a just-so story, a seemingly compelling but inaccurate account. Accordingly, the testing of previously unidentified elliptical assumptions is recommended and illustrated.