ABSTRACT
The authors of this chapter, Sebastian Lemire, Christina Christie, and Steffen Bohni Nielsen, examine a paradox in evaluation: the “theory gap” between the compelling case for incorporating social science theories more explicitly into program evaluation theory, and the fact that this is rarely done in practice. Social science theories, which reflect accumulated knowledge in a particular field, are accepted models describing change mechanisms but are necessarily general and abstract. Program theories, on the other hand, contribute to a better understanding of how a program works in a specific setting (or not), but are hard to generalize. By “knitting” relevant social science theories into the overall fabric of program theory, the authors argue that both the robustness (internal validity) of program evaluation, and its replicability in other contexts (external validity), will be enhanced.
Moreover, theory knitting can simplify both the demand for, and supply of, results information. If program theories can build upon widely accepted causal mechanisms drawn from social science theories, it may be sufficient to report on early results that are easier to assess than eventual outcomes and impact, and still give policymakers and budgetary authorities “something solid” to use for accountability and learning purposes, even in the absence of direct measurement.
