Can We Invent Better Scientific Approaches to Anticipating Failure to Meet Expectations in Randomized Controlled Trials and to Learn from the Failure?

Preamble Health professionals have a disciplined approach to post mortems and autopsies at the individual level, and sometimes at the hospital unit level so as to learn from medical failures. Engineers have " safety factors " in designing structures and systems so as to avert unpredictable failures and a visible, though not always orderly, tradition of learning from their failures. Books on the topic in each arena are in ample supply. Some are good. The applied social sciences, including education sciences and criminology, enjoy less clarity in understanding failure for a variety of reasons. A bridge collapse, for instance, is often plain and spectacular. The failure of an intervention program in education or of a police intervention aimed at crime control is often quieter. Despite journalistic hyperbole and political theater, the outcome is not spectacular and many reasons for it rarely transparent. More to the point, the applied social sciences lack disciplined , well developed, and transparent approaches to anticipating the failure to meet expectations in testing the effectiveness of programs, in analyzing the failures, and in building a cumulative knowledge base on the phenomenon. We can, for instance, identify " what works " pretty well from randomized controlled trials. But little serious attention has been dedicated to understanding why and how a