Abstract
As evidence about a new program or curriculum accumulates from many studies, a time comes to synthesize findings. What does all the research show? Is a program effective?Does it work broadly, or only with certain people in special settings? Literature reviews are designed to answer such questions, but all too often the effort is unsystematic. In this essay, David Pillemer and Richard Light suggest that these reviews are a major scientific challenge. They discuss several ways of extracting information from a group of related studies, and argue that a key ingredient of a strong synthesis is an analyst's attention to detailed features of each study: who participated, in what setting,with what exact program or treatment. Knowing such details is a prerequisite to explaining conflicting outcomes, an effort the authors believe makes fullest use of the varied information the studies offer.
