Almost half the federal government's expenditures are for "human resources including social action pro such as Head Start, delinquency prevention, drug rehabilitation, job training, and many others. Do they work? Or is the money wasted? May they even do more harm than good? The use of scientific research procedures to measure the effectiveness of an action program is called evaluation research [Suchmap, 1967; Abt, 1977; Cook, 1978]. Evaluation research may use any of the kinds of studies described in the preceding pages. Its object is to replace guesswork with knowledge in deciding what programs to continue and how to improve them. (At least in theory, this is the purpose; in practice, the purpose of evaluation research may be to develop proof of the program's "success" so that funding will be continued.) Evaluation research is not easy, for many variables must be controlled. Often the findings of various evaluation studies are so conflicting that no firm conclusions can be drawn. For example, Nancy St. John (1975] reviewed dozens of studies of the effects of school desegregation on pupil learning and found that the reported effects varied so widely that no clear decision could be rendered. Even when numerous studies do agree they may bet disbelieved or ignored. Studies critical of an agency may be quietly buried, and those whose conclusions conflict with popular beliefs are disregarded. liar example, a number of studies have shown that high school driver-training courses have little or no effect upon driver accident rates ( Monomania 1968; Harmon, 1969; Conley and Smiley, 1976], but it appears that people have made a common-sense judgment that driver training "must work"and simply ignore evidence that it does not.
Despite the difficulties and pitfalls, evaluation research is one of the most important and most rapidly growing areas of sociological research, with new books appearing every year (Cuba and Lincoln, 1981; ever, 1981 Crane, 1982], along with an Evaluation Studies Review Annual and a quarterly journal, Evaluate Review While evaluation research is imperfect, the alternative is to rely on hunches and guesswork in designing social action programs.