NCJ Number
58716
Date Published
1976
Length
34 pages
Annotation
THIS ARTICLE EXAMINES TRADITIONAL AND ALTERNATIVE METHODS OF DEVELOPING AND TESTING SOLUTIONS TO SOCIAL PROBLEMS; ITS THRUST IS ON RECOGNIZING BIASES.
Abstract
TRADITIONAL WAYS OF ASSESSING PROGRAMS OR POLICIES MAY NOT ALWAYS BE ACCURATE; NO MEASURE, IN ANY CASE, IS TOTALLY VALID, NOR IS ANY ONE TEST OF A PROGRAM SUFFICIENT. SEVERAL ASSUMPTIONS IN METHODLOGICAL PRACTICE ARE EXAMINED FOR THEIR VALIDITY, INCLUDING THAT RELIABLE AND SOCIALLY USEFUL INFORMATION IS AVAILABLE BY LOOKING AT THE SOCIAL WORLD AS IF IT WERE ORDERED BY CAUSAL RELATIONSHIPS; THAT SOCIAL WORLD DIMENSIONS ARE SEPARABLE BY ANALYSIS; THAT ALLEVIATION OR ELIMINATION OF A SOCIAL PROBLEM IS RIGHT AND BENEFICIAL (WITCHCRAFT AND HOMOSEXUALITY ARE EXAMPLES); AND THAT A MORE EFFECTIVE SOCIAL PROGRAM RESULTS IF FUNDS ARE SPENT ON EVALUATION. TRADITIONAL TECHNIQUES OF EVALUATING PROGRAMS (E.G., THE CLASSICAL EXPERIMENT, QUASI-EXPERIMENT, CORRELATIONAL APPROACH) ARE DIRECTED AT CONVENTIONAL SOCIAL INSTITUTIONS, WITH THE ASSUMPTION THAT ESTABLISHED PROGRAMS MUST BE DOING THE JOB. THIS ARTICLE CAUTIONS THAT AN ANALYST MUST BE AWARE OF THE CAUSES OF ERROR BEFORE ATTRIBUTING THE SOURCES OF SUCCESS OR FAILURE. ORIGINS OF ERROR INCLUDE PROGRAM MATURATION, SELECTION, CONTAMINATION, AND ERRORS IN THE USE OF COLLECTIVITIES. THESE ARE EXPLAINED IN DETAIL WITH EXAMPLES. ALONG WITH DEFINITIONS OF BASIC TERMS SUCH AS METHODOLOGY, PROGRAM, AND EFFECTIVENESS. REFERENCES AND AN ANNOTATED BIBLIOGRAPHY ARE INCLUDED. (RFC)