METHODOLOGY OF EVALUATING SOCIAL ACTION PROGRAMS

• NCJ Number: 16899
• Author(s): G G CAIN; R G HOLLISTER
• Date Published: 1969
• Length: 62 pages
• Sponsoring Agency: US Executive Office of the President
• Corporate Author: University of Wisconsin
• Sale Source: National Institute of Justice/
• Language: English
• Country: United States of America

Annotation

PROBLEMS OF EVALUATION DESIGN FOR SOCIAL ACTION PROGRAMS, SPECIFICATION OF OBJECTIVES, USE OF CONTROL GROUPS, REPLICABILITY CRITERION, STATISTICAL AND ECONOMIC CONSIDERATIONS, AND VALIDITY OF RESULTS ARE DISCUSSED.

Abstract

THE REPORT MAINTAINS THAT DATA COLLECTION AND EVALUATION METHODS EXIST WHICH, WHILE PERHAPS NOT SATISFYING THE METHODOLOGICAL PURISTS, CAN PROVIDE EVIDENCE FOR JUDGING THE DEGREE TO WHICH SOCIAL ACTION PROGRAMS HAVE SUCCEEDED OR FAILED. THE THEME DEVELOPED INVOLVES AN EVALUATION PROCEDURE THAT PROVIDES A MODEL SUITABLE FOR STATISTICAL TESTING, THE ESTABLISHMENT OF A WIDE RANGE IN THE VALUES OF THE VARIABLES REPRESENTING THE PROGRAM INPUTS, AND THE JUDICIOUS USE OF CONTROL GROUPS. IT IS EMPHASIZED THAT SOCIAL ACTION PROGRAMS ARE UNIQUELY COMPLEX IN THEIR VARIETY OF INPUTS AND MULTIPLICITY OF OBJECTIVES, SUCH THAT DECISIONS AS TO WHETHER OR NOT TO ABANDON A PROGRAM CANNOT BE EASILY MADE ON THE BASIS OF THE RESULTS OF ANY EVALUATION RESULTS. THE SUGGESTION IS THAT THE OBJECTIVE OF EVALUATIONS SHOULD MOST OFTEN BE THAT OF PROVIDING A BASIS FOR MODIFICATIONS AND INCREASED EFFECTIVENESS IN EXISTING PROGRAMS.