NCJ Number
58074
Date Published
1978
Length
164 pages
Annotation
NOTING AVAILABLE EVALUATION DESIGN OPTIONS, AS WELL AS SUGGESTIONS FOR DESIGN IMPROVEMENTS, THIS VOLUME ONE OF AN EIGHT-PART SERIES ON PROGRAM EVALUATION OUTLINES HOW TO DEVELOP AN EFFECTIVE EVALUATION DESIGN.
Abstract
AIMED AT PROGRAM EVALUATORS, THIS VOLUME, IN ITS INTRODUCTORY SECTION, DESCRIBES THE LOGIC UNDERLYING THE USE OF EVALUATION DESIGN AND ITS IMPORTANCE IN OBTAINING RELEVANT DATA. THE INTRODUCTION ALSO DISCUSSES DESIGN OPTIONS OPEN TO FORMATIVE AND SUMMATIVE EVALUATORS; THE LACK OF ATTENTION TO DESIGN IN FUNDED PROGRAM EVALUATIONS IS UNDERSCORED AND SUGGESTIONS ARE MADE FOR GRADUALLY IMPROVING EVALUATION PRACTICE. THROUGH DIAGRAMS, STEP-BY-STEP DIRECTIONS, FLOW CHARTS, AND EXAMPLES, SUBSEQUENT SECTIONS SHOW HOW SEVERAL DESIGNS CAN BE ACHIEVED AND IMPLEMENTED. THESE OPTIONS INCLUDE THE FOLLOWING: (1) CONTROL GROUP DESIGNS--TRUE CONTROL PRETEST-POSTTEST DESIGNS, TRUE CONTROL POSTTEST ONLY DESIGNS, AND NONEQUIVALENT CONTROL PRETEST-POSTTEST DESIGNS; (2) TIME SERIES DESIGNS--SINGLE GROUP TIMES SERIES AND TIME SERIES WITH A NONEQUIVALENT CONTROL GROUP; (3) BEFORE-AND-AFTER DESIGNS; AND (4) ANALYSIS OF VARIANCE (ANOVA) DESIGNS. WHAT TO DO WHEN EVALUATION DESIGNS GO WRONG AND HOW TO RANDOMIZE ARE ALSO COVERED, ALONG WITH METHODS FOR COLLECTING, ANALYZING, AND PRESENTING DATA FOR EACH DESIGN. FURTHER READINGS ARE SUGGESTED. AN INDEX IS INCLUDED.