METHODOLOGICAL HAZARDS IN CONDUCTING EVALUATION RESEARCH

• NCJ Number: 50522
• Author(s): M SIZEMORE; M ERNST
• Date Published: 1978
• Length: 10 pages
• Sale Source: National Institute of Justice/
• Publication Type: Program/Project Evaluation
• Language: English
• Country: United States of America
• Note: PRESENTED AT THE SOUTHWEST SOCIAL SCIENCE CONVENTION, HOUSTON, TEXAS, APRIL 12-15, 1978

Annotation

SEVERAL PROBLEMATIC AREAS IN EVALUATION RESEARCH ARE DISCUSSED TO PREPARE RESEARCHERS TO DEAL WITH REAL SITUATIONS DURING PROGRAM EVALUATION.

Abstract

HAZARDS WHICH MAY IMPEDE THE PROGRESS OF THE EVALUATION EFFORT AND THE DECISION MADE FROM THE DATA ARE REVIEWED FOR THESE PROBLEM AREAS: (1) DEFINITION OF THE VARIABLES TO BE ASSESSED; (2) DELINEATION OF THE RESEARCH POPULATION; (3) PROCESS OF SAMPLE SELECTION; (4) IMPLEMENTATION OF THE RESEARCH IN THE FIELD; AND (5) STATISTICAL PRESENTATION OF THE DATA. AGREEMENT SHOULD BE MADE AMONG RESEARCHERS ON THE DEFINITION OF WHAT IS TO BE STUDIED BEFORE ACTUAL MEASUREMENT AND ALL RELEVANT VARIABLES SHOULD BE CLARIFIED. THE IDENTIFICATION OF SAMPLE SUBTYPES MUST BE AGREED UPON, AND THE RESEARCHER SHOULD MAINTAIN A CONCEPTUAL PURITY REGARDING THE MIX OF SERVICE TYPES WITHIN ANY PARTICULAR SUBGROUP. THE TIME FRAME USED FOR EVALUATION SHOULD BE SUFFICIENT TO OBTAIN LEGITIMATE RESULTS. TO ENSURE REPRESENTATIVE SAMPLE SELECTION, AGENCY COMMUNICATION BETWEEN ADMINISTRATORS AND SERVICE PERSONNEL SHOULD ALWAYS BE KEPT OPEN. PROBLEMS IN RESEARCH IMPLEMENTATION INCLUDE REFUSAL RATES AND NONCOOPERATIVE OR NONTRUTHFUL RESPONSES FROM THE SAMPLE POPULATION. IN PRESENTING DATA, THE EVALUATOR SHOULD CONSIDER POSSIBLE DIFFERENCES IN THE INTERESTS OF THE AGENCY PERSONNEL AND AVOID PREMATURE EXPOSURE TO DATA WHICH MIGHT LATER BE REVISED AFTER FURTHER ANALYSIS.