NCJ Number
16034
Journal
Public Administration Review Dated: (JANUARY/FEBRUARY 1972) Pages: 49-52
Date Published
1972
Length
4 pages
Annotation
LIAR (LINKED INDICES FOR ASSESSING RELEVANCE) IS A MATHEMATICAL MODEL TO PREDICT THE PROBABLE UTILIZATION OF EVALUATIVE RESEARCH FINDINGS IN PUBLIC PROGRAMS.
Abstract
THIS COMPUTERIZED MODEL GIVES AN OBJECTIVE OBSERVER A WAY OF PREDICTING AN EVALUATION'S PROBABLE UTILIZATION AT ANY TIME FROM THE BIRTH OF THE IDEA TO THE COMPLETION OF THE REPORT. IN ITS SIMPLEST FORM, WHICH REQUIRES THE INPUT OF ONLY SIX OBSERVATIONS, THE LIAR MODEL IS BASED ON THE FOLLOWING TYPES OF VARIABLES - INCENTIVES TO UTILIZATION, DISCENTIVES TO UTILIZATION, AND ENVIRONMENTAL CONSTRAINTS. EACH VARIABLE GIVEN A WEIGHT FROM ONE TO TEN. IT IS RECOMMENDED THAT OBSERVATIONS BE TAKEN AT LEAST TWO POINTS, PREFERABLY AT THE START OF THE EVALUATION AND JUST BEFORE THE FINAL REPORT IS CIRCULATED. AT THE START OF A STUDY, THE MODEL SERVES AS A TOOL FOR THE HEAD OF AN EVALUATION DIVISION IN DECIDING WHETHER OR NOT TO PERFORM AN EVALUATION BY PERMITTING HIM TO SELECT PROJECTS WHERE THE REPORT IS LIKELY TO ATTRACT FAVORABLE COMMENT AND PERHAPS BE ACTED UPON. AT THE LATER STAGE, IT PERMITS A CHANGED PREDICTION IN THE LIGHT OF NEW CIRCUMSTANCES AND HELPS THE EVALUATOR TO DECIDE HOW MUCH OF HIS PRESTIGE SHOULD BE COMMITTED TO PUBLICIZING THE FINAL REPORT. CRITICISMS OF THE LIAR MODEL INCLUDE A LACK OF MATHEMATICAL PURITY IN THE PROBABILITY FORMULA AND A FAILURE TO TAKE INTO ACCOUNT THE FACT THAT AN ORGANIZATION'S ENTHUSIASM FOR MAKING EVALUATIONS VARIES INVERSELY WITH ITS DISPOSITION TO TAKE ANY OTHER ACTION. (AUTHOR ABSTRACT MODIFIED)