NCJ Number
58072
Date Published
1978
Length
144 pages
Annotation
AIMED AT PROGRAM EVALUATORS, THIS VOLUME OF AN EIGHT-PART SERIES ON PROGRAM EVALUATION OUTLINES, IN LANGUAGE ACCESSIBLE TO THE LAYMAN, HOW TO CALCULATE STATISTICS RELEVANT TO A PROGRAM'S EVALUATION PHASE.
Abstract
EACH OF THE THREE MAIN SECTIONS OF THIS BOOK DEALS WITH ONE OF THE THREE MAJOR FUNCTIONS OF STATISTICS IN EVALUATIONS: (1) SUMMARIZING SCORES THROUGH MEASURES OF CENTRAL TENDENCY AND VARIANCE, (2) TESTING SIGNIFICANCE OF DIFFERENCES FOUND AMONG PERFORMANCES OF GROUPS, AND (3) CORRELATION. THE FIRST SECTION INVOLVES SUMMARIZING A SET OF SCORES FROM THE ADMINISTRATION OF A MEASURE TO ONE GROUP BY GRAPHING A DISTRIBUTION OF SCORES, CALCULATING A MEAN, AND CALCULATING THE STANDARD DEVIATION AND VARIANCE. THE SECOND CENTERS ON EXAMINING THE DIFFERENCES BETWEEN SCORES FROM TWO GROUPS ON THE SAME MEASURE. COMPARING SCORES FROM UNMATCHED GROUPS INVOLVES CALCULATING THE CONFIDENCE LIMITS FOR THE DIFFERENCE BETWEEN TWO MEANS, THE T-TEST, OR THE MANN-WHITNEY U TEST, WHILE COMPARING SCORES FROM MATCHED GROUPS INVOLVES THE SIGN TEST, IN ADDITION TO CONFIDENCE LIMITS AND THE T-TEST. THE THIRD SECTION RELIES ON EXAMINING RELATIONSHIPS BETWEEN TWO MEASURES MADE ON THE SAME GROUP BY USING A GRAPH TO ILLUSTRATE A RELATIONSHIP, PEARSON'S PRODUCT MOMENT CORRELATION COEFFICIENT, SPEARMAN'S RANK ORDER CORRELATION COEFFICIENT, PHI COEFFICIENT, RANK BISERIAL CORRELATION COEFFICIENT, POINT BISERIAL CORRELATION COEFFICIENT, DETERMINING CONFIDENCE INTERVALS, AND CHI-SQUARE TESTING. A FINAL SECTION ON THE USE OF COMPUTERS TO ANALYZE DATA FOCUSES ON WHAT EVALUATORS SHOULD KNOW ABOUT DATA ANALYSIS BY COMPUTER, PREPARING DATA, TYPES OF ANALYSES THAT CAN BE PERFORMED, ANALYSIS OF VARIANCE, DISCRIMINANT ANALYSIS, AND FACTOR ANALYSIS. TABULAR AND GRAPHIC ILLUSTRATIONS ARE PROVIDED, ALONG WITH AN INDEX.