NCJ Number
45900
Date Published
1973
Length
671 pages
Annotation
THIS TEXT FOR A COURSE IN EXPERIMENTAL DESIGN AND STATISTICAL ANALYSIS PRESENTS THE FUNDAMENTAL RULES FOR GENERATING ANALYSES AND FACTORIAL EXPERIMENT DESIGNS; THE RESEARCH APPLICATIONS OF STATISTICS ARE STRESSED.
Abstract
FOLLOWING A BRIEF INTRODUCTION TO RESEARCH, THE ROLE OF STATISTICS, AND MEASURES OF GENERAL TENDENCY AND VARIABILITY, THE NEXT SEVEN CHAPTERS TREAT HYPOTHESIS TESTING LOGIC, PARTITIONING THE TOTAL SUM OF SQUARES, VARIANCE ESTIMATES AND THE F RATIO, ASSUMPTIONS AND ADDITIONAL CONSIDERATIONS, COMPARISONS AMONG TREATMENT MEANS, ORTHOGONAL ANALYSES, AND MULTIPLE COMPARISONS. OTHER TOPICS INCLUDE THE ADVANTAGES OF FACTORIAL DESIGN, INTERACTION, THE RATIONALE AND RULES FOR CALCULATING MAJOR EFFECTS, COMPARISONS AMONG MARGINAL MEANS AND SIMPLE MAIN EFFECTS, AND ADDITIONAL COMPARISONS INVOLVING CELL MEANS. DISCUSSIONS OF THE THREE-FACTOR CASE, THE GENERAL CASE STRUCTURAL MODELS, THE ANALYSIS OF VARIANCE, AND THE ANALYSIS OF EXPERIMENTS BOTH WITH UNEQUAL SAMPLE SIZES AND WITH NESTED FACTORS FOLLOW. LATER CHAPTERS COVER DESIGNS WITH REPEATED MEASURES (SINGLE-FACTOR DESIGNS AND HIGHER-ORDER FACTORIALS) COVARIANCE ANALYSIS, RANDOMIZED BLOCKS, SENSITIVITY OF EXPERIMENTAL DESIGNS, AND SOME CONTROVERSIAL TOPICS (E.G., MAGNITUDE OF TREATMENT EFFECTS, DATA TRANSFORMATIONS, AND POOLING IN THE ANALYSIS OF VARIANCE). MATHEMATICAL ARGUMENTS ARE HELD TO A MINIMUM, AND EMPHASIS IS PLACED ON AN INTUITIVE UNDERSTANDING OF THE MATHEMATICAL OPERATIONS INVOLVED. A SPECIALIZED NOTATIONAL SYSTEM MINIMIZES CONFUSION FOR THE BEGINNING STUDENT AND TIES IN DIRECTLY WITH THE GENERAL COMPUTATIONAL RULES. EXERCISES APPEAR AT THE END OF MOST MAJOR SECTIONS, AND ANSWERS ARE GIVEN IN SUFFICIENT DETAIL TO ENABLE THE STUDENT TO LOCATE THE SOURCE OF AN ERROR. REFERENCES AND AUTHOR AND SUBJUCT INDEXES ARE PROVIDED. (AUTHOR ABSTRACT MODIFIED--DAS)