. 2007 by Taylor & Francis Group, LLC
Lawrence Erlbaum Associates is an imprint of Taylor & Francis Group, an Informa business
Preface xi
Chapter 1
INTRODUCTION 1
1.1 Focus and Overview of Topics
1.2 Some Basic Descriptive Statistics
1.3 Summation Notation
1.4 t Test for Independent Samples
1.5 t Test for Dependent Samples
1.6 Outliers
1.7 SPSS and SAS Statistical Packages
1.8 SPSS for Windows—Release 12.0
1.9 Data Files
1.10 Data Entry
1.11 Editing a Dataset
1.12 Splitting and Merging Files
1.13 Two Ways of Running Analyses on SPSS
1.14 SPSS Output Navigator
1.15 SAS and SPSS Output for Correlations, Descriptives, and t Tests
1.16 Data Sets on Compact Disk
Appendix Obtaining the Mean and Variance on the Tl–30Xa Calculator
Chapter 2
ONE WAY ANALYSIS OF VARIANCE 45
2.1 Introduction
2.2 Rationale for ANOVA
2.3 Numerical Example
2.4 Expected Mean Squares
2.5 MSw and MSb as Variances
v
2.6 A Linear Model for the Data
2.7 Assumptions in ANOVA
2.8 The Independence Assumption
2.9 ANOVA on SPSS and SAS
2.10 Post Hoc Procedures
2.11 Tukey Procedure
2.12 The Scheffé Procedure
2.13 Heterogeneous Variances and Unequal Group Sizes
2.14 Measures of Association (Variance Accounted For)
2.15 Planned Comparisons
2.16 Test Statistic for Planned Comparisons
2.17 Planned Comparisons on SPSS and SAS
2.18 The Effect of an Outlier on an ANOVA
2.19 Multivariate Analysis of Variance
2.20 Summary
Appendix
Chapter 3
POWER ANALYSIS 105
3.1 Introduction
3.2 t Test for Independent Samples
3.3 A Priori and Post Hoc Estimation of Power
3.4 Estimation of Power for One Way Analysis of Variance
3.5 A Priori Estimation of Subjects Needed for a Given Power
3.6 Ways of Improving Power
3.7 Power Estimation on SPSS MANOVA
3.8 Summary
Chapter 4
FACTORIAL ANALYSIS OF VARIANCE 123
4.1 Introduction
4.2 Numerical Calculations for Two Way ANOVA
4.3 Balanced and Unbalanced Designs
4.4 Higher Order Designs
4.5 A Comprehensive Computer Example Using Real Data
4.6 Power Analysis
4.7 Fixed and Random Factors
4.8 Summary
Appendix Doing a Balanced Two Way ANOVA With a Calculator
vi CONTENTS
Chapter 5
REPEATED MEASURES ANALYSIS 181
5.1 Introduction
5.2 Advantages and Disadvantages of Repeated Measures Designs
5.3 Single Group Repeated Measures
5.4 Completely Randomized Design
5.5 Univariate Repeated Measures Analysis
5.6 Assumptions in Repeated Measures Analysis
5.7 Should We Use the Univariate or Multivariate Approach?
5.8 Computer Analysis on SAS and SPSS for Example
5.9 Post Hoc Procedures in Repeated Measures Analysis
5.10 One Between and One Within Factor—A Trend Analysis
5.11 Post Hoc Procedures for the One Between and One Within Design
5.12 One Between and Two Within Factors
5.13 Totally Within Designs
5.14 Planned Comparisons in Repeated Measures Designs
5.15 Summary
Chapter 6
SIMPLE AND MULTIPLE REGRESSION 219
6.1 Simple Regression
6.2 Assumptions for the Errors
6.3 Influential Data Points
6.4 Multiple Regression
6.5 Breakdown of Sum of Squares in Regression and F Test for Multiple Correlation
6.6 Relationship of Simple Correlations to Multiple Correlation
6.7 Multicollinearity
6.8 Model Selection
6.9 Two Computer Examples
6.10 Checking Assumptions for the Regression Model
6.11 Model Validation
6.12 Importance of the Order of Predictors in Regression Analysis
6.13 Other Important Issues
6.14 Outliers and Influential Data Points
6.15 Further Discussion of the Two Computer Examples
6.16 Sample Size Determination for a Reliable Prediction Equation
6.17 ANOVA as a Special Case of Regression Analysis
6.18 Summary of Important Points
Appendix The PRESS Statistic
CONTENTS vii
Chapter 7
ANALYSIS OF COVARIANCE 285
7.1 Introduction
7.2 Purposes of Covariance
7.3 Adjustment of Posttest Means
7.4 Reduction of Error Variance
7.5 Choice of Covariates
7.6 Numerical Example
7.7 Assumptions in Analysis of Covariance
7.8 Use of ANCOVA with Intact Groups
7.9 Computer Example for ANCOVA
7.10 Alternative Analyses
7.11 An Alternative to the Johnson–Neyman Technique
7.12 Use of Several Covariates
7.13 Computer Example with Two Covariates
7.14 Summary
Chapter 8
HIERARCHICAL LINEAR MODELING 321
8.1 Introduction
8.2 Problems Using Single-Level Analyses of Multilevel Data
8.3 Formulation of the Multilevel Model
8.4 Two-Level Model—General Formulation
8.5 HLM6 Software
8.6 Two Level Example—Student and Classroom Data
8.7 HLM Software Output
8.8 Adding Level One Predictors to the HLM
8.9 Addition of a Level Two Predictor to a Two Level HLM
8.10 Evaluating the Efficacy of a Treatment
8.11 Final Comments on Hlm
Appendix A
DATA SETS 365
A.1 Clinical Data
A.2 Alcoholics Data
A.3 Sesame Street Data
A.4 Headache Data
A.5 Cartoon Data
viii CONTENTS
A.6 Attitude Data
A.7 National Academy of Sciences Data
A.8 Agresti Home Sales Data
Appendix B
STATISTICAL TABLES 399
Table B.1 Critical Values for F
Table B.2 Percentile Points of Studentized Range Statistic
Table B.3 Critical Values for Dunnett’s Test
Table B.4 Critical Values for F (max) Statistic
Table B.5 Critical Values for Bryant-Paulson Procedure
Appendix C
POWER TABLES 413
Table C.1 Power of F Test at α = .05, u = 1
Table C.2 Power of F Test at α = .05, u = 2
Table C.3 Power of F Test at α = .05, u = 3
Table C.4 Power of F Test at α = .05, u = 4
Table C.5 Power of F Test at α = .10, u = 1
Table C.6 Power of F Test at α = .10, u = 2
Table C.7 Power of F Test at α = .10, u = 3
Table C.8 Power of F Test at α = .10, u = 4
References 423
Answers to Selected Exercises 431
Author Index 453
Subject Index 457