The Forest before the Trees / 1: |
Why Statistics? / 1.0: |
Statistics as a Form of Social Control / 1.01: |
Objections to Null Hypothesis Significance Testing / 1.02: |
Should Significance Tests be Banned? / 1.03: |
Math Modeling's the Ultimate Answer / 1.04: |
Some Recent Developments in Univariate Statistics / 1.05: |
Why Multivariate Statistics? / 1.1: |
Bonferroni Adjustment: An Alternative to Multivariate Statistics / 1.1.1: |
Why Isn't Bonferroni Adjustment Enough? / 1.1.2: |
A Heuristic Survey of Statistical Techniques / 1.2: |
Student's t test / 1.2.1: |
One-Way Analysis of Variance / 1.2.2: |
Hotelling's T[superscript 2] / 1.2.3: |
One-Way Multivariate Analysis of Variance / 1.2.4: |
Higher Order Analysis of Variance / 1.2.5: |
Higher Order Manova / 1.2.6: |
Pearson r and Bivariate Regression / 1.2.7: |
Multiple Correlation and Regression / 1.2.8: |
Path Analysis / 1.2.9: |
Canonical Correlation / 1.2.10: |
Analysis of Covariance / 1.2.11: |
Principal Component Analysis / 1.2.12: |
Factor Analysis / 1.2.13: |
Structural Equation Modeling / 1.2.14: |
Learning to Use Multivariate Statistics / 1.3: |
A Taxonomy of Linear Combinatons / 1.3.1: |
Why the Rest of the Book? / 1.3.2: |
Quiz 1 See How Much You Know after Reading Just One Chapter! |
Sample Answers to Quiz 1 |
Multiple Regression: Predicting One Variable from Many / 2: |
Data Set 1 |
The Model / 2.1: |
Choosing Weights / 2.2: |
Least Squares Criterion / 2.2.1: |
Maximum Correlation Criterion / 2.2.2: |
The Utility of Matrix Algebra / 2.2.3: |
Independence of Irrelevant Parameters / 2.2.4: |
Relating the Sample Equation to the Population Equation / 2.3: |
R[subscript x] versus S[subscript x] versus x'x as the Basis for MRA / 2.3.1: |
Specific Comparisons / 2.3.2: |
Illustrating Significance Tests / 2.3.3: |
Stepwise Multiple Regression Analysis / 2.3.4: |
Computer Programs for Multiple Regression / 2.4: |
Computer Logic and Organization / 2.4.1: |
Sage Advice on Use of Computer Programs / 2.4.2: |
Computerized Multiple Regression Analysis / 2.4.3: |
Some General Properties of Covariance Matrices / 2.5: |
Measuring the Importance of the Contribution of a Single Variable / 2.6: |
Anova via MRA / 2.7: |
Alternatives to the Least-Squares Criterion / 2.8: |
Path analytic Terminology / 2.9: |
Preconditions for Path Analysis / 2.9.2: |
Estimating and Testing Path coefficients / 2.9.3: |
Decomposition of Correlations into Components / 2.9.4: |
Overall Test of Goodness of fit / 2.9.5: |
Examples / 2.9.6: |
Some Path-Analysis References / 2.9.7: |
Demonstration Problem |
Answers |
Some Real Data and a Quiz Thereon |
Path Analysis Problem |
Answers to Path Analysis Problem |
Hotelling's T[superscript 2]: Tests on One or Two Mean Vectors / 3: |
Single-Sample t and T[superscript 2] / 3.1: |
Linearly Related Outcome Variables / 3.2: |
Two-Sample t and T[superscript 2] / 3.3: |
Profile Analysis / 3.4: |
Discriminant Analysis / 3.5: |
Relationship between T[superscript 2] and MRA / 3.6: |
Assumptions Underlying T[superscript 2] / 3.7: |
The Assumption of Equal Covariance Matrices / 3.7.1: |
Known Covariance Matrix / 3.7.2: |
The Assumption of Multivariate Normality / 3.7.3: |
Analyzing Repeated-Measures Designs via T[superscript 2] / 3.8: |
Single-Symbol Expressions for Simple Cases / 3.9: |
Computerized T[superscript 2] / 3.10: |
Single-Sample and Two-Sample T[superscript 2] / 3.10.1: |
Within-Subjects Anova / 3.10.2: |
Demonstration Problems |
Multivariate Analysis of Variance: Differences Among Several Groups on Several Measures / 4: |
One-Way (Univariate) Analysis of Variance / 4.1: |
The Overall Test / 4.1.1: |
Multiple Profile Analysis / 4.1.2: |
Multiple Discriminant Analysis / 4.4: |
Greatest Characteristic Roots versus Multiple-Root Tests in Manova / 4.5: |
"Protected" Univariate Tests / 4.5.1: |
Simulataneous Test Procedures and Union Intersection / 4.5.2: |
Invalidity of Partitioned-U Tests of Individual Roots / 4.5.3: |
Simplified Coefficients as a Solution to the Robustness Problem / 4.5.4: |
Finite-Intersection Tests / 4.5.5: |
Simple Cases of Manova / 4.6: |
Higher Order Anova: Interactions / 4.7: |
Within-Subject Univariate Anova Versus Manova / 4.8: |
Computerized Manova / 4.10: |
Generic Setup for SPSS MANOVA / 4.10.1: |
Supplementary Computations / 4.10.2: |
Pointing and Clicking to a Manova on SPSS PC / 4.10.3: |
Generic Setup for SAS PROC GLM / 4.10.4: |
Canonical Correlation: Relationships Between Two Sets of Variables / 5: |
Formulae for Computing Canonical Rs / 5.1: |
Heuristic Justification of Canonical Formulae / 5.1.1: |
Simple Cases of Canonical Correlations / 5.1.2: |
Example of a Canonical Analysis / 5.1.3: |
Relationships to Other Statistical Techniques / 5.2: |
Likelihood-Ratio Tests of Relationships between Sets of Variables / 5.3: |
Generalization and Specialization of Canonical Analysis / 5.4: |
Testing the Independence of m Sets of Variables / 5.4.1: |
Repeated-Battery Canona / 5.4.2: |
Rotation of Canonical Variates / 5.4.3: |
The Redundancy Coefficient / 5.4.4: |
What's Missing from Canonical Analysis? / 5.4.5: |
Computerized Canonical Correlation / 5.5: |
Matrix-Manipulation Systems / 5.5.1: |
SAS PROC CANCORR / 5.5.2: |
Canona via SPSS MANOVA / 5.5.3.: |
SPSS Canona From Correlation Matrix: Be Careful / 5.5.4: |
Demonstration Problems and Some Real Data Employing Canonical Correlation |
Principal Component Analysis: Relationships Within a Single Set of Variables / 6: |
Definition of Principal Components / 6.1: |
Terminology and Notation in PCA and FA / 6.1.1: |
Scalar Formulae for Simple Cases of PCA / 6.1.2: |
Computerized PCA / 6.1.3: |
Additional Unique Properties (AUPs) of PCs / 6.1.4: |
Interpretation of Principal Components / 6.2: |
Uses of Principal Components / 6.3: |
Uncorrelated Contributions / 6.3.1: |
Computational Convenience / 6.3.2: |
Principal Component Analysis as a Means of Handling Linear Dependence / 6.3.3: |
Examples of PCA / 6.3.4: |
Quantifying Goodness of Interpretation of Components / 6.3.5: |
Significance Tests for Principal Components / 6.4: |
Sampling Properties of Covariance-Based PCs / 6.4.1: |
Sampling Properties of Correlation-Based PCs / 6.4.2: |
Rotation of Principal Components / 6.5: |
Basic Formulae for Rotation / 6.5.1: |
Objective Criteria for Rotation / 6.5.2: |
Examples of Rotated PCs / 6.5.3: |
Individual Scores on Rotated PCs / 6.5.4: |
Uncorrelated-Components Versus Orthogonal-Profiles Rotation / 6.5.5: |
Factor Analysis: The Search for Structure / 7: |
Communalities / 7.1: |
Theoretical Solution / 7.2.1: |
Empirical Approximations / 7.2.2: |
Iterative Procedure / 7.2.3: |
Is the Squared Multiple Correlation the True Communality? / 7.2.4: |
Factor Analysis Procedures Requiring Communality Estimates / 7.3: |
Principal Factor Analysis / 7.3.1: |
Triangular (Choleski) Decomposition / 7.3.2: |
Centroid Analysis / 7.3.3: |
Methods Requiring Estimate of Number of Factors / 7.4: |
Other Approaches to Factor Analysis / 7.5: |
Factor Loadings versus Factor Scores / 7.6: |
Factor Score Indeterminacy / 7.6.1: |
Relative Validities of Loadings-Derived versus Scoring-Coefficient-Derived Factor Interpretations / 7.6.2: |
Regression-Based Interpretation of Factors is Still a Hard Sell / 7.6.3: |
Relative Merits of Principal Component Analysis versus Factor Analysis / 7.7: |
Similarity of Factor Scoring Coefficients / 7.7.1: |
Bias in Estimates of Factor Loadings / 7.7.2: |
Computerized Exploratory Factor Analysis / 7.8: |
Confirmatory Factor Analysis / 7.9: |
Sas Proc Calis / 7.9.1: |
The Forest Revisited / 8: |
Scales of Measurement and Multivariate Statistics / 8.1: |
Effects of Violations of Distributional Assumptions in Multivariate Analysis / 8.2: |
Nonlinear Relationships in Multivariate Statistics / 8.3: |
The Multivariate General Linear Hypothesis / 8.4: |
General Approach and Examples / 8.5: |
SEM Is Not a General Model for Multivariate Statistics / 8.5.2: |
Other User-Friendly SEM Programs / 8.5.3: |
Where to Go from Here / 8.6: |
Summing Up / 8.7: |
Finding Maxima and Minima of Polynomials / Digression 1: |
Derivatives and Slopes / D1.1: |
Optimization Subject to Constraints / D1.2: |
Matrix Algebra / Digression 2: |
Basic Notation / D2.1: |
Linear Combinations of Matrices / D2.2: |
Multiplication of Matrices / D2.3: |
Permissible Manipulations / D2.4: |
Inverses / D2.5: |
Determinants / D2.6: |
Some Handy Formulae for Inverses and Determinants in Simple Cases / D2.7: |
Rank / D2.8: |
Matrix Calculus / D2.9: |
Partitioned Matrices / D2.10: |
Characteristic Roots and Vectors / D2.11: |
Solution of Homogeneous Systems of Equations / D2.12: |
Solution of Cubic Equations / Digression 3: |
Statistical Tables / Appendix A: |
(Why omitted from this edition) / A.1 - A.4: |
Greatest Characteristic Root Distribution / A.5: |
Computer Programs Available from the Author / Appendix B: |
cvinter: p values and Critical Values for Univariate Statistics / B.1: |
gcrinter: Critical Values for the Greatest Characteristic Root (g.c.r.) Distribution / B.2: |
Derivations / Appendix C: |
Per-Experiment and Experimentwise Error Rates for Bonferroni-Adjusted Tests / Derivation 1.1: |
Scalar Formulae for MRA with One, Two, and Three Predictors / Derivation 2.1: |
Coefficients That Minimize Error Also Maximize Correlation / Derivation 2.2: |
Maximizing r via Matrix Algebra / Derivation 2.3: |
Variances of b[subscript j]s and of Linear Combinations Thereof / Derivation 2.4: |
Drop in R[superscript 2] = b[superscript 2][subscript j](1 - [characters not reproducible]) / Derivation 2.6: |
MRA on Group-Membership Variables Yields Same F As Anova / Derivation 2.7: |
Unweighted Means and Least-Squares Anova Are Identical in the 2[superscript n] Design / Derivation 2.8: |
T[superscript 2] and Associated Discriminant Function / Derivation 3.1: |
Single-Sample T[superscript 2] |
Two-Sample T[superscript 2] |
Two-Sample t Versus Pearson r With Group-Membership Variables / Derivation 3.2: |
Single-Sample t Test versus "Raw-Score" r[subscript xy] |
T[superscript 2] Versus MRA |
Maximizing F(a) in Manova / Derivation 4.1: |
Canonical Correlation and Canonical Variates / Derivation 5.1: |
Canonical Correlation as "Mutual Regression Analysis" / Derivation 5.2: |
Relationship between Canonical Analysis and Manova / Derivation 5.3: |
Principal Components / Derivation 6.1: |
PC Coefficients Define Both Components in Terms of Xs and Xs in Terms of PCs / Derivation 6.2: |
What Does Rotation of Loadings Do to Coefficients? / Derivation 6.3: |
Near Equivalence of PCA and Equal-Communalities PFA / Derivation 7.1: |
References |
Index |