| |
| |
Preface | |
| |
| |
About the Authors | |
| |
| |
The CD-ROM | |
| |
| |
| |
Statistical Models | |
| |
| |
| |
Mathematical and Statistical Models | |
| |
| |
| |
Functional Aspects of Models | |
| |
| |
| |
The Inferential Steps--Estimation and Testing | |
| |
| |
| |
t-Tests in Terms of Statistical Models | |
| |
| |
| |
Embedding Hypotheses | |
| |
| |
| |
Hypothesis and Significance Testing--Interpretation of the p-Value | |
| |
| |
| |
Classes of Statistical Models | |
| |
| |
| |
The Basic Component Equation | |
| |
| |
| |
Linear and Nonlinear Models | |
| |
| |
| |
Regression and Analysis of Variance Models | |
| |
| |
| |
Univariate and Multivariate Models | |
| |
| |
| |
Fixed, Random, and Mixed Effects Models | |
| |
| |
| |
Generalized Linear Models | |
| |
| |
| |
Errors in Variable Models | |
| |
| |
| |
Data Structures | |
| |
| |
| |
Introduction | |
| |
| |
| |
Classification by Response Type | |
| |
| |
| |
Classification by Study Type | |
| |
| |
| |
Clustered Data | |
| |
| |
| |
Clustering through Hierarchical Random Processes | |
| |
| |
| |
Clustering through Repeated Measurements | |
| |
| |
| |
Autocorrelated Data | |
| |
| |
| |
The Autocorrelation Function | |
| |
| |
| |
Consequences of Ignoring Autocorrelation | |
| |
| |
| |
Autocorrelation in Designed Experiments | |
| |
| |
| |
From Independent to Spatial Data--a Progression of Clustering | |
| |
| |
| |
Linear Algebra Tools | |
| |
| |
| |
Introduction | |
| |
| |
| |
Matrices and Vectors | |
| |
| |
| |
Basic Matrix Operations | |
| |
| |
| |
Matrix Inversion--Regular and Generalized Inverse | |
| |
| |
| |
Mean, Variance, and Covariance of Random Vectors | |
| |
| |
| |
The Trace and Expectation of Quadratic Forms | |
| |
| |
| |
The Multivariate Gaussian Distribution | |
| |
| |
| |
Matrix and Vector Differentiation | |
| |
| |
| |
Using Matrix Algebra to Specify Models | |
| |
| |
| |
Linear Models | |
| |
| |
| |
Nonlinear Models | |
| |
| |
| |
Variance-Covariance Matrices and Clustering | |
| |
| |
| |
The Classical Linear Model: Least Squares and Alternatives | |
| |
| |
| |
Introduction | |
| |
| |
| |
Least Squares Estimation and Partitioning of Variation | |
| |
| |
| |
The Principle | |
| |
| |
| |
Partitioning Variability through Sums of Squares | |
| |
| |
| |
Sequential and Partial Sums of Squares and the Sum of Squares Reduction Test | |
| |
| |
| |
Factorial Classification | |
| |
| |
| |
The Means and Effects Model | |
| |
| |
| |
Effect Types in Factorial Designs | |
| |
| |
| |
Sum of Squares Partitioning through Contrasts | |
| |
| |
| |
Effects and Contrasts in The SAS System | |
| |
| |
| |
Diagnosing Regression Models | |
| |
| |
| |
Residual Analysis | |
| |
| |
| |
Recursive and Linearly Recovered Errors | |
| |
| |
| |
Case Deletion Diagnostics | |
| |
| |
| |
Collinearity Diagnostics | |
| |
| |
| |
Ridge Regression to Combat Collinearity | |
| |
| |
| |
Diagnosing Classification Models | |
| |
| |
| |
What Matters? | |
| |
| |
| |
Diagnosing and Combating Heteroscedasticity | |
| |
| |
| |
Median Polishing of Two-Way Layouts | |
| |
| |
| |
Robust Estimation | |
| |
| |
| |
L[subscript 1]-Estimation | |
| |
| |
| |
M-Estimation | |
| |
| |
| |
Robust Regression for Prediction Efficiency Data | |
| |
| |
| |
M-Estimation in Classification Models | |
| |
| |
| |
Nonparametric Regression | |
| |
| |
| |
Local Averaging and Local Regression | |
| |
| |
| |
Choosing the Smoothing Parameter | |
| |
| |
Appendix A on CD-ROM | |
| |
| |
| |
Mathematical Details | |
| |
| |
| |
Least Squares | |
| |
| |
| |
Hypothesis Testing in the Classical Linear Model | |
| |
| |
| |
Diagnostics in Regression Models | |
| |
| |
| |
Ridge Regression | |
| |
| |
| |
L[subscript 1]-Estimation | |
| |
| |
| |
M-Estimation | |
| |
| |
| |
Nonparametric Regression | |
| |
| |
| |
Nonlinear Models | |
| |
| |
| |
Introduction | |
| |
| |
| |
Models as Laws or Tools | |
| |
| |
| |
Linear Polynomials Approximate Nonlinear Models | |
| |
| |
| |
Fitting a Nonlinear Model to Data | |
| |
| |
| |
Estimating the Parameters | |
| |
| |
| |
Tracking Convergence | |
| |
| |
| |
Starting Values | |
| |
| |
| |
Goodness-of-Fit | |
| |
| |
| |
Hypothesis Tests and Confidence Intervals | |
| |
| |
| |
Testing the Linear Hypothesis | |
| |
| |
| |
Confidence and Prediction Intervals | |
| |
| |
| |
Transformations | |
| |
| |
| |
Transformation to Linearity | |
| |
| |
| |
Transformation to Stabilize the Variance | |
| |
| |
| |
Parameterization of Nonlinear Models | |
| |
| |
| |
Intrinsic and Parameter-Effects Curvature | |
| |
| |
| |
Reparameterization through Defining Relationships | |
| |
| |
| |
Applications | |
| |
| |
| |
Basic Nonlinear Analysis with The SAS System--Mitscherlich's Yield Equation | |
| |
| |
| |
The Sampling Distribution of Nonlinear Estimators--the Mitscherlich Equation Revisited | |
| |
| |
| |
Linear-Plateau Models and Their Relatives--a Study of Corn Yields from Tennessee | |
| |
| |
| |
Critical NO[subscript 3] Concentrations as a Function of Sampling Depth--Comparing Join-Points in Plateau Models | |
| |
| |
| |
Factorial Treatment Structure with Nonlinear Response | |
| |
| |
| |
Modeling Hormetic Dose Response through Switching Functions | |
| |
| |
| |
Modeling a Yield-Density Relationship | |
| |
| |
| |
Weighted Nonlinear Least Squares Analysis with Heteroscedastic Errors | |
| |
| |
Appendix A on CD-ROM | |
| |
| |
| |
Forms of Nonlinear Models | |
| |
| |
| |
Concave and Convex Models, Yield-Density Models | |
| |
| |
| |
Models with Sigmoidal Shape, Growth Models | |
| |
| |
| |
Mathematical Details | |
| |
| |
| |
Taylor Series Involving Vectors | |
| |
| |
| |
Nonlinear Least Squares and the Gauss-Newton Algorithm | |
| |
| |
| |
Nonlinear Generalized Least Squares | |
| |
| |
| |
The Newton-Raphson Algorithm | |
| |
| |
| |
Convergence Criteria | |
| |
| |
| |
Hypothesis Testing, Confidence and Prediction Intervals | |
| |
| |
| |
Generalized Linear Models | |
| |
| |
| |
Introduction | |
| |
| |
| |
Components of a Generalized Linear Model | |
| |
| |
| |
Random Component | |
| |
| |
| |
Systematic Component and Link Function | |
| |
| |
| |
Generalized Linear Models in The SAS System | |
| |
| |
| |
Grouped and Ungrouped Data | |
| |
| |
| |
Parameter Estimation and Inference | |
| |
| |
| |
Solving the Likelihood Problem | |
| |
| |
| |
Testing Hypotheses about Parameters and Their Functions | |
| |
| |
| |
Deviance and Pearson's X[superscript 2] Statistic | |
| |
| |
| |
Testing Hypotheses through Deviance Partitioning | |
| |
| |
| |
Generalized R[superscript 2] Measures of Goodness-of-Fit | |
| |
| |
| |
Modeling an Ordinal Response | |
| |
| |
| |
Cumulative Link Models | |
| |
| |
| |
Software Implementation and Example | |
| |
| |
| |
Overdispersion | |
| |
| |
| |
Applications | |
| |
| |
| |
Dose-Response and LD[subscript 50] Estimation in a Logistic Regression Model | |
| |
| |
| |
Binomial Proportions in a Randomized Block Design--the Hessian Fly Experiment | |
| |
| |
| |
Gamma Regression and Yield Density Models | |
| |
| |
| |
Effects of Judges' Experience on Bean Canning Quality Ratings | |
| |
| |
| |
Ordinal Ratings in a Designed Experiment with Factorial Treatment Structure and Repeated Measures | |
| |
| |
| |
Log-Linear Modeling of Rater Agreement | |
| |
| |
| |
Modeling the Sample Variance of Scab Infection | |
| |
| |
| |
A Poisson/Gamma Mixing Model for Overdispersed Poppy Counts | |
| |
| |
Appendix A on CD-ROM | |
| |
| |
| |
Mathematical Details and Special Topics | |
| |
| |
| |
Exponential Family of Distributions | |
| |
| |
| |
Maximum Likelihood Estimation | |
| |
| |
| |
Iteratively Reweighted Least Squares | |
| |
| |
| |
Hypothesis Testing | |
| |
| |
| |
Fieller's Theorem and the Variance of a Ratio | |
| |
| |
| |
Overdispersion Mechanisms for Counts | |
| |
| |
| |
Linear Mixed Models for Clustered Data | |
| |
| |
| |
Introduction | |
| |
| |
| |
The Laird-Ware Model | |
| |
| |
| |
Rationale | |
| |
| |
| |
The Two-Stage Concept | |
| |
| |
| |
Fixed or Random Effects | |
| |
| |
| |
Choosing the Inference Space | |
| |
| |
| |
Estimation and Inference | |
| |
| |
| |
Maximum and Restricted Maximum Likelihood | |
| |
| |
| |
Estimated Generalized Least Squares | |
| |
| |
| |
Hypothesis Testing | |
| |
| |
| |
Correlations in Mixed Models | |
| |
| |
| |
Induced Correlations and the Direct Approach | |
| |
| |
| |
Within-Cluster Correlation Models | |
| |
| |
| |
Split-Plots, Repeated Measures, and the Huynh-Feldt Conditions | |
| |
| |
| |
Applications | |
| |
| |
| |
Two-Stage Modeling of Apple Growth over Time | |
| |
| |
| |
On-Farm Experimentation with Randomly Selected Farms | |
| |
| |
| |
Nested Errors through Subsampling | |
| |
| |
| |
Recovery of Inter-Block Information in Incomplete Block Designs | |
| |
| |
| |
A Split-Strip-Plot Experiment for Soybean Yield | |
| |
| |
| |
Repeated Measures in a Completely Randomized Design | |
| |
| |
| |
A Longitudinal Study of Water Usage in Horticultural Trees | |
| |
| |
| |
Cumulative Growth of Muskmelons in Subsampling Design | |
| |
| |
Appendix A on CD-ROM | |
| |
| |
| |
Mathematical Details and Special Topics | |
| |
| |
| |
Henderson's Mixed Model Equations | |
| |
| |
| |
Solutions to the Mixed Model Equations | |
| |
| |
| |
Likelihood Based Estimation | |
| |
| |
| |
Estimated Generalized Least Squares Estimation | |
| |
| |
| |
Hypothesis Testing | |
| |
| |
| |
The First-Order Autoregressive Model | |
| |
| |
| |
Nonlinear Models for Clustered Data | |
| |
| |
| |
Introduction | |
| |
| |
| |
Nonlinear and Generalized Linear Mixed Models | |
| |
| |
| |
Toward an Approximate Objective Function | |
| |
| |
| |
Three Linearizations | |
| |
| |
| |
Linearization in Generalized Linear Mixed Models | |
| |
| |
| |
Integral Approximation Methods | |
| |
| |
| |
Applications | |
| |
| |
| |
A Nonlinear Mixed Model for Cumulative Tree Bole Volume | |
| |
| |
| |
Poppy Counts Revisited--a Generalized Linear Mixed Model for Overdispersed Count Data | |
| |
| |
| |
Repeated Measures with an Ordinal Response | |
| |
| |
Appendix A on CD-ROM | |
| |
| |
| |
Mathematical Details and Special Topics | |
| |
| |
| |
PA and SS Linearizations | |
| |
| |
| |
Generalized Estimating Equations | |
| |
| |
| |
Linearization in Generalized Linear Mixed Models | |
| |
| |
| |
Gaussian Quadrature | |
| |
| |
| |
Statistical Models for Spatial Data | |
| |
| |
| |
Changing the Mindset | |
| |
| |
| |
Samples of Size One | |
| |
| |
| |
Random Functions and Random Fields | |
| |
| |
| |
Types of Spatial Data | |
| |
| |
| |
Stationarity and Isotropy--the Built-in Replication Mechanism of Random Fields | |
| |
| |
| |
Semivariogram Analysis and Estimation | |
| |
| |
| |
Elements of the Semivariogram | |
| |
| |
| |
Parametric Isotropic Semivariogram Models | |
| |
| |
| |
The Degree of Spatial Continuity (Structure) | |
| |
| |
| |
Semivariogram Estimation and Fitting | |
| |
| |
| |
The Spatial Model | |
| |
| |
| |
Spatial Prediction and the Kriging Paradigm | |
| |
| |
| |
Motivation of the Prediction Problem | |
| |
| |
| |
The Concept of Optimal Prediction | |
| |
| |
| |
Ordinary and Universal Kriging | |
| |
| |
| |
Some Notes on Kriging | |
| |
| |
| |
Extensions to Multiple Attributes | |
| |
| |
| |
Spatial Regression and Classification Models | |
| |
| |
| |
Random Field Linear Models | |
| |
| |
| |
Some Philosophical Considerations | |
| |
| |
| |
Parameter Estimation | |
| |
| |
| |
Autoregressive Models for Lattice Data | |
| |
| |
| |
The Neighborhood Structure | |
| |
| |
| |
First-Order Simultaneous and Conditional Models | |
| |
| |
| |
Parameter Estimation | |
| |
| |
| |
Choosing the Neighborhood Structure | |
| |
| |
| |
Analyzing Mapped Spatial Point Patterns | |
| |
| |
| |
Introduction | |
| |
| |
| |
Random, Aggregated, and Regular Patterns--the Notion of Complete Spatial Randomness | |
| |
| |
| |
Testing the CSR Hypothesis in Mapped Point Patterns | |
| |
| |
| |
Second-Order Properties of Point Patterns | |
| |
| |
| |
Applications | |
| |
| |
| |
Exploratory Tools for Spatial Data--Diagnosing Spatial Autocorrelation with Moran's I | |
| |
| |
| |
Modeling the Semivariogram of Soil Carbon | |
| |
| |
| |
Spatial Prediction--Kriging of Lead Concentrations | |
| |
| |
| |
Spatial Random Field Models--Comparing C/N Ratios among Tillage Treatments | |
| |
| |
| |
Spatial Random Field Models--Spatial Regression of Soil Carbon on Soil N | |
| |
| |
| |
Spatial Generalized Linear Models--Spatial Trends in the Hessian Fly Experiment | |
| |
| |
| |
Simultaneous Spatial Autoregression--Modeling Wiebe's Wheat Yield Data | |
| |
| |
| |
Point Patterns--First- and Second-Order Properties of a Mapped Pattern | |
| |
| |
Appendix A on CD-ROM | |
| |
| |
| |
Mathematical Details and Special Topics | |
| |
| |
Geostatistical Data | |
| |
| |
| |
Estimating the Empirical Semivariogram | |
| |
| |
| |
Parametric Fitting of the Semivariogram | |
| |
| |
| |
Nonparametric Fitting of the Semivariogram | |
| |
| |
| |
Solutions to Kriging Equations | |
| |
| |
| |
Is Kriging Perfect Interpolation? | |
| |
| |
| |
Block and Indicator Kriging | |
| |
| |
Spatial Random Field Models | |
| |
| |
| |
Composite Likelihood in Spatial Generalized Linear Models | |
| |
| |
Lattice Data | |
| |
| |
| |
Maximum Likelihood Estimation in Lattice Models | |
| |
| |
| |
Are Autoregressive Models Stationary? | |
| |
| |
Point Patterns | |
| |
| |
| |
Estimating First- and Second-Order Properties of Point Patterns | |
| |
| |
| |
Point Process Models | |
| |
| |
| |
Spectral Analysis of Spatial Point Patterns | |
| |
| |
Supplementary Application | |
| |
| |
| |
Point Patterns--Spectral Analysis of Seedling Counts | |
| |
| |
Bibliography | |
| |
| |
Author Index | |
| |
| |
Subject Index | |