Statistical and econometric methods for transportation data analysis.

Cover
Half Title
Series Page
Title Page
Copyright Page
Table of Contents
Preface
Authors
Section I: Fundamentals
	1: Statistical Inference I: Descriptive Statistics
		1.1 Measures of Relative Standing
		1.2 Measures of Central Tendency
		1.3 Measures of Variability
		1.4 Skewness and Kurtosis
		1.5 Measures of Association
		1.6 Properties of Estimators
			1.6.1 Unbiasedness
			1.6.2 Efficiency
			1.6.3 Consistency
			1.6.4 Sufficiency
		1.7 Methods of Displaying Data
			1.7.1 Histograms
			1.7.2 Ogives
			1.7.3 Box Plots
			1.7.4 Scatter Diagrams
			1.7.5 Bar and Line Charts
	2: Statistical Inference II: Interval Estimation, Hypothesis Testing, and Population Comparisons Descriptive Statistics
		2.1 Confidence Intervals
			2.1.1 Confidence Interval for µ with Known σ2
			2.1.2 Confidence Interval for the Mean with Unknown Variance
			2.1.3 Confidence Interval for a Population Proportion
			2.1.4 Confidence Interval for the Population Variance
		2.2 Hypothesis Testing
			2.2.1 Mechanics of Hypothesis Testing
			2.2.2 Formulating One- and Two-Tailed Hypothesis Tests
			2.2.3 The p-Value of a Hypothesis Test
		2.3 Inferences Regarding a Single Population
			2.3.1 Testing the Population Mean with Unknown Variance
			2.3.2 Testing the Population Variance
			2.3.3 Testing for a Population Proportion
		2.4 Comparing Two Populations
			2.4.1 Testing Differences between Two Means: Independent Samples
			2.4.2 Testing Differences between Two Means: Paired Observations
			2.4.3 Testing Differences between Two Population Proportions
			2.4.4 Testing the Equality of Two Population Variances
		2.5 Nonparametric Methods
			2.5.1 The Sign Test
			2.5.2 The Median Test
			2.5.3 The Mann–Whitney U Test
			2.5.4 The Wilcoxon-Signed Rank Test for Matched Pairs
			2.5.5 The Kruskal–Wallis Test
			2.5.6 The Chi-Square Goodness-of-Fit Test
Section II: Continuous Dependent Variable Models
	3: Linear Regression
		3.1 Assumptions of the Linear Regression Model
			3.1.1 Continuous Dependent Variable Y
			3.1.2 Linear-in-Parameters Relationship between Y and X
			3.1.3 Observations Independently and Randomly Sampled
			3.1.4 Uncertain Relationship between Variables
			3.1.5 Disturbance Term Independent of X and Expected Value Zero
			3.1.6 Disturbance Terms Not Autocorrelated
			3.1.7 Regressors and Disturbances Uncorrelated
			3.1.8 Disturbances Approximately Normally Distributed
			3.1.9 Summary
		3.2 Regression Fundamentals
			3.2.1 Least Squares Estimation
			3.2.2 Maximum Likelihood Estimation
			3.2.3 Properties of OLS and MLE Estimators
			3.2.4 Inference in Regression Analysis
		3.3 Manipulating Variables in Regression
			3.3.1 Standardized Regression Models
			3.3.2 Transformations
			3.3.3 Indicator Variables
		3.4 Estimate a Single Beta Parameter
		3.5 Estimate Beta Parameter for Ranges of the Variable
		3.6 Estimate a Single Beta Parameter for m - 1 of the m Levels of the Variable
			3.6.1 Interactions in Regression Models
		3.7 Checking Regression Assumptions
			3.7.1 Linearity
			3.7.2 Homoscedastic Disturbances
			3.7.3 Uncorrelated Disturbances
			3.7.4 Exogenous Independent Variables
			3.7.5 Normally Distributed Disturbances
		3.8 Regression Outliers
			3.8.1 The Hat Matrix for Identifying Outlying Observations
			3.8.2 Standard Measures for Quantifying Outlier Influence
			3.8.3 Removing Influential Data Points from the Regression
		3.9 Regression Model Goodness-of-Fit Measures
		3.10 Multicollinearity in the Regression
		3.11 Regression Model-Building Strategies
			3.11.1 Stepwise Regression
			3.11.2 Best Subsets Regression
			3.11.3 Iteratively Specified Tree-Based Regression
		3.12 Estimating Elasticities
		3.13 Censored Dependent Variables—Tobit Model
		3.14 Box–Cox Regression
	4: Violations of Regression Assumptions
		4.1 Zero Mean of the Disturbances Assumption
		4.2 Normality of the Disturbances Assumption
		4.3 Uncorrelatedness of Regressors and Disturbances Assumption
		4.4 Homoscedasticity of the Disturbances Assumption
			4.4.1 Detecting Heteroscedasticity
			4.4.2 Correcting for Heteroscedasticity
		4.5 No Serial Correlation in the Disturbances Assumption
			4.5.1 Detecting Serial Correlation
			4.5.2 Correcting for Serial Correlation
		4.6 Model Specification Errors
	5: Simultaneous Equation Models
		5.1 Overview of the Simultaneous Equations Problem
		5.2 Reduced Form and the Identification Problem
		5.3 Simultaneous Equation Estimation
			5.3.1 Single Equation Methods
			5.3.2 System Equation Methods
		5.4 Seemingly Unrelated Equations
		5.5 Applications of Simultaneous Equations to Transportation Data
		Appendix 5A: A Note on Generalized Least Squares Estimation
	6: Panel Data Analysis
		6.1 Issues in Panel Data Analysis
		6.2 One-Way Error Component Models
			6.2.1 Heteroscedasticity and Serial Correlation
		6.3 Two-Way Error Component Models
		6.4 Variable Parameter Models
		6.5 Additional Topics and Extensions
	7: Background and Exploration in Time Series
		7.1 Exploring a Time Series
			7.1.1 The Trend Component
			7.1.2 The Seasonal Component
			7.1.3 The Irregular (Random) Component
			7.1.4 Filtering of Time Series
			7.1.5 Curve Fitting
			7.1.6 Linear Filters and Simple Moving Averages
			7.1.7 Exponential Smoothing Filters
			7.1.8 Difference Filter
		7.2 Basic Concepts: Stationarity and Dependence
			7.2.1 Stationarity
			7.2.2 Dependence
			7.2.3 Addressing Nonstationarity
			7.2.4 Differencing and Unit-Root Testing
			7.2.5 Fractional Integration and Long Memory
		7.3 Time Series in Regression
			7.3.1 Serial Correlation
			7.3.2 Dynamic Dependence
			7.3.3 Volatility
			7.3.4 Spurious Regression and Cointegration
			7.3.5 Causality
	8: Forecasting in Time Series: Autoregressive Integrated Moving Average (ARIMA) Models and Extensions
		8.1 Autoregressive Integrated Moving Average Models
		8.2 The Box–Jenkins Approach
			8.2.1 Order Selection
			8.2.2 Parameter Estimation
			8.2.3 Diagnostic Checking
			8.2.4 Forecasting
		8.3 Autoregressive Integrated Moving Average Model Extensions
			8.3.1 Random Parameter Autoregressive (RPA) Models
			8.3.2 Stochastic Volatility (SV) Models
			8.3.3 Autoregressive Conditional Duration (ACD) Models
			8.3.4 Integer-Valued ARMA (INARMA) Models
		8.4 Multivariate Models
		8.5 Nonlinear Models
			8.5.1 Testing for Nonlinearity
			8.5.2 Bilinear Models
			8.5.3 Threshold Autoregressive Models
			8.5.4 Functional Parameter Autoregressive Models
			8.5.5 Neural Networks
	9: Latent Variable Models
		9.1 Principal Components Analysis
		9.2 Factor Analysis
		9.3 Structural Equation Modeling
			9.3.1 Basic Concepts in Structural Equation Modeling
			9.3.2 Fundamentals of Structural Equation Modeling
			9.3.3 Nonideal Conditions in the Structural Equation Model
			9.3.4 Model Goodness-of-Fit Measures
			9.3.5 Guidelines for Structural Equation Modeling
	10: Duration Models
		10.1 Hazard-Based Duration Models
		10.2 Characteristics of Duration Data
		10.3 Nonparametric Models
		10.4 Semi-Parametric Models
		10.5 Fully Parametric Models
		10.6 Comparisons of Nonparametric, Semi-Parametric, and Fully Parametric Models
		10.7 Heterogeneity
		10.8 State Dependence
		10.9 Time-Varying Explanatory Variables
		10.10 Discrete-Time Hazard Models
		10.11 Competing Risk Models
Section III: Count and Discrete-Dependent Variable Models
	11: Count Data Models
		11.1 Poisson Regression Model
		11.2 Interpretation of Variables in the Poisson Regression Model
		11.3 Poisson Regression Model Goodness-of-Fit Measures
		11.4 Truncated Poisson Regression Model
		11.5 Negative Binomial Regression Model
		11.6 Zero-Inflated Poisson and Negative Binomial Regression Models
		11.7 Random Effects Count Models
	12: Logistic Regression
		12.1 Principles of Logistic Regression
		12.2 The Logistic Regression Model
	13: Discrete Outcome Models
		13.1 Models of Discrete Data
		13.2 Binary and Multinomial Probit Models
		13.3 Multinomial Logit Model
		13.4 Discrete Data and Utility Theory
		13.5 Properties and Estimation of Multinomial Logit Models
			13.5.1 Statistical Evaluation
			13.5.2 Interpretation of Findings
			13.5.3 Specification Errors
			13.5.4 Data Sampling
			13.5.5 Forecasting and Aggregation Bias
			13.5.6 Transferability
		13.6 The Nested Logit Model (Generalized Extreme Value Models)
		13.7 Special Properties of Logit Models
	14: Ordered Probability Models
		14.1 Models for Ordered Discrete Data
		14.2 Ordered Probability Models with Random Effects
		14.3 Limitations of Ordered Probability Models
	15: Discrete/Continuous Models
		15.1 Overview of the Discrete/Continuous Modeling Problem
		15.2 Econometric Corrections: Instrumental Variables and Expected Value Method
		15.3 Econometric Corrections: Selectivity-Bias Correction Term
		15.4 Discrete/Continuous Model Structures
		15.5 Transportation Application of Discrete/Continuous Model Structures
Section IV: Other Statistical Methods
	16: Random Parameters Models
		16.1 Random Parameters Multinomial Logit Model (Mixed Logit Model)
		16.2 Random Parameters Ordered Probability Models
		16.3 Random Parameters Count Models
		16.4 Random Parameters Duration Models
		16.5 Random Parameters Linear Regression Models
		16.6 Random Parameters with Heterogeneity in Means and Variances
		16.7 Grouped Random Parameters Models
		16.8 Random Thresholds Random Parameters Hierarchical Ordered Probit (HOPIT) Model
		16.9 Zero-Inflated Hierarchical Ordered Probit (HOPIT) Model
		16.10 Correlated Random Parameters Ordered Probit Model
		16.11 Correlated Random Parameters Logit Model
		16.12 Correlated Grouped Random Parameters Binary Logit Model
		16.13 Correlated Grouped Random Parameters Hazard-Based Duration Model
		16.14 Practical Aspects of Random Parameters Model Estimation
	17: Latent Class (Finite Mixture) Models
		17.1 Latent Class Multinomial Logit Model
		17.2 Grouped and Ungrouped Latent Class Ordered Probit Models with and without Class Probability Functions
	18: Bivariate and Multivariate Dependent Variable Models
		18.1 Bivariate Ordered Probit
		18.2 Bivariate Binary Probit
		18.3 Multivariate Binary Probit
		18.4 Simultaneous Estimation of Discrete Outcome and Continuous Dependent Variable Equations with Unrestricted Instruments: An Extension of the Maddala Model
	19: Bayesian Statistical Methods
		19.1 Bayes’ Theorem
		19.2 Markov Chain Monte Carlo (MCMC) Sampling-Based Estimation
		19.3 Flexibility of Bayesian Statistical Models via MCMC Sampling-Based Estimation
		19.4 Convergence and Identifiability Issues with MCMC Bayesian Models
		19.5 Goodness of Fit, Sensitivity Analysis, and Model Selection Criterion Using MCMC Bayesian Models
Appendix A: Statistical Fundamentals
Appendix B: Statistical Tables
Appendix C: Variable Transformations
References
Index