Advanced statistics in criminology and criminal justice / David Weisburd, David B. Wilson, Alese Wooditch, and Chester Britt

Format:

Book

Author/Creator:

Weisburd, David, author.

Wilson, David B., 1961- author.

Wooditch, Alese, author.

Britt, Chester L., author.

Language:

English

Subjects (All):

Criminology.

Social sciences--Statistical methods.

Social sciences.

Physical Description:

1 online resource (552 pages)

Edition:

Fifth edition.

Place of Publication:

Cham, Switzerland : Springer, [2022]

Summary:

This book provides the student, researcher or practitioner with the tools to understand many of the most commonly used advanced statistical analysis tools in criminology and criminal justice, and also to apply them to research problems. The volume is structured around two main topics, giving the user flexibility to find what they need quickly. The first is "the general linear model" which is the main analytic approach used to understand what influences outcomes in crime and justice. It presents a series of approaches from OLS multivariate regression, through logistic regression and multi-nomial regression, hierarchical regression, to count regression. The volume also examines alternative methods for estimating unbiased outcomes that are becoming more common in criminology and criminal justice, including analyses of randomized experiments and propensity score matching. It also examines the problem of statistical power, and how it can be used to better design studies. Finally, it discusses meta analysis, which is used to summarize studies; and geographic statistical analysis, which allows us to take into account the ways in which geographies may influence our statistical conclusions.

Contents:

Intro

Contents

Chapter 1: Introduction

Proportionality Review and the Supreme Court of New Jersey: A Cautionary Tale

Generalized Linear Models

Special Topics

References

Chapter 2: Multiple Regression

Overview of Simple Regression

Extending Simple Regression to Multiple Regression

Assumptions of Multiple Regression

Independence

Normally Distributed Errors

Homoscedasticity of Errors

Linearity

Measurement Error in the Independent Variables

Regression Diagnostics

Dealing with Outliers and Influential Cases

Testing the Significance of Individual Regression Coefficients

Assessing Overall Model Fit and Comparing Nested Models

R2 and Adjusted R2

Comparing Regression Coefficients Within a Single Model: The Standardized Regression Coefficient

Correctly Specifying the Regression Model

Model Specification and Building

An Example of a Multiple Regression Model

Chapter Summary

Key Terms

Symbols and Formulas

Exercises

Computer Exercises

SPSS

Standardized Regression Coefficients (Betas)

F-Test for a Subset of Variables

Residual Plot

Stata

R

Problems

Chapter 3: Multiple Regression: Additional Topics

Nominal Variables with Three or More Categories in Multiple Regression

Nonlinear Relationships

Finding a Nonlinear Relationship: Graphical Assessment

Incorporating Nonlinear Relationships into an OLS Model Using a Quadratic Term of an Independent Variable

Interpreting Nonlinear Coefficients

Note on Statistical Significance

Transforming the Dependent Variable

Review of Nonlinear Relationships

Interaction Effects.

Interaction of a Dummy Variable and a Scaled Variable

An Example: Race and Punishment Severity

Interaction Effects Between Two Scaled Variables

An Example: Punishment Severity

The Problem of Multicollinearity

Dummy Coding Nominal Variables

Computing Nonlinear and Interaction Terms

Nonlinear Terms

Interaction Terms

Estimating the Regression Model

Collinearity Diagnostics

Chapter 4: Logistic Regression

Why Is It Inappropriate to Use OLS Regression for a Dichotomous Dependent Variable?

Logistic Regression

A Substantive Example: Adoption of Compstat in U.S. Police Agencies

Interpreting Logistic Regression Coefficients

The Odds Ratio

The Derivative at Mean

Comparing Logistic Regression Coefficients

Using Probability Estimates to Compare Coefficients

Standardized Logistic Regression Coefficients

Evaluating the Logistic Regression Model

Percent of Correct Predictions

Pseudo-R2

Statistical Significance in Logistic Regression

Chapter 5: Multiple Regression with Multiple Category Nominal or Ordinal Measures

Multinomial Logistic Regression

A Substantive Example: Case Dispositions in California

The Missing Set of Coefficients

Statistical Inference.

Single Coefficients

Multiple Coefficients

Overall Model

A Concluding Observation About Multinomial Logistic Regression Models

Ordinal Logistic Regression

Interpretation of Ordinal Logistic Regression Coefficients

Substantive Example: Severity of Punishment Decisions

Interpreting the Coefficients

Statistical Significance

Parallel Slopes Tests

Score Test

Brant Test

Partial Proportional Odds

Severity of Punishment Example

Formulas

Chapter 6: Count-Based Regression Models

The Poisson Distribution

Poisson Regression

Incident Rate Ratios (IRRs)

Significance Testing

Exposure and Offsets

An Example: California 1999 Uniform Crime Report Data

Over-Dispersion in Count Data

Quasi-Poisson and Negative Binomial Regression

An Example: Reanalysis of the California 1999 Uniform Crime Report Data

Zero-Inflated Poisson and Negative Binomial Regression

Quasi-Poisson Regression

Negative Binomial Regression

Zero-Inflated Poisson/Negative Binomial Regression

Chapter 7: Multilevel Regression Models.

A Simple Multilevel Model

Fixed-Effects and Random-Effects

A Substantive Example: Bail Decision-Making Study

Intraclass Correlation and Explained Variance

Deciding Between and Fixed- and Random-Effects Model

Bail Decision-Making Study

Random Intercept Model with Fixed Slopes

Centering Independent Variables

Between and Within Effects

Testing for Between and Within Effects

Random Coefficient Model

Variance Estimates

Adding Cluster (Level 2) Characteristics

A Substantive Example: Race and Sentencing Across Pennsylvania Counties

Multilevel Negative Binomial Regression

Random Intercept Models

Random Coefficient Models

Chapter 8: Statistical Power

Statistical Power

Setting the Level of Statistical Power

Components of Statistical Power

Statistical Significance and Statistical Power

Directional Hypotheses

Sample Size and Statistical Power

Effect Size and Statistical Power

Estimating Statistical Power and Sample Size for a Statistically Powerful Study

Difference of Means Test

ANOVA

Correlation

Least Squares Regression

Summing Up: Avoiding Studies Designed for Failure

Two-Sample Difference of Means Test

OLS Regression

Chapter 9: Randomized Experiments

The Structure of a Randomized Experiment.

The Main Advantage of Experiments: Isolating Causal Effects

Internal Validity

Selected Design Types and Associated Statistical Methods

The Two-Group Randomized Design

Three or More Group Randomized Design

Factorial Design

Two-Way ANOVA for Between-Subjects Designs

An Example: Perceptions of Children During a Police Interrogation

Mixed Within- and Between-Subjects Factorial Designs

Block Randomized Designs

Block Randomization and Statistical Power

Examining Interaction in a Block Randomized Experiment

Using Covariates to Increase Statistical Power in Experimental Studies

Independent Sample t-Test

One-Way ANOVA

Two-Way Factorial (Type I SS)

Two-Way Factorial (Type II SS)

Two-Way Factorial (Type III SS)

Chapter 10: Propensity Score Matching

The Underlying Logic Behind Propensity Score Matching

Selection of Model for Predicting Propensity for Treatment

Matching Methods

The Case of Work Release in Prison: A Substantive Example

Assessing the Quality of the Matches

Sensitivity Analysis for Average Treatment Effects

Limitations of Propensity Score Matching

Estimating Propensity Score

Matching Cases

Assessing Matches

Estimating Treatment Effect

Problems.

Notes:

Includes bibliographical references and index.

Description based on print version record.

ISBN:

3-030-67738-9

OCLC:

1283851607