A Stata® Companion to Political Analysis

Philip H. Pollock III is a professor of political science at the University of Central Florida. He has taught courses in research methods at the undergraduate and graduate levels for more than thirty years. His main research interests are American public opinion, voting behavior, techniques of quantitative analysis, and the scholarship of teaching and learning. His recent research has been on the effectiveness of Internet-based instruction. Pollock’s research has appeared in the American Journal of Political Science, Social Science Quarterly, and the British Journal of Political Science. Recent scholarly publications include articles in Political Research Quarterly, the Journal of Political Science Education, and PS: Political Science and Politics.Barry C. Edwards writes textbooks and works for Fair Trial Analysis, LLC, a company that conducts research on juries and jurors for civil and criminal litigation. He received his B.A. from Stanford University, a J.D. from New York University, and a Ph.D. from the University of Georgia. He taught survey design and analysis, research methods, and prelaw courses at the University of Central Florida and continues to teach occasional courses for the University of Georgia. His political science interests include American politics, public law, and research methods. He founded the Political Science Data Group and created the PoliSciData.com website. His research has been published in American Politics Research, Congress & the Presidency, Election Law Journal, Emory Law Journal, Georgia Bar Journal, Harvard Negotiation Law Review, Journal of Politics, NYU Journal of Legislation and Public Policy, Political Research Quarterly, Presidential Studies Quarterly, Public Management Review, State Politics and Policy Quarterly, and UCLA Criminal Justice Law Review.

• Figures and TablesPrefaceIntroduction: Getting Started with StataI.1Datasets for Stata CompanionI.2A Quick Tour of StataI.3Running Commands in StataI.4Quick Access to Tutorials and ResourcesChapter 1 Using Stata for Data Analysis1.1General Syntax of Stata Commands1.2Using Stata’s Graphic User Interface Effectively1.3Do-files1.4Printing Results and Copying Output1.5Customizing Your Display1.6Log Files1.7Getting HelpChapter 1 ExercisesChapter 2 Descriptive Statistics2.1Identifying Levels of Measurement2.2Describing Nominal VariablesA Closer Look: Weighted and Unweighted Analysis: What’s the Difference?2.3Describing Ordinal Variables2.4Bar Charts for Nominal and Ordinal Variables2.5Describing Interval VariablesA Closer Look: Stata’s Graphics Editor2.6Histograms for Interval Variables2.7Obtaining Case-Level InformationChapter 2 ExercisesChapter 3 Transforming Variables3.1Creating Dummy Variables3.2Applying Math Operators to Variables3.3Managing Variable Descriptions and Labels3.4Collapsing Variables into Simplified Categories3.5Centering or Standardizing a Numeric Variable3.6Creating an Additive IndexChapter 3 ExercisesChapter 4 Making Comparisons4.1Cross-Tabulation AnalysisA Closer Look: The replace Command4.2Mean Comparison AnalysisA Closer Look: The format Command4.3Making Comparisons with Interval-Level Independent VariablesChapter 4 ExercisesChapter 5 Graphing Relationships and Describing Patterns5.1Graphs for Binary Dependent Variables5.1.1Simple Bar Charts with Nominal-Level Independent Variables5.1.2 Area Chart with Ordinal-Level Independent Variables5.1.3Graphs with Interval-Level Independent Variables5.2Graphs for Nominal-Level Dependent Variables5.2.1Clustered Bar Charts with Nominal-Level Independent Variables5.2.2 Multiple Line Plots with Ordinal-Level Independent Variables5.2.3Graphs with Interval-Level Independent Variables5.3Graphs for Ordinal-Level Dependent Variables5.3.1Using Bars to Represent Select Values5.3.2Stacked Bar Chart for Ordinal-Ordinal Relationship5.3.3Graphs with Interval-Level Independent Variables5.4Graphs for Interval-Level Dependent Variables5.4.1Plotting Means with Bars or Lines5.4.2Box Plots5.4.3ScatterplotsChapter 5 ExercisesChapter 6 Random Assignment and Sampling6.1Random Assignment6.1.1Two Groups with Equal Probability6.1.2Multiple Groups with Varying Probabilities6.1.3Random Assignment to Predetermined-Size Groups6.2Analyzing the Results of an Experiment6.2.1Assessing Random Assignment6.2.2Evaluating the Effect of Treatment6.3Random Sampling6.3.1Simple Random Sampling with Replacement6.3.2Simple Random Sampling without Replacement6.3.3Systematic Random Samples6.3.4Clustered and Stratified Random Samples6.4Selecting Cases for Qualitative Analysis6.4.1Most Similar Systems6.4.2Most Different Systems6.5Analyzing Data Ethically6.5.1Ethical Issues in Data Analysis6.5.2Ten Tips for Writing Replication CodeChapter 6 ExercisesChapter 7 Making Controlled Comparisons7.1Cross-Tabulation Analysis with a Control Variable7.1.1Start with a Basic Cross-Tabulation7.1.2Controlling for Another Variable7.1.3Interpreting Controlled Cross-TabulationsA Closer Look: The If Qualifier7.2Visualizing Controlled Comparisons with Categorical Dependent Variables7.3Mean Comparison Analysis with a Control Variable7.3.1Start with Basic Mean Comparison Table7.3.2Adding Control Variables7.3.3Interpreting a Controlled Mean Comparison7.4Visualizing Controlled Mean ComparisonsChapter 7 ExercisesChapter 8 Foundations of Inference8.1Estimating Population Parameters with Simulations8.2Expected Shape of Sampling Distributions8.2.1Central Limit Theorem and the Normal Distribution8.2.2Normal Distribution of Sample Proportions8.2.3Normal Distribution of Sample Means8.2.4The Standard Normal Distribution8.2.5The Empirical Rule (68-95-99 Rule)8.3Confidence Interval and Margins of Error8.3.1Critical Values for Confidence Intervals8.3.2Reporting the Confidence Interval for a Sample Proportions8.3.2Reporting the Confidence Interval for a Sample Means8.4Student’s t-Distribution: When You’re Not Completely Normal8.4.1The t-Distribution’s Role in Inferential Statistics8.4.1Critical Values of t-DistributionsChapter 8 ExercisesChapter 9 Hypothesis Tests with One and Two Samples9.1Role of the Null Hypothesis9.2Testing Hypotheses with Sample Proportions9.2.1Testing One Sample Proportion Against Hypothesized Value9.2.2Testing Difference Between Two Sample Proportions Using Groups9.2.3 Testing Difference Between Two Sample Proportions Using Variables9.2.4Testing Hypotheses about Proportions with Weighted Data9.3Testing Hypotheses with Sample Means9.3.1Testing One Sample Mean Against Hypothesized Value9.3.2Testing the Difference Between Two Sample Means Using Groups9.3.3Testing the Difference Between Two Sample Means Using Variables9.3.4T-Test Variations from Assumptions about Variance9.3.5Testing Hypotheses about Means with Weighted DataChapter 9 ExercisesChapter 10 Chi-Square Test and Analysis of Variance10.1The Chi-Square Test of Independence10.1.1How the Chi-Square Test Works10.1.2Conducting a Chi-Square Test10.1.3Example with Nominal-Level Independent VariableA Closer Look: Chi-Square Test with Weighted Data10.1.4Reporting and Interpreting ResultsA Closer Look: Other Applications of Chi-Square Tests10.2Measuring the Strength of Association between Categorical Variables10.2.1Lambda10.2.2Somers’ D10.2.3Cramer’s V10.3Chi-Square Test and Measures of Association in Controlled Comparisons10.3.1Analyzing an Ordinal-Level Relationship with a Control Variable10.3.2 Analyzing a Nominal-Level Relationship with a Control Variable (and Weighted Observations)10.4Analysis of Variance (ANOVA)10.4.1How ANOVA Works10.4.2 Single Factor ANOVA10.4.3 Two Factor ANOVA10.4.4 Stata’s F-Distribution FunctionsChapter 10 ExercisesChapter 11 Correlation and Bivariate Regression11.1Correlation Analysis11.1.1Correlation between Two Variables11.1.2Correlation Among More than Two VariablesA Closer Look: Other Types and Application of Correlation Analysis11.2Bivariate Regression AnalysisA Closer Look: Treating Census as a SampleA Closer Look: R-Squared and Adjusted R-Squared: What’s the Difference?11.3Creating a Scatterplot with a Linear Prediction LineA Closer Look: Creating Graphs with Multiple Layered ElementsA Closer Look: What If a Scatterplot Doesn’t Show a Linear Relationship?11.4Correlation and Bivariate Regression Analysis with Weighted DataA Closer Look: Creating Tables of Regression ResultsChapter 11 ExercisesChapter 12 Multiple Regression12.1Multiple Regression Analysis12.1.1Estimating and Interpreting a Multiple Regression Model12.1.2Visualizing Multiple Regression with Bubble Plots12.1.3Multiple Regression with Weighted Observations12.2Regression with Multiple Dummy Variables12.2.1Estimating and Interpreting Regression with Multiple Dummy Variables12.2.2Changing the Reference Category12.2.3Visualizing Regression with Multiple Dummy Variables12.3Interaction Effects in Multiple Regression12.3.1Estimating Regression Model with Interaction Term12.3.2Graphing Linear Prediction Lines for Interaction RelationshipsChapter 12 ExercisesChapter 13 Analyzing Regression Residuals13.1Expected Values, Observed Values, and Regression Residuals13.1.1Example from Bivariate Regression Analysis13.1.2Residuals from Multiple Regression Analysis13.2Squared and Standarized Residuals13.2.1Squared Residuals13.2.2Standardized Residuals13.3Assumptions about Regression Residuals13.4Analyzing Graphs of Regression Residuals13.4.1Histogram of Regression Residuals13.4.2Residual Diagnostic Plots13.5Testing Regression Assumptions with Residuals13.5.1 Testing Assumption that Residuals are Normally Distributed13.5.2Testing the Constant Variance Assumption15.3.3 Regression Diagnostics for Multiple Regression AnalysisA Closer Look: Other Regression Diagnostic Tests13.6What If You Diagnose Problems with Residuals?Chapter 13 ExercisesChapter 14 Logistic Regression14.1Odds, Logged Odds, and Probabilities14.2Estimating Logistic Regression Models14.2.1Logistic Regression with One Independent Variable14.2.2Reporting and Interpreting Odds Ratios14.2.3Evaluating Model FitA Closer Look: Logistic Regression Analysis with Weighted Observations14.3Logistic Regression with Multiple Independent Variables14.4Graphing Predicted Probabilities with One Independent Variable14.4.1Interval-Level Independent Variable14.4.2Categorical Independent VariableA Closer Look: Marginal Effects and Expected Changes in Probability14.5Graphing Predicted Probabilities with Multiple Independent Variables14.5.1One Categorical and One Interval-Level Independent Variable14.5.2Two Categorical Independent VariablesA Closer Look: Stata’s Quiet Mode14.5.3 Plotting Predicted Probabilities with atmeans Option14.5.4 Combining atmeans and over OptionsChapter 14 ExercisesChapter 15 Doing Your Own Political Analysis15.1Doable Research Ideas15.1.1Political Knowledge and Interest15.1.2 Self-Interest and Policy Preferences15.1.3 Economic Performance and Election Outcomes15.1.4Electoral Turnout in Comparative Perspective15.1.5 Correlates of State Policies15.1.6 Religion and Politics15.1.7 Race and Politics15.2Getting Data into Stata15.2.1 Opening Stata Formatted Datasets15.2.2 Importing Microsoft Excel Datasets15.2.3 Using HTML Table Data15.2.4Entering Data with Stata’s Data Editor15.3Writing It Up15.3.1The Research Question15.3.2 Previous Research15.3.3Data, Hypotheses, and Analysis15.3.4Conclusions and ImplicationsChapter 15 ExercisesAppendixTable A-1: Variables in the Debate Dataset in Alphabetical OrderTable A-2: Variables in the GSS Dataset in Alphabetical OrderTable A-3: Variables in the NES Dataset in Alphabetical OrderTable A-4: Variables in the States Dataset by TopicTable A-5: Variables in the World Dataset by Topic