Basic Bivariate Techniques REBECCA M. WARNER © Applied Statistics I Third Edition 1% Pageiof 624 - Location 2 of 15772 Tomystudents:Past,present, andfuture. 1% Pageiiof624 + Location 4 of 15772 Sara Miller McCune founded SAGE Publishing in 1965 to support the dissemination ofusable knowledge and educatea global community. SAGE publishes more than 1000journals and over 800 new books each year, spanning a wide range of subject areas. Our growing selection oflibrary products includes archives, data, case studies and video. SAGE remains majority ownedby our founder and after her lifetime will become owned by a charitable trust that secures the company’s continued independence. Los Angeles | London| New Delhi| Singapore| Washington DC| Melbourne 1% Pageiiof 624 - Location 5 of15772 Applied Statistics I Basic Bivariate Techniques Third Edition Rebecca M. Warner ProfessorEmerita, University ofNewHampshire © SAGE Los Angeles | London | New Delhi Singapore | Washington DC | Melbourne Los Angeles London New Delhi Singapore Washington DC Melbourne 1% Радетof624 » Location 10of15772 SAGEPublications Asia-Pacific Pte. Ltd. Copyright © 2021 by SAGE Publications, Inc. 18 Cross Street #10-10/11/12 All rights reserved. Except as permitted by U.S. China SquareCentral copyright law, no part ofthis work may be reproduced ordistributed in any form or by any Singapore 048423 means,or stored in a database or retrieval system, ISBN 978-1-5063-5280-0 without permission in writing from the This bookis printed on acid-freepaper. publisher. 202122232410987654321 All third-party trademarks referenced or depicted herein are includedsolely for the purpose of Printed in theUnited States ofAmerica illustration and are the propertyoftheir AcquisitionsEditor: Helen Salmon respective owners. Reference to these trademarks in no way indicates any relationship with, or EditorialAssistant: Megan O'Heffernan endorsement by, the trademark owner. ContentDevelopment Editor: Chelsea Neve SPSS is a registered trademark of International Production Editor: LaureenGleason Business Machines Corporation. Excelis a Copy Editor: Jim Kelly registered trademark ofMicrosoft Corporation. All Excel screenshotsin this book are used with Typesetter: Hurix Digital permission from Microsoft Corporation. Proofreader: Scott Oney Indexer: Michael Ferreira OSAGE Cover Designer: Gail Buschman MarketingManager: Shari Countryman FOR INFORMATION: SAGEPublications, Inc. 2455 Teller Road Thousand Oaks, California 91320 E-mail: [email protected] SAGEPublicationsLtd. 1 Oliver's Yard 55 City Road London, EC1Y 1SP United Kingdom SAGEPublications India Pvt. Ltd. B 1/11 Mohan Cooperative IndustrialArea Mathura Road, New Delhi 110 044 India 1% Page ivof 624 » Location 14 of 15772 BriefContents Preface Acknowledgments Aboutthe Author Chapter1 - Evaluating Numerical Information Chapter2 - Basic Research Concepts Chapter3 - Frequency Distribution Tables Chapter4 - Descriptive Statistics Chapter5 - Graphs: Bar Charts, Histograms, and Boxplots Chapter6 - The Normal Distribution and z Scores Chapter7 - Sampling Error and Confidence Intervals Chapter8 - The One-Samplet Test: Introduction to Statistical Significance Tests Chapter9 Issuesin Significance Tests: Effect Size, Statistical Power, and Decision Errors Chapter 10 - Bivariate Pearson Correlation Chapter11 - Bivariate Regression Chapter 12 - The Independent-Samplest Test Chapter 13 - One-Way Between-Subjects Analysis ofVariance Chapter14 - Paired-Samplest Test Chapter 15 - One-Way Repeated-Measures Analysis ofVariance Chapter16 - FactorialAnalysis ofVariance Chapter 17 - Chi-Square Analysis of ContingencyTables Chapter18 - Selection ofBivariate Analyses and Review ofKey Concepts Appendices Glossary References Index 1% Pageviof624 » Location38of15772 1.8 Biases of Information Consumers 1.8.1 Confirmation Bias (A Detailed Contents Social Influence and Consensus Preface 1.9 Ethical Issues in Data Collection and Acknowledgments Analy: Aboutthe Author 1.9.1 Ethical Guidelines for Chapter1 - Evaluating Numerical Researchers: DataCollection Information 1.9.2 Ethical Guidelines for 1.1 Introduction Statisticians: Data Analysis and 1.2 Guidelines for Numeracy Reporting 1.3 SourceCredibility 1.10 Lying With Graphs and Statistics 1.3.1 Self-Interest or Bias 1.11 Degreesof Belief 1.3.2 Bias and “Cherry-Picking” 1.12 Summary 1.3.3 Primary, Secondary, and Chapter 2 - Basic Research Concepts Third-Party Sources 2.1 Introduction 1.3.4 Communicator Credentials 2.2 Types ofVariables andSkills 2.2.1 Overview 1.3.5 Track Record for Truth-Telling 2.2.2 Categorical Variables 1.4 Message Content 2.2.3 Quantitative Variables 1.4.1 Anecdotal Versus Numerical 2.2.4 Ordinal Variables Information 2.2.5 Variable Type and Choice of 1.4.2 Citation of Supporting Analysis Evidence 2.2.6 Rating Scale Variables 1.5 Evaluating Generalizability 2.2.7 Scores That Represent Counts 1.6 Making Causal Claims 2.3 Independent and Dependent 1.6.1 The “Post Hoc, Ergo Propter Variables Hoc”Fallacy 2.4 Typical Research Questions 1.6.2 Correlation (by Itself) Does Not 2.4.1 AreX and Correlated? Imply Causation 2.4.2 Does X Predict Y? 1.6.3 Perfect Correlation Versus 2.4.3 Does X Cause Y? Imperfect Correlation 2.5 Conditionsfor Causal Inference 1.6.4 “Individual Results Vary” 2.6 Experimental Research Design 1.6.5 Requirementsfor Evidence of 2.7 Nonexperimental Research Design Causal Inference 2.8 Quasi-Experimental Research 1.7 Quality Control Mechanisms in Designs Science 2.9 Other Issues in Design and Analysis 1.7.1 Peer Review 2.10 Choice ofStatistical Analysis 1.7.2 Replication and Accumulation Preview of Evidence 2.11 Populations and Samples: Ideal 1.7.3 Open Science and Study Versus Actual Situations Preregistration 1% Pagexix of624 » Location 61 of 15772. 2.11.1 Ideal Definition of Population 3.8.2 Evaluation of Score Location and Sample Using Cumulative Percentage 2.11.2 Two Real-World Research 3.8.3 Grouped or Binned Frequency Situations Similar to the Ideal Distributions Population and Sample Situation 3.9 Frequency Tables for Categorical 2.11.3 Actual Research Situations Versus Quantitative Variables That Are Not Similar to Ideal 3.10 Reporting Data Screening for Situations Quantitative Variables 2.12 Common Problems in 3.11 What We Hopeto See in Frequency Interpretation of Results Tables for Categorical Variables Appendix 2A: More About Levels of 3.11.1 Categorical Variables That Measurement Represent Naturally Occurring Appendix 2B: Justification for the Use of Groups Likert and Other Rating Scales as 3.11.2 Categorical Variables That Quantitative Variables (in Some Represent Treatment Groups Situations’ 3.12 What We Hopeto See in Frequency Chapter3 - Frequency Distribution Tables Tables for Quantitative Variables 3.1 Introduction 3.13 Summary 3.2 Use of Frequency Tablesfor Data Appendix 3A: Getting Startedin IBM Screening SPSS” Version 25 3.3 Frequency Tablesfor Categorical 3.A.1 The Bare Minimum: Using an Variables Existing SPSS Data File to Obtain, 3.4 Elementsof Frequency Tables Print, and Save Results 3.4.1 Frequency Counts (n or f 3.A.2 Moving Between Windowsin 3.4.2 Total Numberof Scores in a SPSS Sample (N 3.A.3 Creating a File and Entering 3.4.3 Missing Values (ifAny Data 3.4.4 Proportions 3.A.4 Defining Variable Names and 3.4.5 Percentages Properties ofVariables 3.4.6 Cumulative Frequencies or Appendix 3B: Missing Values in Cumulative Percentages Frequency Tables 3.5 Using SPSS to Obtain a Frequency Appendix 3C: Dividing Scores Into Table Groups or Bins 3.6 Mode, Impossible Score Values, and Chapter4 - Descriptive Statistics Missing Values 4.1 Introduction 3.7 Reporting Data Screening for 4.2 Questions About Quantitative Categorical Variables Variables 3.8 Frequency Tables for Quantitative 4.3 Notation Variables 4.4 Sample Median 3.8.1 Ungrouped Frequency 4.5 Sample Mean (M) Distribution 4.6 An Important Characteristic ofM: 1% Pagexix of624 - Location94 of 15772 The Sum ofDeviations From M = 0 5.1 Introduction 4.7 Disadvantage ofM: It Is Not Robust 5.2 Pie Chartsfor CategoricalVariables Against Influence ofExtreme Scores 5.3 Bar Charts for Frequencies of 4.8 Behavior ofMean, Median, andMode CategoricalVariables in Common Real-World Situations 5.4 GoodPractice for Construction ofBar 4.8.1 Example 1: Bell-Shaped Charts Distribution 5.5 Deceptive Bar Graphs 4.8.2 Example 2: Bimodal or 5.6 Histogramsfor Quantitative Polarized Distribution Variables 4.8.3 Example 3: Skewed 5.7 Obtaining a Histogram Using SPSS Distribution 5.8 Describing and Sketching Bell- 4.8.4 Example 4: No Clear Mode Shaped Distributions 4.9 ChoosingAmongMean, Median, and 5.9 GoodPractices in SettingUp Mode Histograms 4.10 Using SPSS to Obtain Descriptive 5.10 Boxplot (Box andWhiskers Plot: Statistics for a Quantitative Variable 5.10.1 How to Set Up a Boxplot by 4.11 Minimum, Maximum, and Range: Hand Variation Among Scores 5.10.2 How to Obtain a Boxplot 4.12 The Sample Variance s? Using SPSS 4.12.1 Step 1: Deviation ofEach 5.11 Telling Stories About Distributions Score From the Mean 5.12 UsesofGraphsin Actual Research 4.12.2 Step 2: Sum ofSquared 5.13 Data Screening: Separate Bar Charts Deviations or Histograms for Groups 4.12.3 Step 3: Degrees ofFreedom 5.14 Use ofBar Charts to Represent 4.12.4 Puttingthe Pieces Together: Group Means Computing a Sample Variance 5.15 Other Examples 4.13 Sample Standard Deviation (S or SD! 5.15.1 Scatterplots 4.14 How a Standard Deviation Describes 5.15.2 Maps Variation Among Scores in a Frequency 5.15.3 Historical Example Table 5.16 Summary 4.15 Why Is There Variance? Chapter6 - The Normal Distribution and z 4.16 Reports ofDescriptiveStatistics in Scores ournalArticles 6.1 Introduction 4.17 Additional Issues in Reporting 6.2 LocationsofIndividual Scores in Descriptive Statistics Normal Distributions 4.18 Summary 6.3 Standardized or z Scores Appendix 4A: OrderofArithmetic 6.3.1 First Step in Finding az Score Operations for X: The Distance ofX From M Appendix 4B: Rounding 6.3.2 Second Step: Divide the (X-M) Chapter 5 - Graphs: Bar Charts, Histograms, Distance by SD to Obtain aUnit-Free and Boxplots or Standardized Distance ofScore 1% Pagexix of624 » Location123of15772