Scaling Procedures : Issues and Applications

Acknowledgments xi Foreword xiii 1 Introduction and Overview 1(17) Purpose of the Book 1(1) Perspectives on Measurement in the Social Sciences 2(5) What Is Measurement? 2(1) Usefulness of Social Science Measures 3(1) Scaling of Latent Psychological Constructs With Multiple Items 4(2) Recent Trends in Scaling Latent Psychological Constructs 6(1) Latent Constructs 7(2) Theory and Validity 7(1) Importance of the Literature Review 8(1) Overview of Dimensionality, Reliability, and Validity 9(5) Dimensionality 9(1) Reliability 10(1) Construct Validity 11(3) Overview of Recommended Procedures and 14(3) Steps in Scale Development Step 1: Construct Definition and Content Domain 16(1) Step 2: Generating and Judging Measurement Items 16(1) Step 3: Designing and Conducting Studies to Develop and Refine the Scale 16(1) Step 4: Finalizing the Scale 16(30) Summary and Preview of the Text 17(1) 2 Dimensionality 18(23) Introduction 18(1) Dimensionality of a Construct, Items, and a Set of Items 19(4) Unidimensionality 19(2) Multidimensionality 21(6) Does Unidimensionality of a Set of Items Imply Unidimensionality of the Items or the Construct? 23(3) Relevance of Unidimensionality 26(1) How to Assess Dimensionality of Constructs 27(12) Exploratory Factor Analysis (EFA) 27(9) Confirmatory Factor Analysis (CFA) 36(5) Summary 39(1) Appendix 2A: Computing Partial Correlations 40(1) 3 Reliability 41(30) Introduction 41(1) The True-Score Model 41(2) Relationship Between the True-Score Model and the Factor Model 43(1) Types of Reliability 43(6) Test-Retest Reliability 44(1) Alternative-Form Reliability 45(1) Internal Consistency Methods 46(3) Split-Half Reliability 47(16) Coefficient Alpha ((x) 49(10) Coefficient Alpha Example 53(1) Coefficient Alpha and Dimensionality 54(3) Coefficient Alpha, Scale Length, Interitem Correlation, and Item Redundancy 57(2) Generalizability Theory 59(10) G-Study Illustration 63(5) Generalizability Coefficient 65(30) Decision Studies 68(4) Summary 69(2) 4 Validity 71(17) Overview of Construct Validity 71(1) Translation Validity 72(4) Face Validity 72(1) Content Validity 73(3) Criterion Validity 76(6) Predictive and Post-dictive Validity 76(1) Concurrent Validity 76(1) Convergent and Discriminant Validity 77(3) Known-Group Validity 80(9) Nomological Validity 82(1) Socially Desirable Response Bias 83(2) Summary 85(3) 5 Steps 1 and 2: Construct Definition and Generating and Judging Measurement Items 88(27) Introduction 88(1) Step 1: Construct Definition and Content Domain 89(5) The Importance of Clear Construct Definition 89(1) The Importance of Theory, Literature Review, and Judging 90(2) The Focus on Effeet Items Versus Formative Indicators 92(1) The Role of Theory in Specifying Dimensionality 93(1) Step 2: Generating and Judging Measurement Items 94(9) Domain Sampling 94(1) Generating an Item Pool 95(7) Item Sources 96(1) Item Writing 97(4) Number of Items in the Initial Pool 101(43) Judging Items for Content and Face Validity 102(1) Guidelines for Content and Face Validation 102(13) Applications of Steps 1 and 2 103(4) Summary 107(1) Appendix 5A 108(6) Appendix 5B 114(1) 6 Step 3: Designing and Conducting Studies to Develop a Scale 115(28) Introduction 115(1) Pilot Testing 115(3) Example of a Pilot Test 117(4) Conducting Multiple Studies for Initial Development and Validation 118(1) Including Constructs for Assessing Validity 118(3) Initial Item Analysis: Exploratory Factor Analysis (EFA) 121(1) EFA Options 121(4) Number of Factors to Extract 122(2) Rotational Methods 124(1) Item Retention via EFA 125(2) Initial Item and Reliability Analysis 125(1) A Final Caveat 126(1) EFA and Item and Reliability Analyses Examples From the Literature 127(14) An Empirical Example of EFA 127(4) Reliability and Item-Based Statistics 131(12) Summary 141(2) 7 Step 4: Finalizing a Scale 143(41) Introduction 143(1) EFA and Additional Item Analyses 143(4) EFA 144(1) Additional Item Analyses 144(3) Item-to-Total Correlations 144(1) Interitem Correlations 145(1) Scale Length Considerations 145(5) Confirmatory Factor Analysis (CFA) 147(14) Overview 147(1) Three Key Issues to Consider 148(2) Evaluating CFA Models 150(7) Model Convergence and an "Acceptable Range" of Parameter Estimates 150(1) Fit Indices 151(1) Significance of Parameter Estimates and Related Diagnostics 152(2) Standardized Residuals and Modification Indices 154(1) Measurement Invariance Across Samples 155(2) CFA Examples From the Literature 157(1) An Empirical Example of Evaluating a CFA Model 158(3) Additional Evaluations of Validity 161(3) Convergent and Discriminant Validity 162(1) Predictive and Concurrent Validity 163(1) Establishing Norms 164(2) Means and Standard Deviations 164(2) Known-Group Differences 166(1) Applying Generalizability Theory 166(3) Summary 169(2) Appendix 7A: LISREL Input and Output for Small Business Owners and Real Estate Salespeople Samples 171(11) Appendix 7B: LISREL Input for Multigroup Analyses: Full Measurement Invariance Model 182(2) 8 Concluding Remarks 184 References 187(8) Author Index 195(4) Subject Index 199(6) About the Authors 205