A Step-by-Step Approach to Using SAS for Factor Analysis and Structural Equation Modeling, Second Edition, 2nd Edition

Book description

This easy-to-understand guide makes SEM accessible to all users. This second edition contains new material on sample-size estimation for path analysis and structural equation modeling. In a single user-friendly volume, students and researchers will find all the information they need in order to master SAS basics before moving on to factor analysis, path analysis, and other advanced statistical procedures.

Table of contents

  1. About This Book
  2. Acknowledgments from the First Edition
  3. Chapter 1: Principal Component Analysis
    1. Introduction: The Basics of Principal Component Analysis
      1. A Variable Reduction Procedure
      2. An Illustration of Variable Redundancy
      3. What Is a Principal Component?
      4. Principal Component Analysis Is Not Factor Analysis
    2. Example: Analysis of the Prosocial Orientation Inventory
      1. Preparing a Multiple-Item Instrument
      2. Number of Items per Component
      3. Minimal Sample Size Requirements
    3. SAS Program and Output
      1. Writing the SAS Program
      2. Results from the Output
    4. Steps in Conducting Principal Component Analysis
      1. Step 1: Initial Extraction of the Components
      2. Step 2: Determining the Number of “Meaningful” Components to Retain
      3. Step 3: Rotation to a Final Solution
      4. Step 4: Interpreting the Rotated Solution
      5. Step 5: Creating Factor Scores or Factor-Based Scores
      6. Step 6: Summarizing the Results in a Table
      7. Step 7: Preparing a Formal Description of the Results for a Paper
    5. An Example with Three Retained Components
      1. The Questionnaire
      2. Writing the Program
      3. Results of the Initial Analysis
      4. Results of the Second Analysis
    6. Conclusion
    7. Appendix: Assumptions Underlying Principal Component Analysis
    8. References
  4. Chapter 2: Exploratory Factor Analysis
    1. Introduction: When Is Exploratory Factor Analysis Appropriate?
    2. Introduction to the Common Factor Model
      1. Example: Investment Model Questionnaire
      2. The Common Factor Model: Basic Concepts
    3. Exploratory Factor Analysis versus Principal Component Analysis
      1. How Factor Analysis Differs from Principal Component Analysis
      2. How Factor Analysis Is Similar to Principal Component Analysis
    4. Preparing and Administering the Investment Model Questionnaire
      1. Writing the Questionnaire Items
      2. Number of Items per Factor
      3. Minimal Sample Size Requirements
    5. SAS Program and Exploratory Factor Analysis Results
      1. Writing the SAS Program
      2. Results from the Output
    6. Steps in Conducting Exploratory Factor Analysis
      1. Step 1: Initial Extraction of the Factors
      2. Step 2: Determining the Number of “Meaningful” Factors to Retain
      3. Step 3: Rotation to a Final Solution
      4. Step 4: Interpreting the Rotated Solution
      5. Step 5: Creating Factor Scores or Factor-Based Scores
      6. Step 6: Summarizing the Results in a Table
      7. Step 7: Preparing a Formal Description of the Results for a Paper
    7. A More Complex Example: The Job Search Skills Questionnaire
      1. The SAS Program
      2. Determining the Number of Factors to Retain
      3. A Two-Factor Solution
      4. A Four-Factor Solution
    8. Conclusion
    9. Appendix: Assumptions Underlying Exploratory Factor Analysis
    10. References
  5. Chapter 3: Assessing Scale Reliability with Coefficient Alpha
    1. Introduction: The Basics of Response Reliability
      1. Example of a Summated Rating Scale
      2. True Scores and Measurement Error
      3. Underlying Constructs versus Observed Variables
      4. Reliability Defined
      5. Test-Retest Reliability
      6. Internal Consistency
      7. Reliability as a Property of Responses to Scales
    2. Coefficient Alpha
      1. Formula
      2. When Will Coefficient Alpha Be High?
    3. Assessing Coefficient Alpha with PROC CORR
      1. General Form
      2. A 4-Item Scale
      3. How Large Must a Reliability Coefficient Be to Be Considered Acceptable?
      4. A 3-Item Scale
    4. Summarizing the Results
      1. Summarizing the Results in a Table
      2. Preparing a Formal Description of the Results for a Paper
    5. Conclusion
    6. Notes
    7. References
  6. Chapter 4: Path Analysis
    1. Introduction: The Basics of Path Analysis
      1. Some Simple Path Diagrams
      2. Important Terms Used in Path Analysis
      3. Why Perform Path Analysis with PROC CALIS versus PROC REG?
      4. Necessary Conditions for Path Analysis
      5. Overview of the Analysis
    2. Sample Size Requirements for Path Analysis
      1. Statistical Power and Sample Size
      2. Effect Sizes
      3. Estimating Sample Size Requirements
    3. Example 1: A Path-Analytic Investigation of the Investment Model
    4. Overview of the Rules for Performing Path Analysis
    5. Preparing the Program Figure
      1. Step 1: Drawing the Basic Model
      2. Step 2: Assigning Short Variable Names to Manifest Variables
      3. Step 3: Identifying Covariances among Exogenous Variables
      4. Step 4: Identifying Residual Terms for Endogenous Variables
      5. Step 5: Identifying Variances to Be Estimated
      6. Step 6: Identifying Covariances to Be Estimated
      7. Step 7: Identifying the Path Coefficients to Be Estimated
      8. Step 8: Verifying that the Model Is Overidentified
    6. Preparing the SAS Program
      1. Overview
      2. The DATA Input Step
      3. The PROC CALIS Statement
      4. The LINEQS Statement
      5. The VARIANCE Statement
      6. The COV Statement
      7. The VAR Statement
    7. Interpreting the Results of the Analysis
      1. Making Sure That the SAS Output File “Looks Right”
      2. Assessing the Fit between Model and Data
      3. Characteristics of an Ideal Fit
    8. Modifying the Model
      1. Problems Associated with Model Modification
      2. Recommendations for Modifying Models
      3. Modifying the Present Model
    9. Preparing a Formal Description of the Analysis and Results for a Paper
      1. Preparing Figures and Tables
      2. Preparing Text
    10. Example 2: Path Analysis of a Model Predicting Victim Reactions to Sexual Harassment
      1. Comparing Alternative Models
      2. The SAS Program
      3. Results of the Analysis
    11. Conclusion: How to Learn More about Path Analysis
    12. Note
    13. References
  7. Chapter 5: Developing Measurement Models with Confirmatory Factor Analysis
    1. Introduction: A Two-Step Approach to Analyses with Latent Variables
    2. A Model of the Determinants of Work Performance
      1. The Manifest Variable Model
      2. The Latent Variable Model
    3. Basic Concepts in Latent Variable Analyses
      1. Latent Variables versus Manifest Variables
      2. Choosing Indicator Variables
      3. The Confirmatory Factor Analytic Approach
      4. The Measurement Model versus the Structural Model
    4. Advantages of Covariance Structure Analyses
    5. Necessary Conditions for Confirmatory Factor Analysis
    6. Sample Size Requirements for Confirmatory Factor Analysis and Structural Equation Modeling
      1. Calculation of Statistical Power
      2. Calculation of Sample Size Requirements
    7. Example: The Investment Model
      1. The Theoretical Model
      2. Research Method and Overview of the Analysis
    8. Testing the Fit of the Measurement Model from the Investment Model Study
      1. Preparing the Program Figure
      2. Preparing the SAS Program
      3. Making Sure That the SAS Log and Output Files “Look Right”
      4. Assessing the Fit between Model and Data
      5. Modifying the Measurement Model
      6. Estimating the Revised Measurement Model
      7. Assessing Reliability and Validity of Constructs and Indicators
      8. Characteristics of an “Ideal Fit” for the Measurement Model
    9. Conclusion: On to Covariance Analyses with Latent Variables?
    10. References
  8. Chapter 6: Structural Equation Modeling
    1. Basic Concepts in Covariance Analyses with Latent Variables
      1. Analysis with Manifest Variables versus Latent Variables
      2. A Two-Step Approach to Structural Equation Modeling
      3. The Importance of Reading Chapters 4 and 5 First
    2. Testing the Fit of the Theoretical Model from the Investment Model Study
      1. The Rules for Structural Equation Modeling
      2. Preparing the Program Figure
      3. Preparing the SAS Program
      4. Interpreting the Results of the Analysis
      5. Characteristics of an “Ideal Fit” for the Theoretical Model
      6. Using Modification Indices to Modify the Present Model
    3. Preparing a Formal Description of Results for a Paper
      1. Figures and Tables
      2. Preparing Text for the Results Section of the Paper
    4. Additional Example: A SEM Predicting Victim Reactions to Sexual Harassment
    5. Conclusion: To Learn More about Latent Variable Models
    6. References
  9. Appendix A.1: Introduction to SAS Programs, SAS Logs, and SAS Output
  10. What Is SAS?
  11. Three Types of SAS Files
    1. The SAS Program
    2. The SAS Log
    3. The SAS Output File
  12. SAS Customer Support
  13. Conclusion
  14. Reference
  15. Appendix A.2: Data Input
  16. Introduction: Inputting Questionnaire Data versus Other Types of Data
  17. Entering Data: An Illustrative Example
  18. Inputting Data Using the DATALINES Statement
  19. Additional Guidelines
    1. Inputting String Variables with the Same Prefix and Different Numeric Suffixes
    2. Inputting Character Variables
    3. Using Multiple Lines of Data for Each Participant
    4. Creating Decimal Places for Numeric Variables
    5. Inputting “Check All That Apply” Questions as Multiple Variables
  20. Inputting a Correlation or Covariance Matrix
    1. Inputting a Correlation Matrix
    2. Inputting a Covariance Matrix
  21. Inputting Data Using the INFILE Statement Rather Than the DATALINES Statement
  22. Conclusion
  23. References
  24. Appendix A.3: Working with Variables and Observations in SAS Datasets
  25. Introduction: Manipulating, Subsetting, Concatenating, and Merging Data
  26. Placement of Data-Manipulation and Data-Subsetting Statements
    1. Immediately Following the INPUT Statement
    2. Immediately after Creating a New Dataset
    3. The INFILE Statement versus the DATALINES Statement
  27. Data Manipulation
    1. Creating Duplicate Variables with New Variable Names
    2. Duplicating Variables versus Renaming Variables
    3. Creating New Variables from Existing Variables
    4. Priority of Operators in Compound Expressions
    5. Recoding Reversed Variables
    6. Using IF-THEN Control Statements
    7. Using ELSE Statements
    8. Using the Conditional Statements AND and OR
    9. Working with Character Variables
    10. Using the IN Operator
  28. Data Subsetting
    1. Using a Simple Subsetting Statement
    2. Using Comparison Operators
    3. Eliminating Observations with Missing Data for Some Variables
  29. A More Comprehensive Example
  30. Concatenating and Merging Datasets
    1. Concatenating Datasets
    2. Merging Datasets
  31. Conclusion
  32. Reference
  33. Appendix A.4: Exploring Data with PROC MEANS, PROC FREQ, PROC PRINT, and PROC UNIVARIATE
  34. Introduction: Why Perform Simple Descriptive Analyses?
  35. Example: An Abridged Volunteerism Survey
  36. Computing Descriptive Statistics with PROC MEANS
    1. The PROC MEANS Statement
    2. The VAR Statement
    3. Reviewing the Output
  37. Creating Frequency Tables with PROC FREQ
    1. The PROC FREQ and TABLES Statements
    2. Reviewing the Output
  38. Printing Raw Data with PROC PRINT
  39. Testing for Normality with PROC UNIVARIATE
    1. Why Test for Normality?
    2. Departures from Normality
    3. General Form for PROC UNIVARIATE
    4. Results for an Approximately Normal Distribution
    5. Results for a Distribution with an Outlier
    6. Understanding the Stem-Leaf Plot
    7. Results for Distributions Demonstrating Skewness
  40. Conclusion
  41. Reference
  42. Appendix A.5: Preparing Scattergrams and Computing Correlations
  43. Introduction:  When Are Pearson Correlations Appropriate?
  44. Interpreting the Coefficient
  45. Linear versus Nonlinear Relationships
  46. Producing Scattergrams with PROC PLOT
  47. Computing Pearson Correlations with PROC CORR
    1. Computing a Single Correlation Coefficient
    2. Determining Sample Size
    3. Computing All Possible Correlations for a Set of Variables
    4. Computing Correlations between Subsets of Variables
    5. Options Used with PROC CORR
  48. Appendix: Assumptions Underlying the Pearson Correlation Coefficient
  49. References
  50. Appendix B: Datasets
  51. Dataset from Chapter 1: Principal Component Analysis
  52. Datasets from Chapter 2: Exploratory Factor Analysis
  53. Dataset from Chapter 3: Assessing Scale Reliability with Coefficient Alpha
  54. Appendix C: Critical Values for the Chi-Square Distribution
  55. Index

Product information

  • Title: A Step-by-Step Approach to Using SAS for Factor Analysis and Structural Equation Modeling, Second Edition, 2nd Edition
  • Author(s): Psych. D. Psych. Norm O'Rourke, Ph. D. Larry Hatcher
  • Release date: March 2013
  • Publisher(s): SAS Institute
  • ISBN: 9781629592442