Confirmatory Factor Analysis in lavaan with examples
What is lavaan?
lavaan is a free, open source R package for latent variable analysis. You can use lavaan to estimate a large variety of multivariate statistical models, including path analysis, confirmatory factor analysis, structural equation modeling and growth curve models.
What is "Confirmatory Factor Analysis for Applied Research"?
It is a very comprehensive book by Timothy A. Brown emphasizing practical and theoretical aspects of Confirmatory Factor Analysis (CFA), a well-established multivariate statistics method.
And what exactly can I find in this page? CFA/SEM is mostly performed via commercial software, such as Mplus, EQS, Lisrel and Amos. The R package lavaan offers a very promising open source alternative. The companion website of Brown's book on CFA currently contains the syntax code for running the book examples in most (if not all) of the above commercial packages. So what's missing? A lavaan implementation (this package is here to stay, I can feel it). Here it is, following the chapters of the book.
Chapter 4 - Specification and Interpretation of CFA Models
Contains two examples a) Two-Factor CFA (Neuroticism, Extraversion) and b) CFA with Single Indicators: Health Status
Chapter 5 - CFA Model Revision and Comparison
Contains two examples a) a Three-Factor CFA Model (Alcohol Motives) and b) E/CFA (EFA in the CFA framework)
Chapter 6 - CFA of Multitrait–Multimethod Matrices
Contains two examples a) a Correlated Methods Model of a Multitrait-Multimethod Matrix (MTMM) and b) a Correlated Uniqueness Model of a MTMM matrix
Chapter 7 - CFA with Equality Constraints, Multiple Groups, and Mean Structures
Contains examples for a) Tau Equivalent and Parallel Indicators b) Longitudinal Measurement Invariance c) Multiple Groups CFA (Depression example) d) MIMIC Models (Phobia example)
Chapter 8 - Higher-Order CFA, Scale Reliability Evaluation, Formative Indicators
Contains examples for a) a Higher Order model CFA b) estimating a scale's reliability using the semtools package c) a Formative Indicators model
Chapter 9 - Data Issues in CFA: Missing, Non-normal, and Categorical Data (coming soon)
Chapter 10 - Statistical Power and Sample Size (coming soon)
Useful links:
(1) A discussion group for the R package lavaan
(2) Y. Rosseel's paper on lavaan in the Journal of Statistical software