Exploratory Factor Analysis and PCA
What factor analysis is for and how PCA differs from common factor analysis.
Slides · Dataset
About 40 minutes to complete
Enrol to open the slides and download the data.
What this module covers
what factor analysis is for and how PCA differs from common factor analysis; sample size requirements; KMO and Bartlett's test of sphericity as the entry ticket; extraction methods; deciding how many factors - eigenvalue > 1, the scree plot, parallel analysis, and interpretability, with the honest statement that the last one usually wins; orthogonal vs oblique rotation and how to choose; reading the pattern matrix, cross-loadings and communalities; naming factors; what to do with items that load nowhere; how to report an EFA.
By the end of it
- What factor analysis is for and how PCA differs from common factor analysis
- Sample size requirements
- KMO and Bartlett's test of sphericity as the entry ticket
- Extraction methods
- Deciding how many factors - eigenvalue > 1, the scree plot, parallel analysis, and interpretability, with the honest statement that the last one usually wins
- Orthogonal vs oblique rotation and how to choose
- Reading the pattern matrix, cross-loadings and communalities
- What to do with items that load nowhere
- How to report an EFA
The dataset
Maternal health knowledge, attitude and practice scale — Sigma teaching dataset
A teaching dataset of 240 cases, generated from a fixed seed so every figure in the module reproduces exactly.
- CodebookIncluded
- Dataset (CSV)Included
- SPSS syntaxIncluded