AI Snips
Chapters
Transcript
Episode notes
Factor Analysis As Dimension Reduction
- Factor analysis reduces many item correlations into fewer, meaningful dimensions so researchers can interpret structure.
- Exploratory FA reveals how items naturally cluster without imposing a predefined pattern.
EFA Isn't Fully Exploratory
- EFA is only partly exploratory because item selection and theory usually guide item pools before analysis.
- The 'exploratory' part mainly concerns how items cluster, not how items were chosen.
Orphan Items Revealed Social Withdrawal
- Gregg recounts a real project where two items behaved as orphans and formed a separate factor of social withdrawal.
- The result led them to drop items and plan targeted item development for that new construct.