Hi All,
When looking at Discriminant and Convergent validity of a measure through Pearson's correlations, do I enter each item in the questionnaires, or the questionnaire totals?
Thank you!
Christian Geiser
You would typically compute sum scores (composites, aggregates across the items that pertain to the same scale) first, then correlate the scale sum scores to examine convergent and discriminant validity across different measures. This implies that the items within each scale/measure are unidimensional--an assumption that you should test first (e.g., by using confirmatory factor analysis or methods of item response theory).
3 votes
1 thanks
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology
- Culture Cells
- Cell Line
Similar questions and discussions