What good methods papers (or step by step 'how to' papers) exist concerning how to assess multiple questionnaire measures including at the individual item level as well as measure subscales? Purpose: to figure out how best to apply measure(s) to asses intervention change where measures were not originally designed for this purpose.
Two of the classic tools for this kind of work are Cronbach's alpha to test for how well items hold together in a single scale and exploratory factor analysis (EFA) to test for the dimensionality of items. There have been so so many questions here about these methods that I have complied a set of resources on them:
https://www.researchgate.net/post/How_can_one_help_people_create_simple_scales_from_Likert-scored_items
Note that the most common application for these methods is to Likert-scored items, but they will actually work on any variables with continuous scores.
If you want to assess the intervention change, you need to employ the Growth Curve Model using Structural Equation Modeling (SEM). Using SEM, you can assess the performance of individual items as well as the performance of their underlying constructs. Also the inter-relationships among the constructs in a model.
Adding to the previous response: Mplus is an excellent program for this purpose. There is very good instruction videos at www.statmodel.com.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis