I'm attempting to understand some data and would greatly appreciate help in picking the appropriate analysis measure.
- I am conducting psychometric analysis of the predictive validity of two separate self-report measures on a dependent variable (also self-report, also continuous).
- All participants completed both IV's and the DV.
- There is reason to believe that the subscales of each measure should contribute to the variance explained in the DV in a particular way.
- Measure 1 (MA) has 5 subscales MA1, MA2, MA3, MA4, MA5.
- Measure 2 (MB) has 5 subscales MB1, MB2, MB3, MB4, MB5.
If I only wanted to know the predictive ability of one measure I would use a hierarchical multiple regression as follows...
- Control for gender
- MA4 and MA5
- What is the best way to assess the best predicting measure?
- Can I run separate heirachical MRs and compare beta values?
- Should I put the competing measures constructs in each step of the hierachical MR and look at what is a significant predictor?