I studied erosion evolution in simulation laboratory with 3 different rainfall intensities and 3 slopes . Also i measured splash erosion in the up and down of splash cup and in the up, medium and down of the study plot. Then i measured sediment particle size. Finally I have 12 depended variables now.
I think you may find the answer at the Andy field book discovering Spss
Way too many variables; results meaningless. Analyze the DVs and make sure none are multicollinear (highly correlated). If they are, they measure the same construct, so you could combine 2 or 3 variables into one.
You did not say how many cases you have, but you need 15 cases per variable in Multiple Linear Regression. I have forgotten how many per variable in MANOVA.
Good luck!
Marcie Zinn
Agree with Jason, while you used simulated data you could have an enough sample size to verify your study. But I suggest if you could categorize dependent variables ( I don't know if you can you didn't mention any thing about them) That will make interpreting results easier.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS