Tell me more ×
Cross Validated is a question and answer site for statisticians, data analysts, data miners and data visualization experts. It's 100% free, no registration required.
up vote 2 down vote favorite
1

I want to compare the strengths of different predictors (all at Level 1) using HLM and am not sure what statistical test would be appropriate.

More specifically, if I have 3 predictors at level 1 and I want to know which of them are statistically different from the others, what is the best way to do this?

share|improve this question
1  
If you standardize all the variables, then whichever has the highest beta will be "most important" in some sense. But be careful - standardized coefficients are dependent on the sd of the variables in the particular data set you are using. In general, I don't think it's a good idea to try to do what you are doing (although other statisticians disagree with me) – Peter Flom Oct 4 '12 at 12:13
(NB, this was posted as an edit by an anonymous user, possibly the OP w/o access to their account.) Thanks for your input @PeterFlom - I think I need to provide a little more context... I am looking at concordance between 2 variables (repeated measures) at 3 time points- pre-treatment, post, and 6 months follow-up, and would like to be able to determine (for example) if the post treatment concordance is significantly higher than pre-treatment concordance. – gung Oct 9 '12 at 2:47
(continued) Therefore, all predictors will already be in the same metric, and so I don't believe that standardizing should make a difference. However, I am hoping that there is some sort of procedure that will allow me to determine if the resulting betas are significantly different...Any thoughts? – gung Oct 9 '12 at 2:47
Standardizing will always make a difference - it changes the units from whatever they were to start to standard deviations. Regression coefficients have means and standard errors and can be tested. – Peter Flom Oct 9 '12 at 10:11

Know someone who can answer? Share a link to this question via email, Google+, Twitter, or Facebook.

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.