|
![](http://www.statmodel.com/discussion/icons/csc/1px.gif) |
Correlations between first and second... |
![Previous](http://www.statmodel.com/discussion/icons/csc/navgen_arrow_left.gif) ![](http://www.statmodel.com/discussion/icons/csc/1px.gif)
|
Message/Author |
|
|
Dear Linda and dear Bengt, I have two questions: 1) I want to present simple correlations between one individual level variable (aal1, measured by two items) and one aggregate variable (zzaa - inhabitants of certain districts are clustered). Can I do this with "Type is complex" oder do I have to use "type is twolevel"? (Design effects for the two items are far below 2). ANALYSIS: TYPE IS complex; MODEL: aal1 by aa1 aa2; aal1 with zzaa; OUTPUT: STDYX; 2) How can I compute the ICC of a latent variable? Thank you in advance! |
|
|
1) Just use Type=complex basic; 2) Estimate a Type=twolevel model where you estimate the factor variances on both levels and based on the factor variances create icc = between var/(between+within var). |
|
Back to top |
|
|