Dear all, 
 
I am looking into modeling risky sexual behaviour in terms of the possibility to catch STDs for sexually active adults in a refugee camp. 
 
The standard knowledge, attitude and practice questionnaire looks at three types of partnerships: regular (living-in); non-regular (occasionnal); and in exchange of money, favour or a service. Then a question is asked about condom use. 
 
If we reduce somewhat the possibilities on condom use, we get about 16 "choices" from not having had any sexual partnership to all three without condom use, and any combination in between. In the data however only 7 choices actually appear. 
 
I have used a stereotype logit but it fails to converge. Nested or cross-nested models seem difficult to use because branches stop whenever a person says no.
 
The public health literature is fond of simple logit.
I could have separate tractable model for simpler combinations. Among those with a regular partner, regular only vs (+ non regular or exchange and consistent condom use) vs (+ non regular and exchange and inconsistent condom use). Etc among people without a regular partner.    
 
What to do?
Thank you in advance. Charles Lor School of International Service American University 
_________________________________________________________________
With Windows Live for mobile, your contacts travel with you.
http://www.windowslive.com/mobile/overview.html?ocid=TXT_TAGLM_WL_Refresh_mobile_052008
**********************************************************
             Political Methodology E-Mail List
   Editors: Melanie Goodrich, <[log in to unmask]>
            Delia Bailey, <[log in to unmask]>
**********************************************************
        Send messages to [log in to unmask]
  To join the list, cancel your subscription, or modify
           your subscription settings visit:

          http://polmeth.wustl.edu/polmeth.php

**********************************************************