Hi Martin and Eva,
Thanks very much! Your advice really helped a lot. Especially, I
learned a way to check out some suspicion with hypothetical conditions
---for an instance, if I want to make sure the problem is caused by
missing values, I could check if the problem becomes more severe when
missing percentage gets larger. I can use this idea to deal with lots
of other problems. I'm happy.
Thanks!
Mandy
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
