Others have mentioned the procedures you might want to use.
You can also use a transformation like logit(y), or any other inverse CDF of a continuous unbounded RV and then use -regress-, though you should be wary of extreme observations (near 0 or 1) as outliers in the transformed space can seem quite innocuous in the untransformed space. I never thought about it before but the Cauchy link might be useful in this regard.
-----Original Message-----
From: "Andrea Rispoli" <[email protected]>
To: [email protected]
Sent: 12/29/2008 2:59 PM
Subject: st: regression with dependent variable ranging from 0 to 1
Dear Statalist,
I would like to run a regression in which the dependent variable has
values that fall between zero and one by definition.
Can anyone suggest me a way to deal with this problem?
Thank you very much,
Best,
AR
*
* 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/
*
* 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/
