Linda
In some form you can use a dummy as a dependent variable in logistic model
or in a probit model. These models are non linear and through them you can
estimate the probability of a particular event. This event is relative a
dichotomous variable (1 = success, 0 = failure). The logit model is of form:
Ln(p/(1-p)) = beta0 + beta1*X1 + beta2*X2 + ... + betap*Xp + u
where:
p = probability of sucess
beta0, beta1, ..., betap are parameters
X1, X2, ...., Xp are covariates (independent variables).
u is the error term
p/(1-p) is the odd ratio and p = probability of X = 1. So you are modeling
not X directly, but the probability of X assuming a determinate value. X is
a dummy variable.
The interpretation of the parameters (estimates) values are as follow:
If the estimate of betap is betaphat so a variation of one unity in Xp
"cause" a variation of 100*(exp(betaphat)-1) per cent in the odd ratio
(p/(1-p)), fixing the others covariates. This is somehow different of the
multiple regression estimates in a linear model.
Henrique
-----Mensagem original-----
De: [email protected]
[mailto:[email protected]] Em nome de Linda Walker
Enviada em: s�bado, 1 de dezembro de 2007 14:30
Para: [email protected]
Assunto: st: Dummy
Hello statalist
Can dummy(ies) be used as dependent variable(s)? Please help.
Many thanks
Linda
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
__________ Informa��o do NOD32 IMON 2694 (20071130) __________
Esta mensagem foi verificada pelo NOD32 sistema antiv�rus
http://www.eset.com.br
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
