[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: Logit transformation

From   [email protected]
To   [email protected]
Subject   Re: st: Logit transformation
Date   Fri, 08 Aug 2008 12:30:58 +0200


Thank you for your explanation. This is what I am trying to do. I want to relate the probability of being poor to the logit coefficients in a linear function.

Zitat von "Maarten buis" <[email protected]>:

--- [email protected] wrote:
I would to ask you a technical question on the transformation of
logit coefficients into non-negative weights.

For example, I have run a logit regression which gave the following

logodds= 0.2198*agrhh +0.7161*ea_trad_boma - 0.0673*ruralnorth
           + 0.1807*ruralcentre - 0.4402*ruralsouth;

odds= p/(1-p) is the odd ratio
p is the probability of being poor the explanatory variables are all
categorical and take the value of 0 or 1.

Now, I want to transform all of the coefficients into non-negative
integers (based on a linear relationship as above) so that the
prediction of the logodds  ranges from 0 to 100.
What I think you want to know is how the explanatory variables
influence the probability of being poor. The fact that the probability
is a number between 0 and 1 (or 0 and 100, if you prefer to think in
terms of percentages) does not mean that the effects have to be
positive. I certainly does not mean that the effect has to be an
integer, this is not true for the probability either. An excelent text
on interpreting these kinds of models is J. Scott Long & Jeremy
Freese's "Regression Models for Categorical Dependent Variables Using
Stata" published by the Stata Press: .

Anyhow, if my assumption of what you want is correct, the command you
are looking for is -mfx-, see -help mfx-.

Hope this helps,

Maarten L. Buis
Department of Social Research Methodology
Vrije Universiteit Amsterdam
Boelelaan 1081
1081 HV Amsterdam
The Netherlands

visiting address:
Buitenveldertselaan 3 (Metropolitan), room Z434

+31 20 5986715

Not happy with your email address?.
Get the one you really want - millions of new email addresses available now at Yahoo!
* For searches and help try:

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index