Michael Ingre wrote:
I have estimated a logit model and would like to present Relative Risk
Ratios (RRR) comparing the risk at different levels of a scale used as a
covariate in the model. Odds Ratios are easily calculated with -lincom-.
RRRs are also easily calculated by just taking the ratio between predicted
probabilities for different levels of the scale. It is also quite straight
forward to calculate a confidence interval of the predicted probabilities.
But how do I calculate Confidence Intervals of the risk ratios?
----------------------------------------------------------------------------
Could you fit a different generalized linear model to your data? One with
the logarithmic link instead of the logit link will give you the risk ratio
directly. I'm not really sure what you mean by comparing the risk at
different levels of scale, but you can set each of the levels of scale of
the covariate as the reference level for estimating the risk ratios (and
their confidence intervals) of the other levels relative to it.
Joseph Coveney
set obs 200
generate byte response = uniform() > 0.5
generate byte category = round(uniform() * 5, 1) // six levels of scale
xi: glm response i.category, family(binomial) link(log) eform nolog
char category[omit] 1 // going to a different level of scale
xi: glm response i.category, family(binomial) link(log) eform nolog
// and so on
exit
*
* 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/
