Is there a good way to graph multiple relative risks into a single graph when they arent part of the same regression or model? I have multiple independent relative risks of single variable which I can't combine into a single model due to the fact that my outcome is very rare and combining them would over fit the data. I was using coefplot when I was using a regression but am struggling to use two way to combine multiple separate coefplot commands.
Thanks!
You could store the estimates after each regression and then use coefplot to plot them together.
Here is an example:
sysuse auto
regress price mpg headroom trunk length turn if foreign==0
estimates store domestic
regress price mpg headroom trunk length turn if foreign==1
estimates store foreign
coefplot domestic foreign, drop(_cons) xline(0)
I am able to get coefplot to plot things now just fine but it seems to be plotting the wrong numbers
I am using binreg <var>, rr but ceofplot keeps plotting numbers with negative point estimates or point estimates which clearly arent the relative risk.
I have also tried Logit to plot OR and CS to plot RR all with odd point estimates.
As an example I am trying
binreg culturepos lastrryn, rr (which gives a point estimate of 0.569)
coefplot drop(_cons) xline(1)
The plot is giving a negative point estimate which looks like near -1.
I am not sure what is going wrong with the plotting of the relative risks or odds ratios.
I think you need to use the eform option, which gives you the exponentiated version of the coefficients. Like this:
binreg culturepos lastrryn, rr
coefplot drop(_cons) xline(1) eform
