Re: st: Use of the term 'relative risk ratio'

[Michael Ingre <Michael.Ingre@ipm.ki.se>]

Many thanks to Roberto G. Gutierrez, StataCorp, for your very clear and 
pedagogical description of the RRR concept and how it is used by Stata. 
Maybe some of the confusion about RRR is related to pronunciation (or 
me not being a native in English)?

(1) relative-risk ratio - a ratio of two relative-risks
(2) relative-risk-ratio - a ratio of two relative-risks
(3) relative risk-ratio - a risk-ratio that the describe the relative 
of two (absolute) risks

I'm currently writing a paper using a special case of number 3 above 
that I don't know how to present (used to call it RRR). The normal use 
is just risk ratio (RR) comparing the risk of having a disease compared 
to not have it. However, suppose you estimate a -logit- and want to 
describe the ratio between the risk given one level of the exposure to 
the risk given another level. These ratios will not behave like normal 
RRs across the levels of the exposure because of the S-shaped 
(logistic) probability function. What would you call these ratios: RRs 
or RRRs ...?

 p1  = P(Y=1 | x+1)     q1 = p(Y=0 | x)
 p2  = P(Y=1 | x=2)
 p10 = P(Y=1 | x=10)

	  p1
RR  = ----
       q1

	  p10
??  = ----
       p2

Any suggestions?

Thanks

Michael

*
*   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/