[pileated]

Ok so, I am terrible at biostats...This is so fundamental I am a little embarrassed to post it but I can't find the answer and don't quite know where to look....

When I am looking at a CI, how do I know when the null is 0 vs when it is 1?

For example in most questions I have seen, it is not statistically significant if the CI includes 1 CI (0.8-1.2) example
There is a question on uworld that is stat significant because the null does not cross 0 CI(1-1.4)

so how do I differentiate between CI crossing 1 vs 0?

 

[Dimas]

Hi ..

If given CI includes 1, H0 is not rejected which means no statistically significant effect of exposure
Examples
C.I. (0.8 – 2.46) not statistically significant (have same risk)
C.I. (1.22 – 2.45) increase risk (RR=1.77) ---> 1.77-1= 0.77 or 77% more cases
C.I. (0.45 – 0.95) decrease risk (RR= 0.78) ---> 1-0.78= 0.22 or 22% .. so one group has 22% reduction in risk

taken from Kaplan book
hope this will help