In logistic regression analysis, is there time where odds ratio is less than 1 and alignment of Confidence interval greater than 1 or odds ratio greater than 1 and alignment of confidence interval less than 1? example, is that possible to have logistic regression analysis result as
AOR--- 0.34 and at the same time 95%CI-------2.34 to 7.01 or AOR ----3.51 and CI ---0.01 to 0.87 irrespective of p value?
No. Your CI should always contain the point estimate (here, odds ratio).
No. It would be like saying that it's probably in Texas, but I'm pretty certain it's somewhere in Kenya.
What is your package? Some give both an odds ratio and a regression coefficient confidence interval. I suspect that is what you are seeing. The relationship is discussed here: https://www.google.com/search?client=firefox-b-1-d&q=relatioship+betwwen+odds+ratio+and+regression+coefficients+in+logistic+regression Use your calculator to check it out. If that is not the case get a new package. Best, D. Booth
PS If you are in the market for a new package I suggest R. See the link:
https://b-ok.cc/book/5825760/ed313f
- Badges
- Science topic
- Similar topics
- Correction