0
$\begingroup$

question 1: I have 6 variables where 2 binary predictor variables have a much higher odds ratio than the other variables. One variable has 8.40 odds ratio and the second has 3.16 odds ratio. The other variables are between 1.42 and 1.54 odds ratio. It seems like the variable with 1.42 odds ratio would be a much more reasonable result to get, because 1 unit increase in that variable, would give a 42% higher probability of success. 8.40 odds ratio would on the other hand give something like 740 % higher probability of success. Does that make sense, or is there something wrong with my regression?

Question 2: The other variables with an odds ratio between 1.42 and 1.54 have an insignificant p-value (0.10 and 0.25 at a 0.05 significant-level). Does this mean that i can't interpret on those variables' odds ratio, or is that still an option?

Screen shot of the analysis: http://prntscr.com/i959r7

Improve this question
$\endgroup$
2
  • $\begingroup$ It might help if you edited in information about the scales of your predictor variables. The coefficients are for unit increase in the predictor and obviously that is scale dependent. $\endgroup$
    – mdewey
    Feb 2, 2018 at 13:39
  • $\begingroup$ The predictor variables with odds ratio 8.40 and 3.16 are binary. $\endgroup$
    – Jacob Rosschou
    Feb 2, 2018 at 13:51

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.