Odds Ratio vs. P-value

[canadianamazon]

Hi Everyone,

I am a Master's student in South Africa investigating genetic susceptibility to tuberculosis. I was running a haplotype analysis and got a non-significant p-value of 0.056 but a significant Odds Ratio 0.65 (0.45 – 0.94).

Could anyone explain which if my results are significant or not? Clearly there is not a major association but I'm confused none the less.

Thanks!

[lumhearts]

What test was the p-value associated with? I would say that there IS a fairly strong association. First of all, don't let the .05 rule ruin your day. The .05 is arbitrary. I would consider the .056 very suggestive, especially if your sample size is small.

Second, don't let the OR less than 1 (decreased odds) fool you. A 0.65 is actually pretty strong. This is equivalent to an OR of 1.54 (Increased odds), which is strong.

[Link]

yeah...its likely that your p-value isn't the one corresponding to the Odds Ratio. Good luck though.