odds ratio

[LarsFolkestad]

Dear List
I have a quick question.
I have two groups 0 and 1 and want to calculate the odds ratio with CI for a given event.
The event is a 0 or 1 event.
When i manually calculate the Odds Ratio it is about 1.96. In stata i use the logistic command:

.logistic event group

and now i get an odds ratio of 2.41

1) is there a way to get stata to calculate the odds ratio using a 2 by 2 table (ad/bc) with CI.
2) i'm guessing the odds ratios are different because the latter is a logistic regression model. I have not tester for GOF.

thank you

[Carlo Lazzaro]

Lars:
see -help cci-

[Maarten Buis]

You should have gotten the same odds ratio, and in the example below that is indeed the case:

sysuse nlsw88, clear

// compute manually
tab collgrad married
di (616*344)/(188*1098)

// use logit
logit collgrad married, or

I suggest you check again your exect manual computation and your logit command, for example what did you do with missing value?

[LarsFolkestad]

Thank you both.
Maarten, there are no missing values - but i did not use your logit command, i used the logistic event group
not sure if that was it though. I'll have a new look at it.
Lars

[LarsFolkestad]

Maarten,
you were right. gives the same OR.
lars