0
$\begingroup$

I have been asked by co-authors on a paper to fill in the Odds Ratios marked as (?) in a table like the one below.

Country of origin Land A (Odds Ratio) Land B (Odds Ratio)
Number of people who meet the outcome of interest 12 (?) 4 (?)
Median age 33 (?) 28 (?)
Number of males 16 (?) 14 (?)
Number of females 17 (?) 14 (?)

Would you need to run a multivariate logistic regression to obtain an OR while considering both independent variables (Land A = Yes, Male = Yes) in relation to the dependent variable (Outcome of interest = Yes)?

Or am I on the wrong track entirely? Intuitively I feel like this isn't right somehow.

Improve this question
$\endgroup$
3
  • 1
    $\begingroup$ Odds ratio is composed of two odds, which are composed of probabilities. You don't have either in this table. $\endgroup$
    – user2974951
    Commented Jan 24 at 12:06
  • $\begingroup$ You have three independent variables: Age, sex, and land. $\endgroup$
    – Peter Flom
    Commented Jan 24 at 12:25
  • $\begingroup$ What about in the case of cell B4, reading 16? In the context of the table, I would take this to mean outcome ~ land + sex. $\endgroup$
    – jserv
    Commented Jan 24 at 12:33

1 Answer 1

Reset to default
2
$\begingroup$

First, this would be multiple logistic regression, not multivariate. The latter term applies when there are multiple outcome or dependent variables. Also, the cells mix counts and statistics.

Second, this table is oddly formatted and doesn't seem to contain the information needed to run a logistic regression. If "outcome of interest" is the DV, then it should be the columns and Land should be one of the rows. And "median age" won't work well.

If you have the raw data, then you can use any statistical program to run a logistic regression along the lines of

outcome ~ land + age + sex

(with details depending on the program you use). The output will include ORs for each independent variable.

Improve this answer
$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.