Close Home Forum Sign up / Log in

stats help :(

J

The 2x2 was what I was originally thinking but the poster said they were interested in the conditional probability of the response given that the user has a need which means you are looking at just one column (or row) of the 2x2 contingency table. An odds ratio only applies if you are comparing the two groups.

C

Despite having said early I thought you needed an odds ratio, I think Jewel is right.

All you can really say is given someone had a need, they were x% likely to have that need met (or x/y where x is the number who had need met and y is the total number of people who had a need)

(I hope this is what you're saying Jewel, I really don't want to confuse this thread further. If it is not then I'll bow out!)


9545