Thread: Checking stat. significant difference between combined measures of performance

1. Checking stat. significant difference between combined measures of performance

Hello!

I wonder how to check is there statistically significant difference between my combined measures of performance of diagnosis methods. But let me describe the problem.

We have roughly 1000 patients in a sample. All of them have a tumor: benign or malignant.
For 80% of the sample we can perform a diagnosis of tumor by method A and for 70% (not always the same patients) by method B.
For each of method we can calculate accuracy, sensitivity and specificity. Now I have my combined measure of performance of method which is x = accuracy*sensitivity*specificity.

My question is: how can I check whether there is statistically significant difference between x_A and x_B?

2. Re: Checking stat. significant difference between combined measures of performance

I have not seen this before "combined measure of performance of method which is x = accuracy*sensitivity*specificity".

Does this combination represent something in particular or is a defined metric?

So are you asking if the Accuracy*SEN*SPEC for A and B compared differ significantly? What would this matter? Please provide some more details, in that I am interested in what you are doing.

I think you may have issues comparing the two tests on different samples. Do some have only one test while others have both or just the other one? How was the use of either test or both assigned, randomly or is this observations data?

 Tweet

Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts