Hello,
i have a study in which i want to prove that test A is not inferior / equivalent to Test B in detecting disease X.
i have 300 subjects, of which 100 have disease X. I used both test on all the subjects and the detection rate for test A was 89% and for test B was 91%. What statistical test/method should i use?
Thank you
Before you can select a statistical test, you need to determine what makes a screening test inferior or superior compared to another screening test.
In your question, you only consider the test sensitivity "the detection rate for test A was 89% and for test B was 91%". If that was your only concern, then it would be very easy to design the perfect screening test: Simply let your test always produce a positive result and you get 100% sensitivity.
I hope that makes clear that you should also somehow consider the specificity of the screening test.
Then you should consider what is the (relative) value of a false negative and a false positive screening test result.
Another important consideration is whether you are only concerned with how a screening test in a population with 1/3 prevalence.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size