I have to detect certain bacteria and used two methods for same 200 samples. I have tabulated two methods as below:

Sample  Method1   Method2
1         Yes       No
2         No        yes
200       Yes       yes

I have used chi-square test and found there is significant difference between two different methods. Is there any other statistical test I can use? As there is a gold standard, if both method1 and method2 turns out to be yes, then the sample does necessarily contains the bacteria. Can I use this? Can some one guide please?

  • $\begingroup$ What is your research question / hypothesis? What is the goal? $\endgroup$ – user2974951 Feb 4 at 7:16
  • $\begingroup$ Do you have the ground truth for each of these samples? Meaning that do you know which ones of these 200 contain bacteria or you have to rely on the methods’ outputs only $\endgroup$ – Vasilis Vasileiou Feb 4 at 8:24
  • $\begingroup$ Because you have observations paired within each sample, you might want to ask a somewhat different question. Is the count of a) method1 reporting a yes when method2 reports a no; greater than b) method1 reporting a no when method2 reports a yes. This is McNemar's test. Note that this test ignores the cases when both tests answer yes and when both tests answer no. $\endgroup$ – Sal Mangiafico Feb 4 at 13:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.