How do I do a chi-square test when one of my expected values is 0?

  • Thread starter Thread starter deadbilly
  • Start date Start date
D

deadbilly

Guest
Or does this even work? I'm trying to test a set of values against an expected ratio of 2:0:1:1. Obviously this causes divide-by-zero problems in my final chi-square. In this case do I ignore the zero term or is there some other way? (like reversing observed and expected for that value?)

(In this case, I know the test will show < 0.01, but I have to do it anyway).
Thanks!
 
Back
Top