Chi-square versus G-square versus Fisher's

Discussion in 'SPSS' started by LogReg, Feb 2, 2006.

  1. LogReg

    LogReg Guest

    Can anyone recommend a paper or text or website that clearly and
    concisely compares the appropriateness of these tests (Chi-square
    versus G-square versus Fisher's) under different conditions. I
    appreciate that Fisher's is usually preferred in the 2X2 contingency
    with small samples, but at what sample size would we select a
    chi-square? When would G-square be preferred over chi-square? And how
    are each of these tests effected by rare events (eg. low expected
    frequencies)?

    Thanks in advance.

    Ian
     
    LogReg, Feb 2, 2006
    #1
    1. Advertisements

  2. LogReg

    Bruce Weaver Guest

    See the note on Pearson vs likelihood ratio chi-square at the bottom of
    this page:

    www.angelfire.com/wv/bwhomedir/notes/chi_square_assumptions.txt
     
    Bruce Weaver, Feb 2, 2006
    #2
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.