+ Reply to Thread
Results 1 to 4 of 4

Thread: Chi square vs Mcneamar's test

  1. #1
    Points: 1,017, Level: 17
    Level completed: 17%, Points required for next Level: 83

    Posts
    13
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Chi square vs Mcneamar's test




    Hello I have a cohort study where a cohort of subjects with a specific characteristic has been matched to a control group according to age (+/ 3 years) and sex. Before running the main analysis I am analyzing differences at baseline about whether the two groups smoke and drink alcohol (boh variables binary 0/1).
    Specifically I would want to check whether there are differences of sex smoking status and drinking status between matches and controls. Shall I use the Chi square or the Mcneamar's test in this case?
    I would go for Chi square for sex and Mcneamar's for smoking and alcohol consumption. What do you think?
    And also, I am quite familiar with the interpretation of the Chi square, not as much for the McNeamar's test. If I find a significant result for the alcohol consumption what does it mean?
    Many thanks in advance

  2. #2
    Omega Contributor
    Points: 38,253, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,989
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: Chi square vs Mcneamar's test

    Superficially seems about right. And you would be hoping the Chi sq has a high pvalue to validate the process.

    You amight aso think about conditional logistic regression, since you may have more options with that procedure.
    Stop cowardice, ban guns!

  3. The Following User Says Thank You to hlsmith For This Useful Post:

    Ralph84 (02-15-2015)

  4. #3
    Points: 1,017, Level: 17
    Level completed: 17%, Points required for next Level: 83

    Posts
    13
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Re: Chi square vs Mcneamar's test

    Yes, the p value for sex is quite high (as the two populations were matched according to gender too). It is just that using the Mcneamar's for sex doesn't sound quite right to me even if I read that is an appropriate test for testing the difference of distribution of a variable in matched groups

  5. #4
    Points: 1,017, Level: 17
    Level completed: 17%, Points required for next Level: 83

    Posts
    13
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Re: Chi square vs Mcneamar's test


    I have another question about the topic, is the McNeamar still the right approach to use when you have matched case-control population with a ratio 1:2 and you want to test whether there is a different frequency of the binary variable (i.e. smoking status categorized as "yes" "no") between cases and controls ?
    In this case the paired data would appear twice as matched twice to different controls, therefore I was wondering if this is still the right approach to use to test such things.
    Many thanks in advance

+ Reply to Thread

           




Posting Permissions

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






Advertise on Talk Stats