+ Reply to Thread
Results 1 to 5 of 5

Thread: Difference between two independent proportions

  1. #1
    Points: 9, Level: 1
    Level completed: 17%, Points required for next Level: 41

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Difference between two independent proportions




    Hello,

    I am not a statistician, however, for my latest small research, I would like to test whether there is a difference between two independent groups of samples in terms of proportions (one with a condition and one without a condition).

    I have researched this problem and find out that different researchers did use two different approaches:
    a) z-test to test the difference between two independent proportions
    b) chi-quadrate

    Now I would like to kindly ask you personal opinion/experience which one is more accurtae in this matter and why ?

    Looking forward to hearing (reading) your precious thoughts. Thank you very much !

    Cheers,
    Zuzana.

  2. #2
    Omega Contributor
    Points: 38,432, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,006
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Difference between two independent proportions

    What is your sample size and proportion of outcome in each group?
    Stop cowardice, ban guns!

  3. #3
    Points: 9, Level: 1
    Level completed: 17%, Points required for next Level: 41

    Posts
    2
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Difference between two independent proportions

    Actually, I am at the stage of defining a sample size for my study. This is why I have been evaluating these different approaches. But my expected proportions (also for the purpose of power analysis) are 50% to 50% in the intervention group and 70% to 30% in the control group. Does this influence the choice of the test ? (but I would expect a sample size of approx. 300 samples - 150 in each group). Thx for your thoughts.

  4. #4
    Omega Contributor
    Points: 38,432, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,006
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Difference between two independent proportions

    Yeah, the typical approach for this question would be to use the chi-square test. Though, if any expected cell frequency is <5, the Fisher Exact test gets used. I am sure the z-test is fine to use, when samples get larger things normalize.


    It might also come down to your field. I am in medicine and we hardly ever use the Z over chi-square.


    I am not familiar with the chi-quadrate. What is that in particular? Do you have a link. So is it just a phrasing that says you are comparing two rates from a 2x2 table?
    Stop cowardice, ban guns!

  5. #5
    Points: 5, Level: 1
    Level completed: 9%, Points required for next Level: 45

    Posts
    1
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Difference between two independent proportions


    Hi, following up on this discussion. I'd like to ask, WHY, for example in medicine, is chi-square typically used instead of z? I am learning Statistics at a beginner level and came across this problem of a fair coin: https://frank.itlab.us/datamodel/node42.html
    I wondered why you would not just use the z-test? I cannot figure out at this point when to use the chi-square instead of the z-test if working with (1 or 2) proportions. Help is appreciated; thanks.

+ Reply to Thread

           




Tags for this 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