+ Reply to Thread
Results 1 to 3 of 3

Thread: T-test and large but unequal sample sizes

  1. #1
    Points: 1,044, Level: 17
    Level completed: 44%, Points required for next Level: 56

    Posts
    4
    Thanks
    0
    Thanked 0 Times in 0 Posts

    T-test and large but unequal sample sizes




    Hi,

    I'm working with a very large data set and would like to compare two independent populations. For the first population, I have approximately 30,000 responses. For the second population, I have 3,000 responses. Can I still report the examples from an independent samples T-test?

    I've looked up this question elsewhere, and realize that this usually doesn't work with smaller samples because it is so difficult to establish significance. Yet, when I run this through SPSS, I've got p < .001 and Cohen's d (based on means and SD) is like 3.8.

    Can I report this in my write-up of the data, despite the very unequal (but large) sample sizes?

    Thanks!

  2. #2
    TS Contributor
    Points: 11,073, Level: 69
    Level completed: 56%, Points required for next Level: 177
    Karabiner's Avatar
    Location
    Schalke 04, Germany
    Posts
    1,697
    Thanks
    31
    Thanked 418 Times in 399 Posts

    Re: T-test and large but unequal sample sizes

    Independent samples t-tests can well be peformed with unequal sample sizes.
    Problems discussed often concern unequal sample sizes in combination with
    marked differences of standard deviations (this might lead to biased test
    results).

    You have a very very very large d and a very large sample; is there really any
    doubt that the two groups stem from populations with different means?

    Regards

    K.

  3. #3
    Points: 1,044, Level: 17
    Level completed: 44%, Points required for next Level: 56

    Posts
    4
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: T-test and large but unequal sample sizes


    Hi Karabiner,

    You are correct. I have a very, very, very large d.

    I am embarrassed to confess that this was because I was applying an independent sample t-test with generation as the IV and age as the DV. Oops! Of course it was a large effect size.

    Nevertheless, I do plan to make some comparisons on other variables of interest using the unequal sample sizes, and your answer to my question helps out.

    Thanks!
    Aaron

+ Reply to Thread

           




Similar Threads

  1. Replies: 2
    Last Post: 02-19-2011, 03:35 AM
  2. unequal sample sizes
    By Ali59 in forum Statistics
    Replies: 0
    Last Post: 11-10-2009, 10:07 AM
  3. T-Test and unequal sample sizes
    By Azndude51 in forum Statistics
    Replies: 1
    Last Post: 06-29-2009, 06:44 AM
  4. unequal sample sizes
    By janse030 in forum Psychology Statistics
    Replies: 0
    Last Post: 10-08-2008, 11:54 AM
  5. Unequal sample sizes in JMP?
    By trmstat in forum Other Software
    Replies: 0
    Last Post: 07-31-2008, 07:58 PM

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