+ Reply to Thread
Results 1 to 11 of 11

Thread: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Stat

  1. #1
    Omega Contributor
    Points: 38,303, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,993
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Stat




    I am trying to get an effect size for a Wilcoxon rank sum test. There is a postulated approach of dividing the z-score from the normalization of test by the square root of the sample size.


    Though this approach does not have a respective way to get the 95% CI for the above calculated effect size. Does it make any sense to use 95% Percentile Bootstrap confidence intervals? So bootstrap the dataset and calculate the ES for each sample and use the 2.5 and 97.5 centile values?
    Stop cowardice, ban guns!

  2. #2
    Cookie Scientist
    Points: 13,431, Level: 75
    Level completed: 46%, Points required for next Level: 219
    Jake's Avatar
    Location
    Austin, TX
    Posts
    1,293
    Thanks
    66
    Thanked 584 Times in 438 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    Yes it makes sense
    “In God we trust. All others must bring data.”
    ~W. Edwards Deming

  3. The Following 3 Users Say Thank You to Jake For This Useful Post:

    bryangoodrich (03-28-2016), hlsmith (03-28-2016), spunky (03-28-2016)

  4. #3
    Devorador de queso
    Points: 95,705, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,931
    Thanks
    307
    Thanked 2,629 Times in 2,245 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    No. It makes no sense at all.

    I just wanted to provide a contrasting view. Jake is right of course but he needs to be disagreed with from time to time.
    I don't have emotions and sometimes that makes me very sad.

  5. The Following 2 Users Say Thank You to Dason For This Useful Post:

    bryangoodrich (03-28-2016), hlsmith (03-28-2016)

  6. #4
    Omega Contributor
    Points: 38,303, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,993
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    Great. It seemed fine with me as well.

    Bootstrap nation - long live Efron!
    Stop cowardice, ban guns!

  7. #5
    Cookie Scientist
    Points: 13,431, Level: 75
    Level completed: 46%, Points required for next Level: 219
    Jake's Avatar
    Location
    Austin, TX
    Posts
    1,293
    Thanks
    66
    Thanked 584 Times in 438 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    Quote Originally Posted by Dason View Post
    I just wanted to provide a contrasting view.
    We always try to be fair and balanced here at Talk Stats.
    “In God we trust. All others must bring data.”
    ~W. Edwards Deming

  8. #6
    Devorador de queso
    Points: 95,705, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,931
    Thanks
    307
    Thanked 2,629 Times in 2,245 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    No we don't. JAKE IS STUPID AND IS A COTTON-HEADED-NINNY-MUGGINS.
    I don't have emotions and sometimes that makes me very sad.

  9. #7
    Points: 3,006, Level: 33
    Level completed: 71%, Points required for next Level: 44

    Posts
    177
    Thanks
    1
    Thanked 29 Times in 29 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    hah, I reread what you are trying to do.

    One comment though. To do this you need to correct for bias.

  10. #8
    Omega Contributor
    Points: 38,303, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,993
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    Are you talking about the slanderous posts or my original post

    Do you want to elaborate! I have seen bootstrap used under both scenarios. It seems when the average BS is calculated and subtracted it is a bias correction. Though it can be used as above. I heard a talk awhile back that I seem to recall the presenter mentioning correction for sampling variability and also need to control for bias - but it was a simulation approach not a BS approach.
    Stop cowardice, ban guns!

  11. #9
    Omega Contributor
    Points: 38,303, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,993
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    the42up,


    Perhaps, this is what you are referring to. Though, I don't think I need to correct the estimate in my presented scenario.


    http://www.ats.ucla.edu/stat/sas/faq/bootstrap.htm
    Stop cowardice, ban guns!

  12. #10
    TS Contributor
    Points: 12,227, Level: 72
    Level completed: 45%, Points required for next Level: 223
    rogojel's Avatar
    Location
    I work in Europe, live in Hungary
    Posts
    1,470
    Thanks
    160
    Thanked 332 Times in 312 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta

    hi
    actually I always have a bit of a bad feeling about bootstrapping high percentiles. I know that for a maximum value distribution the bootstrap would be a poor choice - so why would it be better for the 95th percentile?

    Regards

  13. #11
    Omega Contributor
    Points: 38,303, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    6,993
    Thanks
    397
    Thanked 1,185 Times in 1,146 Posts

    Re: Bootstrap Approach for Effect Size 95% Percentile Interval for Non-parametric Sta


    What do you mean in particular by this:


    "maximum value distribution the bootstrap would be a poor choice "?
    Stop cowardice, ban guns!

+ 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