+ Reply to Thread
Results 1 to 3 of 3

Thread: Post-hoc power analysis

  1. #1
    Points: 44, Level: 1
    Level completed: 88%, Points required for next Level: 6

    Posts
    3
    Thanks
    1
    Thanked 0 Times in 0 Posts

    Post-hoc power analysis




    I have analyzed data by performing a regression model in three steps - (a) two psychological independent variables, (b) adding a political independent variable (3 variables), (c) adding a political demographic independent variables (five variables).

    I need to calculate post-hoc power analysis through specific software G * power 3. I wonder if I should use ‘F test: Multiple Regression - omnibus (deviation of R2 form zero), fixed model’ and how to report it.

    Many thanks in advance :-)

  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: Post-hoc power analysis

    Post hoc power can be computed simply from the observed p-value and the sample size. In fact, for moderate sample sizes post hoc power is basically just a simple function of the p-value. See the following: http://www.stat.uiowa.edu/files/stat/techrep/tr378.pdf
    So post hoc power doesn't really add any new information beyond what you already know. If you rejected the null, then post hoc power is > .5. If you failed to reject the null, then post hoc power is < .5. So you may want to reconsider what it is exactly that you think you will learn from a post hoc power analysis.
    “In God we trust. All others must bring data.”
    ~W. Edwards Deming

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

    Vasily_Kot (07-24-2014)

  4. #3
    Points: 44, Level: 1
    Level completed: 88%, Points required for next Level: 6

    Posts
    3
    Thanks
    1
    Thanked 0 Times in 0 Posts

    Re: Post-hoc power analysis


    Thank you very much for the explanation and the link.
    From what I gather , Post hoc power can indicated that the N that was researched was enough for rejected the null.
    Above all, I actually have to report statistical power of the study (a posteriori, if it had not been calculated before data collection, as in my case) using free software GPower*3 because this is the instruction I was given.
    I am not quit sure, though:
    1. How to use G*Power 3 in my case (may it be through ‘F test: Multiple Regression - omnibus (deviation of R2 form zero), fixed model’?)?
    2. How to formally report calculating statistical power using G*Power ?

+ 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