+ Reply to Thread
Results 1 to 6 of 6

Thread: Effect Sizes

  1. #1
    Points: 12, Level: 1
    Level completed: 23%, Points required for next Level: 38

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Effect Sizes




    This may seem like a ridiculous question, but it's bugging me and no text book or paper I've read has answered it. I am looking into research on effect sizes and their use in research. I see a lot of mention of Cohen's D and F2 and the "small", "medium" and "large" effect sizes. Can someone explain (with an example) why a researcher would choose a "medium" effect size for example? I'm just not seeing any justifications and I know it isn't just an arbitrary decision.

  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: Effect Sizes

    Classifications are arbitrary. A 0.1 may be super exciting to one person in their context and not in a different context. Think of saving lives, a little bit is still important, though if you decrease length of hospital stay by 15 minutes that may not mean much.
    Stop cowardice, ban guns!

  3. #3
    TS Contributor
    Points: 18,889, Level: 87
    Level completed: 8%, Points required for next Level: 461
    CowboyBear's Avatar
    Location
    New Zealand
    Posts
    2,062
    Thanks
    121
    Thanked 427 Times in 328 Posts

    Re: Effect Sizes

    Quote Originally Posted by chadwae View Post
    Can someone explain (with an example) why a researcher would choose a "medium" effect size for example?
    I'm not quite sure what you mean by "choose". Do you mean "Why would a researcher interpret an effect size of (say) d = 0.5 as being "medium" in size? Or do you mean, why would a researcher hypothesise a "medium" effect for power analysis? Or do you mean something else?
    Matt aka CB | twitter.com/matthewmatix

  4. #4
    Points: 12, Level: 1
    Level completed: 23%, Points required for next Level: 38

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Effect Sizes

    Quote Originally Posted by CowboyBear View Post
    I'm not quite sure what you mean by "choose". Do you mean "Why would a researcher interpret an effect size of (say) d = 0.5 as being "medium" in size? Or do you mean, why would a researcher hypothesise a "medium" effect for power analysis? Or do you mean something else?
    Good point. Bad wording on that. What I've been reading is researchers hypothesizing effect sizes for studies, but they offer no reason for that hypothesis. It has to be based on something... I know each study will vary, obviously, but how are they deciding that?

  5. #5
    TS Contributor
    Points: 18,889, Level: 87
    Level completed: 8%, Points required for next Level: 461
    CowboyBear's Avatar
    Location
    New Zealand
    Posts
    2,062
    Thanks
    121
    Thanked 427 Times in 328 Posts

    Re: Effect Sizes

    Welll... Cohen's original effect sizes were based on his rough experience of what was most common in psychology. So hypothesising a "medium" effect size is like hypothesising a typical, expected effect size (although in reality Cohen's medium effect size is a bit higher than the average in psychology - see Richard et al.). But most likely the researcher just can't be bothered coming up with a more specific hypothesis, and is going with convention.
    Matt aka CB | twitter.com/matthewmatix

  6. #6
    Points: 12, Level: 1
    Level completed: 23%, Points required for next Level: 38

    Posts
    3
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Effect Sizes


    Well, that explains how vague all of these "explanations" have been. Thank you! Very helpful!

+ 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