+ Reply to Thread
Results 1 to 4 of 4

Thread: Paired samples with competition

  1. #1
    Points: 2,005, Level: 27
    Level completed: 4%, Points required for next Level: 145

    Posts
    32
    Thanks
    1
    Thanked 0 Times in 0 Posts

    Paired samples with competition




    Hi all,

    I have abit of a puzzle that at first may seem straightforward but I believe there is a complication.

    I have a paired design with each subject being a pot containing 2 plants, 1 treated and 1 untreated. Note that the treatments are applied before planting so they so not run onto the other plant. The problem is that the two plants in the pot compete. My understanding is that a paired t-test makes the assumption that the two plants in the same pot are not independent but I also believe that a paired t-test assumes that the result for one plant does not directly affect the result for the other. Any suggestions as to whether this is a valid layout and if so what analysis can be done?

    Thanks

  2. #2
    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: Paired samples with competition

    hi,
    It does seem a bit problematic, but I think it could be ok in the end. If there is no reason to suspect that the treated ones are sytematically better/worse then the untreated ones the effect of the competition should average out.

    regards
    rogojel

  3. #3
    Points: 2,005, Level: 27
    Level completed: 4%, Points required for next Level: 145

    Posts
    32
    Thanks
    1
    Thanked 0 Times in 0 Posts

    Re: Paired samples with competition

    Thanks Rogojel. The key objective is to compare treated and untreated though so if there was a true effect then it wouldn't cancel out.

    Emily

  4. #4
    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: Paired samples with competition


    Quote Originally Posted by emily.stats View Post
    Thanks Rogojel. The key objective is to compare treated and untreated though so if there was a true effect then it wouldn't cancel out.

    Emily
    hi,
    you mean the treatment effect will not cancel out. The effect of the competition on the other hand will, so the experiment should be ok.

    regards
    rogojel
    Last edited by rogojel; 07-30-2013 at 10:50 AM. Reason: typo

+ 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