+ Reply to Thread
Results 1 to 6 of 6

Thread: Nine variable - Looking for a relationship

  1. #1
    Points: 564, Level: 11
    Level completed: 28%, Points required for next Level: 36

    Posts
    6
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Nine variable - Looking for a relationship




    Nope, not a dating advert, just looking for advise on some stats.

    I have 9 variables and looking to see if a relationship exists. Looking at histograms a lot of the data is skewed and are negative

    So, my question is, would I be better off just doing it as a non-parametric analysis?

  2. #2
    Omega Contributor
    Points: 39,128, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,085
    Thanks
    402
    Thanked 1,194 Times in 1,155 Posts

    Re: Nine variable - Looking for a relationship

    Provide more information please. So you have 9 continuous variables and you want to see if there is a relationship between any pairwise comparisons of any 2 of them, correct.


    You can just create a correlation matrix, probably use Spearman rank correlation if they are skewed. Remember is you look at enough variable combinations, so will come-up significant by chance (e.g. 5%), so of your 28 comparisons expect some chance.
    Stop cowardice, ban guns!

  3. #3
    Points: 564, Level: 11
    Level completed: 28%, Points required for next Level: 36

    Posts
    6
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Nine variable - Looking for a relationship

    Quote Originally Posted by hlsmith View Post
    Provide more information please. So you have 9 continuous variables and you want to see if there is a relationship between any pairwise comparisons of any 2 of them, correct.


    You can just create a correlation matrix, probably use Spearman rank correlation if they are skewed. Remember is you look at enough variable combinations, so will come-up significant by chance (e.g. 5%), so of your 28 comparisons expect some chance.
    There are yes/no questionnaires
    Likert scale questionnaires
    and measures taken over fixed time periods and performance on other tasks being measured on a scale.

    So there are quite a few tasks and different measures.

    Then of course I am having data such as (data attached) being thrown in to make things
    slightly more complicated
    Attached Images  

  4. #4
    Omega Contributor
    Points: 39,128, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,085
    Thanks
    402
    Thanked 1,194 Times in 1,155 Posts

    Re: Nine variable - Looking for a relationship

    The use of scatterplots would be beneficial.
    Stop cowardice, ban guns!

  5. #5
    Points: 564, Level: 11
    Level completed: 28%, Points required for next Level: 36

    Posts
    6
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Nine variable - Looking for a relationship

    Quote Originally Posted by hlsmith View Post
    The use of scatterplots would be beneficial.

    Used something different due to the relatively small sample size (N=60) so thought it might be a bit difficult to see any relationship using a scatterplot (not that I would see one using a histogram, but wanted to see if I could use pearsons or non-parametric alternative).

  6. #6
    Omega Contributor
    Points: 39,128, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,085
    Thanks
    402
    Thanked 1,194 Times in 1,155 Posts

    Re: Nine variable - Looking for a relationship


    Yeah, non-parametric is Spearman.
    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