+ Reply to Thread
Results 1 to 4 of 4

Thread: Correlation b/w continuous and categorical variables in SAS?

  1. #1
    Points: 105, Level: 2
    Level completed: 10%, Points required for next Level: 45

    Posts
    5
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Correlation b/w continuous and categorical variables in SAS?




    I would like to look at the correlation of a hormone level (continuous variable) and a genotype activity (categorical variable with the following levels:COMT-low, COMT-high, and COMT-intermediate). How should I do this in SAS?

    Thanks!

    PS I have done the polyserial correlation but that doesn't make sense to me!

  2. #2
    Omega Contributor
    Points: 38,423, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,005
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Correlation b/w continuous and categorical variables in SAS?

    You might get away with Spearman rank order if the genotype activity can be construed as ordinal (e.g., recoded: 1, 2, 3).


    That would mean using something like


    Code: 
    PROC CORR DATA=yours / SPEARMAN;
    P.S., What is the purpose of this analysis? What are your trying to say? Could regression be used? It took me awhile to tell myself that I can just use linear regression instead of correlations and logistic regression instead of contingency tables sometimes. Which these approaches have more options and at times better interpretability.
    Stop cowardice, ban guns!

  3. #3
    Points: 105, Level: 2
    Level completed: 10%, Points required for next Level: 45

    Posts
    5
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Correlation b/w continuous and categorical variables in SAS?

    Thanks for your reply. I want to look at the correlation between methylated estrogen levels (continuous variable) and Catechol-O-methyltransferase (COMT) genotype activity. COMT activity can high, intermediate, or low activity.
    It looks like my most appropriate option here is to treat COMT activity as an ordinal variable and do a spearman corr test.



    Quote Originally Posted by hlsmith View Post
    You might get away with Spearman rank order if the genotype activity can be construed as ordinal (e.g., recoded: 1, 2, 3).


    That would mean using something like


    Code: 
    PROC CORR DATA=yours / SPEARMAN;
    P.S., What is the purpose of this analysis? What are your trying to say? Could regression be used? It took me awhile to tell myself that I can just use linear regression instead of correlations and logistic regression instead of contingency tables sometimes. Which these approaches have more options and at times better interpretability.

  4. #4
    Omega Contributor
    Points: 38,423, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,005
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: Correlation b/w continuous and categorical variables in SAS?


    That seems better than comparing methyl-estrog levels between genotypes for your question? You can put the genotypes into a linear model and try run a trend test as well:


    http://support.sas.com/kb/22/912.html
    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