+ Reply to Thread
Results 1 to 3 of 3

Thread: ROC plotting

  1. #1
    Points: 528, Level: 10
    Level completed: 56%, Points required for next Level: 22

    Posts
    3
    Thanks
    1
    Thanked 0 Times in 0 Posts

    ROC plotting




    I am a true novice of statistics, so would be grateful of easy explanations, as I am trying to create an ROC curve, and understand how to do it conventionally. However I want to display 3 continuous variables as 1 continuous variable instead on my ROC curve. For example a profile signature of 3 continuous variables (eg. progesterone, oestrogen and testosterone) in unison is associated with diabetes. How do I combine 3 variables together to display on the ROC curve as one profile signiture? I considered doing a mean average of the 3 variables, but the range of the values between the different variables are very large (eg. progesterone= 0.1 mg/ml range, oestrogen= 10,000 mg/ml range, testerone=100mg/ml range), and therefore doesnt really do the variables justice. Is there any other choices I have?

  2. #2
    Devorador de queso
    Points: 95,889, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,937
    Thanks
    307
    Thanked 2,630 Times in 2,246 Posts

    Re: ROC plotting

    I'm not sure what you mean by that. What do you think a ROC curve tells you?
    I don't have emotions and sometimes that makes me very sad.

  3. #3
    Omega Contributor
    Points: 38,396, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,001
    Thanks
    398
    Thanked 1,186 Times in 1,147 Posts

    Re: ROC plotting


    I know a program would do it for you (e.g., SAS), but in my mind I cannot conceptualize how to do it solely by hand work. You would plot everypersons sen and 1-spec but what if progesterone is high and testos is low then the next person the progesterone is high and so is testosterone but both have diabetes. You would use a classification table but unsure how you would decide who falls below the cut off.

    What I may do is generate the beta coefficients from the logistic model the plug into the model and see if any levels preset themselves.
    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