+ Reply to Thread
Results 1 to 4 of 4

Thread: Which regression should I use?

  1. #1
    Points: 60, Level: 1
    Level completed: 20%, Points required for next Level: 40

    Posts
    4
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Which regression should I use?




    I am trying to analyze the following research question.

    I hypothesize that smoking status increases risk of heart attack but this effect is modified by the level of a certain hormone. There is no dose-dependent relationship between the level of this hormone and the effect. For example, very low or very high level of this hormone are detrimental but normal values are protective.

    This is the way I coded my variables:

    Smoking status: 3 levels (former, current, never)
    Heart attack: 2 levels (yes/no)
    Hormone level (0-100, 101-200; 201-300 etc.)

    I also want to control for age, gender and some other variables.

    I was thinking about stepwise logistic regression but I might be wrong. I am using SPSS

    I appreciate your help!

  2. #2
    Fortran must die
    Points: 58,790, Level: 100
    Level completed: 0%, Points required for next Level: 0
    noetsi's Avatar
    Posts
    6,532
    Thanks
    692
    Thanked 915 Times in 874 Posts

    Re: Which regression should I use?

    Since your dependent variable heart attack has two levels I think you would use logistic regression. Smoking status and hormone levels would normally (given your coding scheme) be used as dummy predictors.

    Most accounts I have read agree that stepwise is not a good approach. It depends too much on chance results (which might vary by sample) especially when the predictors are correlated.
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

  3. The Following User Says Thank You to noetsi For This Useful Post:

    vicvan (12-06-2013)

  4. #3
    Devorador de queso
    Points: 95,540, Level: 100
    Level completed: 0%, Points required for next Level: 0
    Awards:
    Posting AwardCommunity AwardDiscussion EnderFrequent Poster
    Dason's Avatar
    Location
    Tampa, FL
    Posts
    12,930
    Thanks
    307
    Thanked 2,629 Times in 2,245 Posts

    Re: Which regression should I use?

    Also if you explicitly want to control for variables then you want to make sure they're in the model regardless of how 'significant' they end up being. Stepwise isn't a good way to do that.
    I don't have emotions and sometimes that makes me very sad.

  5. The Following User Says Thank You to Dason For This Useful Post:

    vicvan (12-06-2013)

  6. #4
    Points: 60, Level: 1
    Level completed: 20%, Points required for next Level: 40

    Posts
    4
    Thanks
    3
    Thanked 0 Times in 0 Posts

    Re: Which regression should I use?


    Thank you for your response. Thanks God SPSS creates dummy variables but since I will have 10 levels of hormone and other covariates, my concern is that the analysis is going to have many iterations. Any advise on that? Thanks again. Sean

+ 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