+ Reply to Thread
Results 1 to 4 of 4

Thread: Question on Effect modification analysis

  1. #1
    Points: 666, Level: 13
    Level completed: 32%, Points required for next Level: 34

    Posts
    11
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Question on Effect modification analysis




    Dear All
    I wonder if you kindly would assist me on the basic concept in examining the effect modification.
    The live example is:
    I want to analyze the effect of gender on weight loss after different interventions. What would be the best way to perform it?
    Whether what I did is correct? I split the file according to the type of intervention, I got t test comparing weight loss between genders. then I tried to assess the confounder effect which could be the preoperative weight, so I did a simple linear regression saving the unstandardized residual errors (preoperative weight is the predictor, and weight loss is the outcome) then I compared the residuals.
    I quoted this from way based on the concept that this is the same as partial correlation.

    I wish the answer be easy for me to apply.
    Thanks in advance
    Last edited by drmohamedbekheit; 09-12-2013 at 08:36 AM.

  2. #2
    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: Question on Effect modification analysis

    There are different schools of thought on modeling differencs, can't remember all of the nuances. Another option would be to just have post-op weight as the dependent variable with pre-op and gender in the model.
    Stop cowardice, ban guns!

  3. #3
    Points: 666, Level: 13
    Level completed: 32%, Points required for next Level: 34

    Posts
    11
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Question on Effect modification analysis

    Thank you
    Indeed your reply was useful too

    There are some differences in the output significance (some were stat significant, became significant in this analysis)
    Does this mean that one of the methods is incorrect or we just could elicit different results using different tests.

    many thanks once again

  4. #4
    Points: 1,050, Level: 17
    Level completed: 50%, Points required for next Level: 50

    Posts
    17
    Thanks
    1
    Thanked 4 Times in 4 Posts

    Re: Question on Effect modification analysis


    The way I tend to consider measuring effect modification in regression is stratifying by gender like you did however you then compare the MoA and CI between the aggregate outcome MoA. No EM means both gender MoA's are similar (enough) to the aggregate. Do that between and among each intervention.

+ 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