+ Reply to Thread
Results 1 to 7 of 7

Thread: Interaction / Moderation effect linear regression

  1. #1

    Interaction / Moderation effect linear regression




    Hi,

    I have a question regarding a moderation effect with linear regression. I have hypothesized for a significant main effect and another hypothesis for a significant interaction effect. My tutor told me to create 2 models. In the first model without interaction I have one variable (which is hypothesized for significance) that is insignificant and in the second model with interaction, this variable is significant, but the hypothesized interaction term is not. How do I interpret this? Do i reject h1 and reject h2 or do I accept h1 based on the second model and reject h2?

    Gr.

  2. #2
    Omega Contributor
    Points: 39,304, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,099
    Thanks
    405
    Thanked 1,203 Times in 1,164 Posts

    Re: Interaction / Moderation effect linear regression

    Does your second model look like this:

    y-hat = bo + b1 + b2 + b1*b2
    Stop cowardice, ban guns!

  3. #3
    Points: 1,321, Level: 20
    Level completed: 21%, Points required for next Level: 79

    Posts
    9
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Interaction / Moderation effect linear regression

    Yes it does.

  4. #4
    Omega Contributor
    Points: 39,304, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,099
    Thanks
    405
    Thanked 1,203 Times in 1,164 Posts

    Re: Interaction / Moderation effect linear regression

    What is this for? An internal project or something you plan to try to publish?


    So a priori you say hey X! is associated with Y and also X1*X2 is associated with Y. I was checking before to make sure you had the main terms in your interaction based model and you did.


    So the slight possible conundrum would have been that typically if your interaction term is significant, you no longer interpret your main terms, in that their individual interpretation is no longer of interest, since the variable's effect is conditional on another term.


    So X2 is not significant either? I don't see a problem with writing this up and reporting the X1 model and also reporting the interaction model was null. It is important to just report what your original intentions were and to report null results, so other's know your intent.
    Stop cowardice, ban guns!

  5. #5
    Fortran must die
    Points: 62,031, Level: 100
    Level completed: 0%, Points required for next Level: 0
    noetsi's Avatar
    Posts
    6,538
    Thanks
    693
    Thanked 916 Times in 875 Posts

    Re: Interaction / Moderation effect linear regression

    Just because you have interaction does not mean you are not interested in the main effects. It means you are interested in the main effects at some specific level of the other main effect. This is commonly called simple effects in ANOVA.
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

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

    hlsmith (08-08-2017)

  7. #6
    Omega Contributor
    Points: 39,304, Level: 100
    Level completed: 0%, Points required for next Level: 0
    hlsmith's Avatar
    Location
    Not Ames, IA
    Posts
    7,099
    Thanks
    405
    Thanked 1,203 Times in 1,164 Posts

    Re: Interaction / Moderation effect linear regression

    Yes, conditional on the other variable!
    Stop cowardice, ban guns!

  8. #7
    Points: 1,321, Level: 20
    Level completed: 21%, Points required for next Level: 79

    Posts
    9
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Interaction / Moderation effect linear regression


    Thanks a lot. I used your advice (it was for a school assignment)!

    Gr

+ 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