+ Reply to Thread
Results 1 to 9 of 9

Thread: Multiple Linear Regression - really need help

  1. #1
    Points: 131, Level: 2
    Level completed: 62%, Points required for next Level: 19

    Posts
    12
    Thanks
    8
    Thanked 0 Times in 0 Posts

    Multiple Linear Regression - really need help




    Hi, I'm going out of my mind...really need help.

    I have one continuous dependent variable (test scores) and 6 categorical (with 2 categories - e.g. male/female) independent variables.

    I want to look at the association/relationship between them or does the independent variables have an effect on the dependent.

    What test can I use? I have tried Multiple Linear Regression but the data doesn't pass any of the asssumptions.

    My deadline is so soon so please help..Thank you so much in advance

  2. #2
    TS Contributor
    Points: 17,749, Level: 84
    Level completed: 80%, Points required for next Level: 101
    Karabiner's Avatar
    Location
    FC Schalke 04, Germany
    Posts
    2,540
    Thanks
    56
    Thanked 640 Times in 602 Posts

    Re: Multiple Linear Regression - really need help

    What test can I use? I have tried Multiple Linear Regression but the data doesn't pass any of the asssumptions.
    What do you mean by this?
    And how large is your sample size?

    With kind regards

    K.

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

    101lou (04-02-2014)

  4. #3
    Points: 131, Level: 2
    Level completed: 62%, Points required for next Level: 19

    Posts
    12
    Thanks
    8
    Thanked 0 Times in 0 Posts

    Re: Multiple Linear Regression - really need help

    There are 68 participants.
    I have tried multiple linear regression, i read their are 8 assumptions the data has to pass before you can run the test? linearity, normal distribution etc except their isn't a linear relationship between my dependent and independent variables or they are not normally distributed

  5. #4
    Points: 131, Level: 2
    Level completed: 62%, Points required for next Level: 19

    Posts
    12
    Thanks
    8
    Thanked 0 Times in 0 Posts

    Re: Multiple Linear Regression - really need help

    initially i want to use chi square but the dependent variable would have to be categorical and mine is continuous (can't be categorised)

  6. #5
    Points: 131, Level: 2
    Level completed: 62%, Points required for next Level: 19

    Posts
    12
    Thanks
    8
    Thanked 0 Times in 0 Posts

    Re: Multiple Linear Regression - really need help

    any help please?

  7. #6
    TS Contributor
    Points: 17,749, Level: 84
    Level completed: 80%, Points required for next Level: 101
    Karabiner's Avatar
    Location
    FC Schalke 04, Germany
    Posts
    2,540
    Thanks
    56
    Thanked 640 Times in 602 Posts

    Re: Multiple Linear Regression - really need help

    their isn't a linear relationship between my dependent and independent variables
    You said you had 6 binary variables, so the bivariate
    relationships cannot be non-linear.
    or they are not normally distributed
    I am not sure what you mean by this. You have
    to check the residuals of the multiple regression;
    the residuals should be from a normal population
    (but with n > 50 departures from this are not
    cosidered serious).

    With kind regards

    K.

  8. The Following User Says Thank You to Karabiner For This Useful Post:

    101lou (04-03-2014)

  9. #7
    Points: 131, Level: 2
    Level completed: 62%, Points required for next Level: 19

    Posts
    12
    Thanks
    8
    Thanked 0 Times in 0 Posts

    Re: Multiple Linear Regression - really need help

    OK thanks! So I am doing the right test by doing Multiple Linear Regression?

  10. #8
    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: Multiple Linear Regression - really need help

    I assume you used dummy independent variables. Ordinal independent dvariables are I believe always linear and I assume this includes dummies. Why do you conclude they are not linear?
    "Very few theories have been abandoned because they were found to be invalid on the basis of empirical evidence...." Spanos, 1995

  11. #9
    Points: 1,503, Level: 22
    Level completed: 3%, Points required for next Level: 97

    Posts
    26
    Thanks
    0
    Thanked 0 Times in 0 Posts

    Re: Multiple Linear Regression - really need help


    Noetsi can you clarify your last answer, I have run into this problem as well. I was always told that the relationship between the Dependent variable and the independent variable had to be linear... If you have a dummy variable (say male =1 female=0) then by definition you are comparing groups and the variable are not linearly correlated). it is therefore assumed that since dummy variables basically split the data into groups, they are linerally related to the dependent variable?


    On a related question, does the Independent variable have to be normally distributed related to the dependent variable? meaning if one variable was age and another was income, what if income was FAR more variable at young ages than it is at old ages (it is). SO what is you have heteroscetastic data in a scatter plot of Age vs Income? Does that matter, or is it only the normal assumption of the residuals that matter?

    Thanks so much!

+ 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