SPSS Statistics

 View Only

  • 1.  Writing the Binary Logistic Regression Equation

    Posted Fri November 25, 2022 04:49 PM
    Reply
    Hello!

    I am not sure how to write the binary logistic expression (and determine the probability)
    when I have only one IV, one reference category, and two comparative categories:

    Biological sex (0)

     

    Biological sex(1)

    -.055

    Biological sex(2)

    -.751

    Constant

    .978

    Would that be the constant for the reference group: Li=.978  ?

    For the first comparison group, it would then be:

    Li=B0+B1X1= .978+(-.055) x 1 = 0.923

    I will also need to determine eLi before I move on to the probability calculation. 
    (I tried to follow an example that I found online, but it does not explain how to calculate eLi;

    Li = (−1.791) + (.016)∙(30) + (0.530)∙(1) = −.781

    eLi = 0.458

    p1= 0.458 ÷ (1 + 0.458) = .314).

    Thank you for your input. 



    ------------------------------
    A MC
    ------------------------------

    #SPSSStatistics


  • 2.  RE: Writing the Binary Logistic Regression Equation

    Posted Fri November 25, 2022 04:59 PM
    Reply
    Just save the probabilities using the LOGISTIC REGRESSION command.  You can then plot that against your sex variable (or others).  The exponentiated coefficients are part of the regular output.

    Note that you can enhance the plot using the Chartbuilder and GGRAPH by added the binary dv with an additional ELEMENT statement (make sure that variable measurement level is set to scale).

    --



    Original Message


Related Content