ANCOVA: how to recreate equation based on SPSS output?

[Hrag]

Hi all,

I performed an ANCOVA with 3 covariates (Analyze>GLM>Univariate…). I would like to use the output that SPSS provides to calculate the adjusted values for my independent variable for each individual. I am trying to do this by figuring out the equation that SPSS uses to calculate the estimated adjusted means. Any help on how to do this would be greatly appreciated!

What I've done until now: I clicked on the "parameter estimates" options, which provided me with B values for my covariates and the intercept value. The problem is that SPSS is also giving me a B value for all levels of my dependent variable, except for the last variable. It gives me the note: "This parameter is set to zero because it is redundant".

Thanks in advance for any help you may provide,
-H

 

[Dragan]

What you need to do is compute the adjusted means based on "dummy coding" i.e. 1 and 0's That is, use your B values (parameter estimates) and the overall means of your covariates in the equation (with "dummy codes") and that will give you the adjusted means.

 

[Hrag]

What you need to do is compute the adjusted means based on "dummy coding" i.e. 1 and 0's That is, use your B values (parameter estimates) and the overall means of your covariates in the equation (with "dummy codes") and that will give you the adjusted means.

Hi Dragan,

Thanks for the help. I'm still running into the same problem with the message "this parameter is set to zero because it is redundant" for the last level of my independent variable. Could you please give me more detailed step-by-step instructions? I'm wondering if I'm doing something wrong along the way. Thanks in advance!

 

[Dragan]

Hi Dragan,

Thanks for the help. I'm still running into the same problem with the message "this parameter is set to zero because it is redundant" for the last level of my independent variable. Could you please give me more detailed step-by-step instructions? I'm wondering if I'm doing something wrong along the way. Thanks in advance!

It's giving you that information because the degrees of freedom are the number of groups - 1. In short, just ignore it.

Example: Suppose I conducted an ANCOVA with 4 groups with 3 covariates.

The equation to get the adjusted means is as follows:

Y_bar = Intercept + B1*D1 + B2*D2 +B3*D3 + B4*XBar_1 + B5*XBar_2 + B6*X_Bar_3

As such, if you want the adjusted mean for group 1, then you set D1=1, D2=0, and D3=0

Iif you want the adjusted mean for group 2, then you set D1=0, D2=1, and D3=0

Iif you want the adjusted mean for group 3, then you set D1=0, D2=0, and D3=1.

Iif you want the adjusted mean for group 4, then you set D1=0, D2=0, and D3=0.