Posted by
Ian Martin-2 on
Sep 02, 2014; 6:03pm
URL: http://spssx-discussion.165.s1.nabble.com/data-post-age-correction-tp5727132p5727134.html
Arjan,
You do not say if the ANCOVA is defensible. Did you check the treatmentXcovariate interaction term? Without homogeneity of slopes, the adjustment is meaningless. You said age differs between groups, and this is also potentially a problem if the differences are significant.
Was Levene's test OK too?
I have seen -- too often -- situations where researchers blithely "adjust" for many covariates without first checking to see if it is possible.
regards,
Ian
On Sep 2, 2014, at 5:49 AM, Arianm <
[hidden email]> wrote:
> Dear all,
> I have a question about how to show my new data post-age correction.
> So imagine, protein A is different between group 1, 2 and 3. However these
> groups differ in age, so we have to perform age corrections (I do this with
> GLM--> univariate --> covariate: age)
> now my question is how to depict the data after this age correction in a box
> plot? so one of the methods I was considering was to save: unstanderdized
> predicte values, and then do the box plot with the new saved data. However i
> am not certain if this method is correct, and if not, how should I perform
> this task?
>
> Thank you in advanced
>
> Kind regards
> Arjan
>
>
>
> --
> View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/data-post-age-correction-tp5727132.html> Sent from the SPSSX Discussion mailing list archive at Nabble.com.
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
>
[hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD