Plotting categorical by continuous interactions from a mixed linear model

[Pol]

Hello,

I am fitting a mixed linear model (using SPSS) and I am trying to graph the interactions between a categorical (factor in my design) and a continuous variable. I tried hard to find a solution to this but was not really successful. Any ideas?

Thanks!

[Bruce Weaver]

Use the /SAVE sub-command to save fitted values, then make scatter-plot with X = continuous predictor variable, Y = fitted value, and different colours (or markers) for the levels of the categorical variable.

centaur wrote

Hello,

I am fitting a mixed linear model (using SPSS) and I am trying to graph the interactions between a categorical (factor in my design) and a continuous variable. I tried hard to find a solution to this but was not really successful. Any ideas?

Thanks!

[Pol]

Thanks! This worked perfect! :)

[mariskakret]

Hi,


I would like to plot a three-way interaction from a mixed linear model with multiple levels (subjects and each participant has multiple trials per condition).

I have:
-one between subject variable (group: treatment and placebo group)
-one within subject variable with three categories called 'face type'
-one continous centred variable (questionnaire score)

The dependent variable is a 5-point scale which is treated as continuous.

Ideally, I want to make two plots of the interaction between group * face type where the continuous variable is fixed once at minus 1 standard deviation and once at plus 1 standard deviation.

How can I achieve this?



best wishes,
Mariska
PS I know the SPSS command for the generalized linear mixed model for it, but not for the general linear mixed model.

This works for generalized:
 /EMMEANS_OPTIONS  CONTROL=  continuousvar (-0.2)  SCALE=ORIGINAL PADJUST=LSD. (here 0.2 is the standard deviation)