Moderation analysis SPSS - dummy control variables

#1
Hi all,

I am conducting a moderation analysis in SPSS for which I have standardized my IV's, DV and continuous control variables. My model also employs year-effects, industry-effects and a number of dummy control variables. My question is; do I also have to standardize these variables? Why/why not?

Many thanks in advance!
 

Karabiner

TS Contributor
#2
If you mean center a dummy-variable, what for example should "male gender = -0.65, female gender = 0.35" mean?

With kind regards

Karabiner
 
#3
If you mean center a dummy-variable, what for example should "male gender = -0.65, female gender = 0.35" mean?

With kind regards

Karabiner
My bad, I should have clarified. I only use dummy variables whereby a "1" means the presence of that particular item, and a "0" indicates an absence. For my year effects, a "1" is assigned when a particular observation is from that year.

Jgroup J2018 J2019 J2020 J2021
2019 0 1 0 0

My instructor posted this in his instructions: "If you want to test for the impact of the moderator on the main effect, always use the standardized version of all independent variables (X’s) (including control variables)."

However I was wondering if this meant including all CONTINUOUS control variables, as I don't understand what sense it would make to standardize a binary variable.
 

Karabiner

TS Contributor
#4
Yes, I know what a dummy-variable is. And if you standardize (mean center)
a dummy variable, you get something like -0.65/+0.35 instead of 0/1 (absence/
presence). So he (hopefully) meant to standardize only the continuous variables.

With kind regards

Karabiner
 
Last edited: