0
$\begingroup$

I hope you can help me with these questions.

I am developing a mediation analysis using the mediation package in R. I need to analyze the influence of perceived learning (Mediating Variable –MV-) in the generation of pro-environmental outcomes (Dependent Variables DV).

I have different Independent Variables (IV) to test. One is numeric and another one is categorical. I have already developed the mediating analysis with the numeric IV and I got the results.

However, the motivation IV I have is categorical and has 4 values (Learning, Adventure, Wild, Social). Do you know if creating dummies for the categorical IV is a correct procedure to then run the mediation analysis? This should be like this:

Regression to see the relation between IV vs. DV.

Regression to see the relation between MV vs. DV.

Mediation analysis to see the relation IV+DV vs. DV.

If the above is not possible, do you think that ANOVAs could be used as a substitute for regression models for mediation analysis?

If yes, is it correct to do the following?

ANOVA to see the relation between IV vs. DV.

ANOVA to see the relation between MV vs. DV.

But… How to analyze the relation: IV+MV vs. DV?

Thanks in advance for your advice.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

I strongly recommend that you check out Preacher & Haye's (2014) article

Hayes, A.F. and Preacher, K.J., 2014. Statistical mediation analysis with a multicategorical independent variable. British journal of mathematical and statistical psychology, 67(3), pp.451-470.

It's openly available on Andrew Hayes' website:: http://quantpsy.org/pubs/hayes_preacher_2014.pdf

It's pretty long but it does address your problem exactly. Feel free to jump to section 2.1. When X is a multicategorical variable

Improve this answer
$\endgroup$
1
  • $\begingroup$ Please briefly summarize what the linked document had to say about this particular question, as the answer would be much more valuable then particularly for future visitors to this site. Links to documents can break or the documents themselves can disappear. Just came across an answer on this site earlier today for which both of 2 such links were broken so that the answer was no longer helpful. Whatever you write as part of an answer, on the other hand, should be available so long as this web site or a successor site exists. $\endgroup$
    – EdM
    Commented Sep 3, 2020 at 20:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.