2
$\begingroup$

I am planning to run a moderated regression on my dataset with two IVs.

  • One of them is a straightforward interval data-type, between subjects IV (e.g. anxiety level).
  • The other is a categorical data-type, within subjects IV (e.g. type of social situation).
  • DV is outcome scores on a variable (e.g. psychological adjustment).

I am predicting that anxiety level is a moderator of the effect of social situation on psychological adjustment.

My question is:

  • How can I run moderated regression since all participants will have the same scores on the categorical IV, since it is a within-subjects variable. ie. all participants will have 0 for social sitution A, and 1 for social situation B and so on?
  • Am I better off running a two-way mixed design ANOVA?
Improve this question
$\endgroup$
1
  • $\begingroup$ I have made some substantive edits to your question, because I think you were getting your IVs and DVs confused. Please review to check that these changes are correct. $\endgroup$
    – Jeromy Anglim
    Commented Mar 15, 2012 at 4:57

1 Answer 1

Reset to default
1
$\begingroup$

The anxiety by social situation interaction effect is a moderator effect. In simple terms, interaction means moderation. In more subtle terms, I find interaction has connotations of a pure statistical effect, whereas moderation implies a causal role for one for the moderator. Whether you agree with this causal connotation, moderation is typically assessed by whether there is a statistical interaction.

In the case of your example, examining the anxiety by social situation interaction whether it be by looking at a mixed-design ANOVA or seeing whether anxiety predicts the difference score between psychological adjustment in the two social situations will both provide a basic assessment of this interaction effect of interest.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Hi Jeromy, thank you very much for your reply. When you say "seeing whether anxiety predicts the difference score between psychological adjustment in the two social situations", do you mean that in essence, I am using a linear regression: relationship between anxiety and the difference in scores for Social situation A and B? $\endgroup$
    – Hannah
    Commented Mar 18, 2012 at 7:29
  • $\begingroup$ Yes. that's what I mean $\endgroup$
    – Jeromy Anglim
    Commented Mar 18, 2012 at 8:21

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.