5
$\begingroup$

I am testing a model with a moderator variable in the PROCESS macro available in SPSS. The results are as follows:

enter image description here

So, the effect X*M is positive, which means that as M increases, the impact of X on Y should increase. However, when I plot the result, I get the opposite:

enter image description here

In the image PRIVSCO2 = X VULN_AFF=Y DVP = M

Any idea on how to make sense of these results ?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
5
$\begingroup$

The interaction means that the predicted value of Y should go up faster for X = 1 than for X = 0. That is exactly what you see.

You can see this by writing out the whole equation:

$Y = 6.26 - .89X + .23M + .26XM - .22V$

Then substituting different values for M. You didn't give any range or mean for M, but suppose M = 0, then this becomes

$Y = 6.26 - .89X - .22V$

but if M = 1 then

$Y = 6.26 - .89X + 2.3M + .26X - .22V$

or, simplifying

$Y = 6.26 - .63X + 2.3M - .22V$

so the effect of X is "greater" in that it is closer to 0.

Improve this answer
$\endgroup$
4
  • $\begingroup$ Dear Peter, I thank you very much for your help. But I'm not 100% sure to understand your response. As I see it (let me know where I'm wrong), as X*M is positive, an increase in th M variable should result in a stronger impact of X on Y. So, if I use your example, the diference of Ys between X=0 and x=1 should be more important when M is 1 vs. 0. So, the two lines should not converge, but rather diverge as M (DVP) increases... $\endgroup$ Commented Mar 14 at 10:41
  • $\begingroup$ Your error is with the word "stronger". The main effect of X is negative. It becomes less negative when M is bigger. That's weaker, not stronger. $\endgroup$ Commented Mar 14 at 10:58
  • $\begingroup$ OK, got it! Thanks a thousand times for these explanations. You're the best! $\endgroup$ Commented Mar 14 at 12:40
  • $\begingroup$ You're welcome. If my answer is sufficient for your needs you can accept it by clicking the check mark. $\endgroup$ Commented Mar 14 at 12:54

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.