1
$\begingroup$

I have found a significant interaction with two continuous variables. I proceeded to create simple slopes graphs as recommended by Preacher (i.e., - 1 SD, mean, and +1SD slopes). What exactly to these different lines mean? I understand the mean - but what do the -1 SD and +1SD slopes represent - have we created three different groups (low group, medium group, and high group)? Is the +1SD line representing an individual? Would like any clearing up on this! Thanks.

I understand the interpretation of an interaction etc. Just want some intuition on how to explain the graph (simple slopes) to someone who keeps thinking I took the data set and divided it into tertiles.

$\endgroup$
2
$\begingroup$

The simple slope for -1SD is the model predicted slope for someone with a value of the moderator one standard deviation below the mean of the moderator. The simple slope for +1SD is the model predicted slope for someone with a value of the moderator one standard deviation above the mean of the moderator. You're not creating new groups; you're computing the predicted slope for someone with a given value of the moderator. This comes directly from the one regression model you run.

$\endgroup$
  • $\begingroup$ Thanks. And what difference does it make that I pick one variable as the moderator over the other? Is this purely a theoretical choice. $\endgroup$ – mindhabits Aug 21 '19 at 19:40
  • 1
    $\begingroup$ Yes, this is a theoretical choice. If one variable is more like a manipulable treatment and the other is more like a background characteristic, the latter would be a good choice as the moderator. $\endgroup$ – Noah Aug 21 '19 at 20:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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