I did a moderation analysis, and I see that, when my X value is zero, the Y value is higher at the low level of the moderator than at the high value of the moderator. I would like to test whether the difference in these two Y values is statistically significant. How would I test this?
I don't know what to try first.