Johnson-Neyman procedure

I run an ANCOVA, finding a significant interaction between my treatmente and a covariate. Now I would like to identify the regions of significance in order to state for which values of the covariate my slopes are different. I've triyed "js" function of "probemod" package, but I have some problems.
Message error look like this "Error in cov[interactionterm, interactionterm] : subscript out of bounds".
This is the script that I used:

myModel <- lm(score ~ x * treatt, data=igy)
jnresults <- jn(myModel, dv="score", iv="x", mod="treatt")

Thank you for anyone helping me.


This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.