help deciphering - hierarchical regression report


I've just started by statistical analysis stage and I've had some strange outputs I was hoping someone could help me with.

Two personality measures were entered into a hierarchical regression model and both were found to be significant predictors of heart rate with: N (β = 2.15, p = 0.001), E (β = 1.70, p < 0.01).

However, the strange part is when I entered an interaction of the two dimensions, I got a significant but negative beta value at: (β = -1.57, p < 0.01).

Does anyone have an explanation for why this might have happened?

Estelle x


Less is more. Stay pure. Stay poor.
N and E are your two predictors? What level (cluster) do you have random intercepts for in this model?