Skip to main content
. Author manuscript; available in PMC: 2015 Sep 30.
Published in final edited form as: Psychol Addict Behav. 2015 Sep;29(3):590–602. doi: 10.1037/adb0000109

Figure 1.

Figure 1

Sample cross-lag structural regression model. Hypothesized relations were tested with two multiple regression models. Heavier lines indicate the two paths of primary theoretical interest assessing the cross-lagged effect of initial temperament on subsequent psychopathology (β2) and the degree to which initial marijuana use moderates the cross-lagged effect of initial temperament on subsequent psychopathology (β3).