ted to see the effect of the parenting variables on the adolescent marijuana use differ in the two race studied (black and white). The significant interactions include ethnicity by father rules and ethnicity by fathers time spent with child. If father rules are strong, then the adolescent marijuana use in both races will be low. If the father rules are weak, then the adolescent marijuana use is more at risk for white children than for black children. It is highest for the three children that identified themselves as other for ethnicity. The same result applied to father time spent with child. When the fathers time spent with his child is high, the adolescent marijuana use for both races is low. While fathers time spent with the child is low is more risky for white children to use more marijuana than for black children. Finally, father marijuana use is interacted with each of the parenting variables. Only mother identification was significant in the interaction regression with father marijuana use variable. The plot show if the child is strongly identified with the mother, his/her marijuana use will be low regardless if the father uses marijuana or not. When the child shows weak identification with the mother, child marijuana use will increase; more strongly effected in the presence of fathers marijuana use. Therefore mother identification helped buffered the negative effects of the fathers marijuana use risk factor on the childs marijuana use. Discussion The overall information suggest that a strong parent-child relationship helped offset the negative effects of certain risk factors such as father having AIDS and father using marijuana. Specifically, a more affectionate parent-child relationship, characterized by warmth, good discipline, adolescents identification with the parents (mother in particular), and a non-conflict relationship were related to less frequent marijuana use by the child. We can then say that fathers who spends a go...
