Associations between parenting behaviors of support, behavior control and overcontrol, and psychological control/disrespect with adolescent internalizing and externalizing problems have been studied extensively (Barber et al., 2012; Kerr & Stattin, 2000), and also adolescent behaviors of disclosure and secrecy in the context of these problems (Frijns et al., 2010). However, few studies have assessed how parent and child behaviors might moderate each other’s associations with problems (Keijsers et al., 2009). This study investigates interaction effects of the above-mentioned parent and adolescent behaviors when predicting depression, loneliness, and low self-esteem (internalizing), and delinquency, aggression, and drug/alcohol use (externalizing). Given the variety of behaviors and problems under study, it is hypothesized that various kinds of moderation effects will emerge.

An ethnically diverse sample of 1,281 adolescents attending grades 7 to 10 in a Southern Swedish municipality (age 12.5 to 19.3, M = 15.2, SD = 1.2) filled out questionnaires in class. All scales have been published internationally; however, some items were added to short scales. Each of the internalizing and externalizing problems was regressed on all possible combinations of one of the four parenting variables and one of the two adolescent behaviors under study, resulting in 48 regression analyses.

Confirming previous findings, parent psychological control and overcontrol were associated with internalizing and externalizing problems, and behavior control and insufficient support with internalizing problems. Adolescent disclosure predicted low levels of both kinds of problems and secrecy predicted high levels. Two-way interactions of parent and adolescent behaviors added significantly (p < .05) to the variance in 13 of 48 analyses which is beyond chance level (p < .001). In addition to the inspection of significant effects, t-values across all analyses were analyzed in order to distinguish between more general trends and solitary effects on specific internalizing or externalizing problems only. Confirming the hypothesis, interaction effects varied across the combinations of parent and adolescent behaviors (η² = .26) and were further moderated by the distinction between internalizing and externalizing problems (η² = .38). These effects were grouped into five kinds of interaction effects: In mutually enhancing and mutually exacerbating effects, two positive or two negative, respectively, behaviors increased each other’s associations with problem levels. In protection effects, usually adolescents’ behavior reduced associations between negative parenting and problems. Relationship split effects might reflect an alienated parent-adolescent relationship in which negative behaviors cannot do much additional harm. Finally, maintained relationship/sabotage means that the lowest level of problems occurred if one generation maintained the relationship by a positive behavior and the other generation abstained from “sabotaging” it by a negative behavior. Otherwise, problem behaviors increased sharply without the other generation’s behavior having any large effect any longer.

In conclusion, analyses provide ample evidence that adolescents’ behavior moderates links between parents’ behaviors and adolescents’ internalizing and externalizing problems. Possible causal interpretations include adolescents as “gatekeepers” of parenting efforts, families’ functional and dysfunctional adaptations, and parent and child behavior combinations as consequences of internalizing and externalizing problems.