Paradoxically, knowledge that parents posses about their adolescent children’s activities declines with age, but low levels of knowledge are associated with externalizing and internalizing problems. Might there only be a small group of adolescents with steeply declining parental knowledge? Or, are interindividual differences in knowledge and its normative decline independent of each other? This study will explore different trajectories of knowledge in order to answer this question.

Second, why is low parental knowledge associated with adolescent problems? Focusing on internalizing problems, does parental knowledge really predict them over time, or do they reduce parental knowledge, for example because a depressed or unconfident adolescent tends to withdraw from conversation? This study will determine the direction of effects.

Third, if parental knowledge predicts internalizing problems, why is this so? Previous studies suggest that both knowledge and internalizing might result from family interaction processes (Kerr & Stattin, 2000), but the same results could also be read as mediation from knowledge via family interactions to internalizing. Furthermore, knowledge was only partly explained by parent-adolescent interaction processes, lending doubt to the interpretation of parental knowledge as a mere expression of them (Barber, 2005). Thus, parental knowledge might either be an indicator of parent-adolescent communication or a causal factor in its own. This study will contribute to clarification. Aversive parental behaviors and adolescent non-disclosure and oppositional behavior were chosen as predictors because they belong to problematic parent-adolescent interactions and because of their links to adolescent internalizing problems.

A representative Swedish community sample of 1,744 adolescents of age 10-14 at T0 was re-assessed at four annual occasions T1-4. Each year, adolescents filled out questionnaires at school.

Using Growth Mixture Modeling, three trajectories of parental knowledge, and two trajectories each of self-esteem and depression were revealed across T1-4. The three knowledge trajectories differed in level, but each trajectory had virtually the same age decline.

In all subsequent analyses, the effects of predictor variables at T0 on T1-4 trajectories of either knowledge or depression, or self-esteem were tested, above and beyond the stability of the respective dependent variable since T0. These analyses revealed effects of parental knowledge on trajectories of depression and self-esteem, but not vice versa.

A conceptual model was concluded from a series of analyses including parent-adolescent interaction variables. If parents exerted aversive behaviors such as being harsh or making fun of their children, these disclosed not much information and behaved oppositional which in turn predicted low levels of parental knowledge. Although knowledge had predicted adolescent depression and low self-esteem when entered in the analyses alone, it did not consistently predict these variables if adolescents’ opposition and non-disclosure were taken into account.

In conclusion, the normative decline of parental knowledge and interindividual differences are two independent phenomena which might have different causes. This study has contributed to an understanding of how parent-adolescent interactions lead to interindividual differences in knowledge. Low levels of knowledge were not a consistent causal factor for adolescent internalizing symptoms, but clearly indicated parent-adolescent problems.