Margaret Kerr

Margaret Kerr, margaret.kerr@oru.se Örebro University

Beyond parenting: Toward a general model of reciprocal parent-youth interactions
Projekttid: 2005-2011 The purpose of this project was to do a series of empirical studies aimed at testing new theoretical approaches to the issue of parenting and adolescent problem behavior. When we began this project, the defining feature of virtually allempirical studies of parenting and adolescent problem behavior was that they assumed a causal direction from parenting to adolescent behavior. They assumed that parents’ behaviors produced adolescent behavior, but were not affected by it. We had challenged this view in our earlier work on parental monitoring (Kerr & Stattin, 2000, 2003a, 2003b; Kerr, Stattin, & Trost, 1999; Stattin & Kerr, 2000),and our goal in this project was to continue to do so with theoretically based examinations of adolescents’ effects on parents and bi-directional or transactional processes between parents and adolescents. We used data from a longitudinal study financed by Vetenskapsrådet (dnr 345-2001-6661; 421-2003-1522). The longitudinal study dealt with the development of criminality and other problembehaviors. Concerning family, a unique feature of the study was the inclusion of measures to capture parents’ reactions to problem behavior as well as their proactive strategies to manage it. Thus, our aims were to use these longitudinal data to challenge the view of parenting as a unidirectional process of parental influence on adolescents and test some more bidirectional and transactional theoreticalmodels. THREE MAIN RESULTS OF THE PROJECT One main result of this project was that in models testing for bidirectional effects across a number of different adolescent behaviors and parental behaviors, parental behaviors seem to be better interpreted as reactions to adolescent problem behaviors than causes of it. We have looked at various problem behaviors: delinquency (Kerr, Stattin, & Burk, 2010;Kerr, Stattin, & Pakalniskeine, 2008; Stattin, Kerr, & Tilton-Weaver, 2010), social anxiety (Van Zalk & Kerr, in press), and psychopathic personality traits (Salihovic, Kerr, Pakalniskeine, & Özdemir, 2011). Concerning delinquency, parents seem to react more to negative behavior at home than to delinquency, itself, and their reactions are both to reduce attempts to monitor or control the youth andto withdraw emotionally (Kerr et al., 2008). Taken together, then, these results suggest more influence of adolescents on parental behavior than has previously been recognized in adolescent theory and research. A second main result of the project was the emergence of negative parental behavior as more important than monitoring in predicting changes in delinquency. Specifically, higher levels ofparental monitoring did not predict less delinquency over time (Kerr et al., 2010; Kerr et al., 2008; Stattin et al., 2010), but behaviors such as angry outbursts and emotional coldness or rejection were linked to increases in delinquency over time (Kerr et al., 2008). These results contradict claims from the large literature on parental monitoring, of which many studies have been cross-sectional andmost have used measures of parental monitoring that are questionable because they are made up of items that have been shown to represent adolescent more than parental behavior (Kerr & Stattin, 2000; Stattin & Kerr, 2000). For us,
1

the negative parenting findings are central to thinking about what parents can best do to help their youths navigate adolescence successfully. A third main resultof the project was that theoretically based models of parents’ reactions to adolescent behaviors and transactional processes were supported across a range of parent and adolescent behaviors. We have tested a number of new theoretical explanations for these links between adolescent problem behaviors and parenting behaviors. Our results show support for our Context-choice theory (Persson, Kerr, &...