The Effects of Parenting and Family Functioning on Adolescent Behavior

DESCRIPTION (provided by applicant): The general purpose of this study is to help design more effective family-based prevention interventions for reducing adolescent sexual behavior. Because of many problems with preexisting instruments, the specific aim of this research is to develop and validate a comprehensive and systems based measure of parenting and family functioning that can be reliably and efficiently used in larger studies with diverse American families. The instrument will be developed according to the five steps of Likert's (1932) scaling method. An exhaustive set of items that cover all the constructs of interest will be generated by reviewing existing instruments. Each item will then be rated by a panel of experts. To further reduce the instrument, as well as assess the psychometric properties of each variable, the instrument will be administered to participants and tested in three phases. Items selected for final inclusion will pass a stringent set of criteria at each phase. During the first phase, the initial set of items will be administered online to approximately 300 18-19 year old undergraduate students from various racial backgrounds. To further refine the instrument, focus groups with parents and adolescents will occur during the second phase. During the third phase and main portion of the study, approximately 150 African American and 150 European American adolescents and one of their parents (N=600) will be sampled from a larger pool of eligible students at an economically and racially diverse high school in the greater Chicago area. Equal numbers of males and females will be sampled. Adolescents and parents will complete measures of sexual activity, drug use, psychological health, the new measure of family functioning and preexisting measures of family functioning. The factor structure of the instrument will be tested with variable-level confirmatory factor analyses using standard structural equation modeling procedures. Campbell and Risk's (1959) Multi-Method-Multi-Trait matrix technique will be used to assess the convergent and discriminant validity of the instrument. Possible racial and gender differences, as well as differences between parents and adolescents will also be assessed. Because of many unique features, it is expected that the new measure will be a better predictor of adolescent sexual behavior, especially for African American youth, than other measures. If so, it will have important implications for the design and success of larger studies and family-based prevention interventions.