Research is notably lacking on how fathers impact the mental health and behavioral problems of their adolescent children, though studies have begun to show that fathersdo influence adolescents in important ways. Particularly understudied, but at high risk, are children with stepfathers and children of Mexican American heritage. A Conceptual Model is developed that emphasizes the construct of "meanings" that children give to fathering behaviors. The meanings investigated derive from the attribution theories and the working models and scripts tradition, which focus on the adolescent's schema or representation of important social-emotional aspects of the father-child relationship.
The Model gives rise to 6 Specific Aims: (1) identify father behaviors influencing child mental health and behavioral problems (net of mothering and other controls); (2) identify social/contextual variables predicting father behaviors; (3) identify "meanings" fathering holds for children that influence mental health; (4) identify social/contextual variables predicting children's meanings; (5) determine how meanings mediate between fathering behaviors and child mental health; and (6) determine how the above factors and relationships may differ (or be moderated) by children's genders, and two father-types (birth-father or stepfather).
PAYS is a longitudinal study, at two sites, Phoenix, AZ (Arizona State University site) and Riverside-San Bernardino, CA (University of California-Riverside site) that includes 392 families, each to be studied for three waves over 4 years. Currently we are collecting the second Wave, with the third Wave scheduled for data collection in 2006. Families are approximately evenly divided into two ethnicities (Mexican American and European American), two child genders, and two father-types (birth-father and stepfather). Families were recruited through schools when the child was in 7th grade. Multi-agent reporting is used, with Spanish-speaking respondents interviewed in Spanish. The study uses many established, adapted and new measures. Most of the proposed "meanings" measures have been designed specifically for this study, successfully pre-tested, and include narrative methods. A 4-cohort sequential design is employed and the Specific Aims are analyzed with Latent Growth Models.