Direct observation and assessment of family members' interactions with one another is being increasingly recognized as an important method for empirically investigating the social context of individual development. Longitudinal analyses of observational data and its relation to individual development, however, are rare. As a result, there are few models for analytic techniques that can assess changing patterns of interaction using multivariate, categorical data over several time points. The majority of statistical analyses of observational data are ill-fitted to the task of describing the complexity of behavioral relationships among family members. This research will include the development of an application of log-linear techniques to the analysis of longitudinal data on the family interaction patterns of adolescents. The relationship of these changing patterns to adolescents social and personality development and adaptive functioning will be assessed. In addition, a collaborative study using two longitudinal data sets will be initiated to examine how the context of family interaction alters throughout the life-span and how families at different stages of the life-span adjust to and facilitate the development of their members. This research will both clarify the causal links among behaviors, attitudes, and competencies within the family as a child moves from early to late adolescence and aid in the development of new analytic strategies for examining longitudional observational data.
