Methodological issues in the study of divergent views of the family

A wealth of methodologies, both qualitative and quantitative, are available for conducting research on divergent views in the family. While the pathways for investigating divergent views of the family are clearer for those who chose more quantitative methods (as current literature reflects this tradition), qualitative methods may serve to clarify the process through which divergent views occur and are maintained within the family, both on an individual and on a dyadic or triadic level. In terms of the analysis of data from quantitative designs, divergent views of the family can quite easily be explored by viewing the family as a system and the individual family members as components of that system. This approach minimizes problems arising from having sampled families and included more than one family member from each. In addition, the organization of the data as family (rather than individual) records enables the researcher to directly examine divergences between family members by comparing ratings by different family members on the same construct. The issues involved in such analyses were discussed, first, in terms of the reliability of difference scores and then, in more detail, in terms of alternative regression-based models that estimate the effects of differences on a dependent variable. Although differences between variables are reputed to be (and often are) unreliable, the conditions under which these scores are likely to be reliable are the conditions found when examining divergent family views. Also, the reliability of difference variables can be treated directly. The discussion of regression models indicated the statistical equivalence (or in some instances nonequivalence) of models containing difference variables to alternative models not containing difference variables. Although algebraically equivalent, these alternative linear models provide different views of the same information. When moving from linear models to nonlinear models, however, researchers are cautioned to choose each specific model carefully since the nonlinear terms containing difference variables are actually more complex than they at first appear. The current volume speaks to the importance of variation in family members' views for the adjustment and well-being of adolescents. This chapter has taken on the question of how best to look at these divergences, and how different methods and statistical techniques may yield similar or different information regarding divergent views. The diversity in research methods and analytical strategies creates a challenging task for the investigator, and the continued exploration of questions regarding the implications of divergent views in the family should enhance our knowledge of how and why views diverge, as well as what divergences mean to family members.