Questioned Straus and Gelles' (1986) conclusion that child and wife abuse were substantially lower in 1985 than in 1975. Refined cohort analysis to apply to data, demonstrating that changes in age structures of adult and child population explained most of reported reduction. Much remaining change may have resulted when respondents' reactions to…

Bias due to attrition of respondents poses a threat to the internal and external validity of research findings. Discusses methods of detecting attrition bias in longitudinal family research, and presents Heckman's procedure to correct attrition bias. Data from the University of California Longitudinal Study of Generations are used to illustrate…

Dealing with data from multiple family members presents problems for researchers, as many of the techniques available for dealing with such data are problematic, and there is no way that is unique to each family. Proposes and gives an example of the use of Q factor analysis as a means of combining data from multiple family members in a unique way…