The last decade has seen great growth in the collection of demographic data in which two or more members of a family or household are queried regarding matters of a subjective nature - from attitudes and expectations to evaluations of equity. We shall examine a variety of intra-household response patterns via a measurement model that does not detach the analytic phenomenon from its roots in the social process. The model we propose allows us to test for different evaluations and perceptions of subjective phenomena which are rooted in differential experience. Although the concept of structured differences is central to sociological work on the family, it is often ignored in the analysis of survey data. The ostensible fixity of the survey instrument leads to two alternative interpretations of disagreement in response: l) measurement error, and 2) true dyadic disagreement, particular to the dyad. Lost in between is our sociological understanding of perceptions and valuative standards as role- specific. These fundamental measurement/conceptual issues will be studied in four distinct substantive domains. First using data from the National Survey of Families and Households (NSFH), husbands' and wives' reports of marital happiness and fairness in the division of family labor will be compared. Second, we shall contrast mothers' and children's reports of a child's closeness to fathers using data from the National Survey of Children (NSC). Third, we shall assess the degree of agreement between husbands' and wives' fertility intentions using data from the NSFH and data from' a set of contrasting, developing country contexts. Finally, we shall examine change in fertility related attitudes across panels of the National Longitudinal Survey of the High School Class of 1972 (NLS-72) and the NSFH.
