Although several methods have been developed to allow for the analysis of data in the presence of missing values, no clear guide exists to help family researchers in choosing among the many options and procedures available. We delineate these options and examine the sensitivity of the findings in a regression model estimated in three random samples from the National Survey of Families and Households ( n = 250 -2,000). These results, combined with findings from simulation studies, are used to guide answers to a set of 10 common questions asked by researchers when selecting a missing data approach. Modern missing data techniques were found to perform better than traditional ones, but differences between the types of modern approaches had minor effects on the estimates and substantive conclusions. Our findings suggest that the researcher has considerable flexibility in selecting among modern options for handling missing data.Within the last decade, the practice of analyzing data in the presence of missing values has

No abstract

We use data from two national surveys of married individuals—one from 1980 and the other from 2000—to understand how three dimensions of marital quality changed during this period. Marital happiness and divorce proneness changed little between 1980 and 2000, but marital interaction declined significantly. A decomposition analysis suggested that offsetting trends affected marital quality. Increases in marital heterogamy, premarital cohabitation, wives' extended hours of employment, and wives' job demands were associated with declines in multiple dimensions of marital quality. In contrast, increases in economic resources, decision‐making equality, nontraditional attitudes toward gender, and support for the norm of lifelong marriage were associated with improvements in multiple dimensions of marital quality. Increases in husbands' share of housework appeared to depress marital quality among husbands but to improve marital quality among wives.

Research linking basal cortisol levels with internalizing and externalizing behavior problems in youths has yielded inconsistent results. We hypothesize that the high moment to moment variation in adrenocortical activity requires an analytical strategy that separates variance in cortisol levels attributable to "stable traitlike" versus "state or situationally specific" sources. Early morning saliva samples were obtained from 724 youths~M age ϭ 13.5 years; range ϭ 6-16 years in Year 1! on 2 successive days 1 year apart. Latent state-trait modeling revealed that 70% of the variance in cortisol levels could be attributed to statelike sources, and 28% to traitlike sources. For boys only, higher levels of externalizing problem behaviors were consistently associated with lower cortisol attributable to traitlike sources across 3 years of behavioral assessment. The inverse association between individual differences in cortisol and externalizing problem behavior has previously only been reported in studies of at-risk or clinical groups. The present findings suggest the relationship is a stable phenomenon that spans both normative and atypical child development. Studies are needed to reveal the biosocial mechanisms involved in the establishment and maintenance of this phenomenon, and to decipher whether individual differences in this hormone-behavior link confers risk or resilience.

Study of the effect of transitions on individual and family outcomes is central to understanding families over the life course. There is little consensus, however, on the appropriate statistical methods needed to study transitions in panel data. This article compares lagged dependent variable (LDV) and change score (CS) methods for analyzing the effect of events in two-wave panel data. The methods are described, and their performances are compared both with a simulation and a substantive example using the National Survey of Families and Households two-wave panel. The results suggest that CS methods have advantages over LDV techniques in estimating the effect of events on outcomes in two-wave panel data.