Longitudinal data analysis for husband and wife pairs has not been discussed in depth in the nursing literature. Traditionally, husband data and wife data were analyzed separately. Such analyses ignored the fact that husbands' and wives' scores tend to be related to each other. The purpose of this poster is to show readers to use hierarchical linear models to analyze husbands' and wives' longitudinal data in relation to one another, using couple as the unit of analysis. This was not found in nursing literature. Raudenbush, Brennan and Barnett (1995) used a multivariate hierarchical linear model to study psychological change within married couples. In their study, each person's psychological characteristics were viewed as changing over time as a function of both the individual's personal characteristics and the influence of his/her partner. The methodology they used is flexible in allowing randomly missing data, varying spacing of time points, unbalanced designs, and time-varying and time-invariant covariates. Pinelli et al (2003) studied the family adjustment of parents who had their newborn in the neonatal intensive care unit (NICU). One hundred and fifty-two families were studied. Each father's and mother's scores were gathered while in the NICU, and at 3, 6 and 12 months later. We measured adjustment, coping, resources, and stress for each father or mother at each of these 4 time points. We will describe sample statistics and then model first level (within individuals) data using time as the predictor. Linear and quadratic growth trajectories will be modeled for each subject at the first level, then we will test whether these growth trajectories vary significantly from person to person. If the answer is positive, we will then model the effects of other predictor variables on growth trajectories, using couple as a second level of analysis.
Session #1219 - Poster Session I
The 29th Annual MNRS Research Conference (April 1-4, 2005)