A dyadic process perspective for social interactions: social support and companionship in couples

Abstract

Background: Psychosocial processes such as companionship and support have long been of interest to health scientists and relationship researchers alike. However, most studies so far have focused either on individuals or couples with cross-sectional designs. A new wave of dyadic longitudinal studies promises a better understanding of psychosocial processes and their links with well-being and health behaviour, examining both members of couples, close to real time, and with high ecological validity. We present data from two dyadic longitudinal studies of the links between companionship and social support with well-being and health behaviour. Methods: Both partners of committed couples (Study 1: N = 90, Study 2: N = 99) filled out daily online diaries for over a month. We compare different statistical approaches, with a dyadic score model (Iida, Seidman, & Shrout, in press) allowing to model the dyadic level. Findings: Companionship and support varied between couples, within partners, and from day to day within person in both studies. Companionship and support showed high correlations between partners (r > .49, p < .05 within and between persons), emphasizing the need for understanding the couple level. Both companionship and support emerged as intertwined, yet independent constructs with unique links to affect, relationship satisfaction, and health behaviour. In Study 2, smokers with higher companionship smoked fewer cigarettes, over and above support. Discussion: Our findings demonstrate the need for a dyadic process perspective that integrates appropriate theory building, study designs, and statistical methods to better understand psychosocial predictors of well-being and health behaviour.