Predicting marital satisfaction from behavior: Do all roads really lead toRome?

The concurrent and longitudinal associations between spousal communication and marital satisfaction were investigated using two systems of observation (rating and coding) and two statistical methods. A coding system and a rat ing system were used to assess the amount of positive, negative, and avoida nt behavior exhibited by 30 couples engaged in a 15-minute discussion about a marital problem. Change scores as well as regression methods were used t o calculate correlations between communication behaviors and one-year chang e in Marital Adjustment Test (MAT) scores. Positive behavior by both spouse s predicted an increase in wives' satisfaction, and negative behavior by bo th spouses predicted a decrease in wives' satisfaction. None of the variabl es consistently predicted changes in husbarids' satisfaction. The two syste ms of observation and the two methods of analysis produced similar though n ot identical results. The article discusses the constructs of negative and avoidant behavior, reviews limitations of standard marital research methods , and offers theoretical and methodological recommendations for future rese arch.