A data set of 1030 individuals (including 392 married couples) was employed to create a comprehensive picture of the interactive impact of many variables on marital satisfaction. Predictor variables were eventually combined into 20 composite variables and structural equation modeling resulted in 78.6% of the variance in marital satisfaction being explained for men; 79.8% for women. The primary dependent variable was Relational Satisfaction. Primary predictors (all composite variables) included emotional engagement, emotional-regulation skills, destructive interactions, shared activities, family and friend support, compatibility, strength of personal identity, accuracy of perception (of their partner), personality traits, temperaments (from the DISC measure), improvement over time, and positive illusions. To measure change over time, participants answered questions for both “now” and in the “first year of marriage”. Further, a criss-cross technique (rate self and partner across all variables) facilitated many comparative predictors. The structural models found the primary predictors of relational satisfaction (with only minor differences between mens’ and womens’ models) to be: emotional engagement (with β values of .56 for both), family and friend support, improvement over time, accuracy of perception, (absence of) destructive interactions, compatibility and positive traits. Equally important were predictors of emotional engagement—the greatest predictor of relational satisfaction: emotional-regulation skills (men), emotional-regulation skills (women), shared activities, accuracy of perception, family and friend support, and looking for the good explained 75% of the variance in the emotional engagement.
Read full abstract