关键词: 婚姻满意度, 相似性, APIM, 以夫妻为中心, 家庭生命周期

Abstract: Similarity constitutes an essential component of marital satisfaction,yet current research has yielded inconsistent conclusions regarding the relationship between the two factors. First,previous studies have not adequately examined the length of marriage,resulting in a lack of dynamic adaptability in the research conclusions. Second,owing to the limitations of the research content,the results of previous studies are not comparable. Third,the similarity indices utilized in previous studies vary and therefore lack comparability. Finally,the majority of prior studies have adopted the variable-centered approach (VCA),which can be employed only to investigate the similarity of a couple on the basis of a single characteristic. Consequently,the aim of this study is to adopt the couple-centered approach (CCA) for couples at different marital stages and to integrate the actor effect,partner effect,similarity effect,and difference effect by combining profile-level similarity (PCS) and attribute-level difference (ADS) indicators to examine the relationship between the similarity of individual,interaction,and family of origin variables and marital satisfaction. We utilize dyadic data to explore the following questions: (1) Are real couples more similar than random couples? (2) In which dimension are real couples at different stages of marriage more similar? (3) How do these similarities affect marital satisfaction?
A total of 638 Chinese couples were recruited and divided into four marriage stages according to the family life cycle. The average ages of the husbands and wives were 48.83 years (SD = 13.42) and 47.40 years (SD = 13.36),respectively. After arriving at the laboratory,couples completed the questionnaire independently under the guidance of the interviewer. We first recorded the scores of the couples on all the items of each scale,then used RStudio to calculate the correlation coefficient between each column of these data (the correlation coefficient represents the PCS) and to calculate the absolute value of the score difference between the couple on each variable (the absolute difference score represents the ADS). Finally,we used an improved Actor-Partner Interdependence Model (APIM) incorporating ADS and PCS in Mplus to test the model.
The results indicate that first,real couples are more similar than random couples are,with the exception of family of origin variables. Second,compared with individual traits and interaction variables,couples are more similar in their family of origin. Third,regarding the similarity effect,husbands' marital satisfaction is most strongly influenced by similarities in their family of origin during the first marital stage,whereas in all other marital stages,similarities in personal traits have the greatest effect. For wives,personal trait similarity has the most significant influence during the first and fourth marital stages,whereas similarities in family of origin dominate the second and third marital stages. Finally,among the four effects,the actor effect has the strongest predictive power.
This study reveals the relationship between the similarity of diverse variables and marital satisfaction. It emphasizes that people should not only pay attention to the similarities between couples but also attach greater importance to the cultivation and display of their own personal traits for the sake of enhancing marital satisfaction. Overall,these findings hold crucial reference value and practical significance for our future research on the relevant mechanisms of marital satisfaction and marriage counseling.

Key words: marital satisfaction, similarity, APIM, couple-centered approach, family life cycle