夫妻互动行为是影响婚姻质量的最重要的因素，从20世纪70年代起，越来越多的西方学者开始使用观察法对夫妻的互动行为进行编码分析，并发展出许多成熟的、且在婚姻观察研究领域影响力较大的编码系统。然而，一方面现有成熟的编码系统大多是基于西方文化背景发展而来，还鲜有基于中国文化背景研发的夫妻互动行为编码系统，而跨文化研究和国内外关于婚姻的不一致研究结果提示，中西方文化背景下夫妻互动行为可能存在较大差异，因此，从理论角度来看，直接选用基于西方文化背景发展而来的编码系统进行中国夫妻互动行为研究显然是不合适的；另一方面，对以西方编码系统为研究工具的国内实证研究进行进一步分析发现，中文翻译版往往存在编码者一致性较低或因素载荷较低等缺陷。即使在研究者已经对编码系统进行本土化修订的前提下，仍然发现某些具有中国夫妻特色的典型互动行为无法被现有编码体系所涵盖。 因此，本研究采用文献分析法对已有的西方夫妻互动研究中所涉及的编码系统及调查问卷共同关注的指标进行梳理，并应用访谈法对130对不同家庭发展阶段的中国夫妻的典型互动行为进行归纳（访谈数据来源于方晓义教授主持的国家自然科学基金“夫妻互动对婚姻关系的影响：行为与认知神经机制”项目），最后通过整合文献分析及访谈的结果发展出包含4个维度、14个二级指标及32个一级指标的中国夫妻互动行为编码系统。该编码系统共包含夫妻互动、原生家庭／拓展家庭、第三方/无关互动、对互动的整体评价4个维度，其中，夫妻互动维度包括心理攻击、身体攻击、敌意、回避、防御、控制、建设性沟通、情感表达8个二级指标，下属涵盖29个一级指标；原生家庭／拓展家庭维度包括孩子卷入及原生家庭卷入2个二级指标，下属涵盖4个一级指标；第三方/无关互动维度包括外部评价1个二级指标，下属涵盖2个一级指标；对互动的整体评价维度包括夫妻任务卷入程度、任务解决程度和沟通氛围3个二级指标。中国夫妻互动行为编码系统主要采用微观编码方式及动作采样程序，而在对互动的整体评价维度上采用宏观编码方式及时间采样程序。经18位婚姻研究与实践领域专家评估，中国夫妻互动行为编码系统中14个二级指标及32个一级指标的内容效度良好。针对该系统的编码者一致性的探索性研究结果显示，中国夫妻互动行为编码系统的信度在可接受范围内，但受样本量的影响，该信度检验结果可能仍需要后续研究进一步验证。

The interaction between husband and wife is the most important factor that is affecting the quality of marriage. Since the 1970s, more and more Western scholars have used the observation method to analyze marital interaction, and have developed many classic marital interaction coding systems with great influence in the research field. However, on the one hand, the existing coding systems are mostly based on Western culture, and there are few marital interaction coding systems that are based on Chinese culture. Cross-cultural studies and the inconsistent results of marriage researches at home and abroad suggest that there may be large differences in the interaction between couples in Chinese and Western cultures. Therefore, theoretically, it is obviously inappropriate to directly use the coding systems that are developed from Western cultures to study the interaction of Chinese couples. On the other hand, further analysis of domestic empirical researches using Western coding systems finds that the Chinese versions often have shortcomings such as low inter-coder correlations or low load factors. Even if the researchers have already localized and revised the coding system, it is still found that some typical interactions between Chinese couples cannot be covered by the Chinese version. Therefore, this study first used literature analysis to sort out the common indicators of the existing Western coding systems and questionnaires and then used the interview method to collect the typical interaction behaviors of 130 Chinese spouses of different marriage ages. (The interview data comes from the program funded by the Chinese National Natural Science Foundation, chaired by Professor Fang, whose name is “the Influence of Marital Interaction on Marriage: Behavioral and Cognitive Neuroscience Mechanisms”.) Finally, through the integration of literature analysis and interview results, a new coding system that designed for Chinese couples was developed. This new coding system has four dimensions, 14 level-two indicators, and 32 level-one indicators. The new coding system includes four dimensions, which are Couple Interaction, the Family of Origin/Extended Family, the Third-party/Independent Interaction, and Overall Evaluation of Interaction. Among them, the Couple Interaction dimension includes 29 level-one indicators and 8 level-two indicators, which are Psychological Attack, Physical Attack, Hostility, Avoidance, Defense, Control, Constructive Communication, and Emotional Expression. The Family of Origin/Extended Family dimension includes 4 level-one indicators and 2 level-two indicators, which are Children Involvement and the Family of Origin involvement. The Third-party/Independent Interaction dimension includes 2 level-one indicators and 1 level-two indicator, which is external evaluation. The Overall Evaluation of Interaction dimension includes 3 level-two indicators, which are the Engagement of the Couple, the Resolution of the Task, and the Communication Atmosphere. The new coding system mainly adopts the micro-coding method and the event sampling program and adopts the macro-coding method and the state sampling program in terms of the Overall Evaluation of Interaction dimension. According to 18 experts in the field of marriage research and practice, the content validity of 14 level-two indicators and 32 level-one indicators in the new coding system is good. The pilot study of the inter-coder reliability of the new coding system shows that the reliability of the new coding system is within the acceptable range. However, the result of the reliability test may still need to be further confirmed on account of the small sample size.