Abstract: Background　Traditional Chinese Medicine (TCM) health behaviors play an important role in maintaining and promoting public health. Scientifically assessing the level of TCM health behaviors has practical significance for health management and intervention. However, there is currently a lack of standardized localized assessment tools for TCM health behaviors in China. Objective　To develop a TCM Health Behavior Scale and examine its reliability and validity. Methods　Guided by the conceptual framework of TCM health behaviors, the initial scale was developed through literature review, semi-structured interviews, and group discussions. A test version was formed after expert consultation and a pilot survey. From October to December 2024, 590 community residents aged 18-80 from Sichuan, Hubei, and Hunan provinces were selected by convenience sampling. A total of 641 questionnaires were distributed, with 590 valid responses (92.04% effective rate). Item analysis and reliability/validity tests were conducted. Results　An initial scale of 10 dimensions and 41 items was constructed based on literature review and qualitative interviews. After expert revision, both rounds of consultation achieved 100% response rates, with suggestion rates of 95% and 74%, respectively. Expert authority coefficients were 0.889 and 0.903, and Kendall's W values were 0.168 and 0.073 (P<0.05). A test version with 9 dimensions and 40 items was finalized through expert feedback and panel discussions. After psychometric testing, the final scale comprised 8 dimensions and 34 items. Exploratory factor analysis extracted 8 common factors with a cumulative variance contribution of 70.944%, item loadings ranged from 0.548 to 0.899. Confirmatory factor analysis indicated good model fit: χ2 /df=1.713, RMSEA=0.049, GFI=0.860, CFI=0.945, NFI=0.879, TLI=0.937, IFI=0.946. All item factor loadings exceeded 0.5, CR values were > 0.7, AVE ranged 0.478-0.682, and inter-dimension correlation coefficients ranged 0.007-0.583, indicating acceptable convergent and discriminant validity. S-CVI was 0.935, and I-CVI ranged 0.890-1.000. The overall Cronbach's α was 0.918, and split-half reliability was 0.962, demonstrating good reliability. Conclusion　The TCM Health Behavior Scale demonstrates good reliability and validity, and can serve as a localized tool for health managers to assess the TCM health behavior levels of Chinese residents.

Key words: Chinese medicine health behaviors, Health promotion, Scale, Validity, Reliability

摘要： 背景　中医健康行为是维护和促进居民健康的重要方式，科学评估其中医健康行为水平对健康管理和干预具有重要实践意义，但目前国内缺乏标准化中医健康行为本土化测评工具。目的　编制中医健康行为量表并检验其信效度。方法　以中医健康行为概念框架为理论指导，通过文献研究、半结构式访谈与小组讨论构建初始量表；经专家函询和预调查后形成测试版量表。2024年10—12月，采用便利抽样法选取四川省、湖北省、湖南省590名18~80岁的社区居民作为调查对象，共发放问卷641份，回收有效问卷590份，有效回收率92.04%，对量表进行项目分析及信效度检验。结果　通过文献回顾、质性访谈初步构建了9个维度41个条目的初始量表。经专家函询完善量表，两轮专家积极性均为100%，意见提出率分别为95%和74%，专家权威系数为0.889、0.903，肯德尔协调系数分别为0.168、0.073（均P<0.05），结合专家意见和小组讨论后，形成9个维度40个条目的测试版量表。经信效度检验后，形成8个维度34个条目的正式版量表。探索性因子分析提取8个公因子，累计方差贡献率为70.944%，各条目载荷值为0.548~0.899。验证性因子分析显示模型拟合指标卡方自由度比值（χ2/df）=1.713，渐进残差均方和平方根（RMSEA）=0.049，拟合优度指数（GFI）=0.860，比较拟合指数（CFI）=0.945，规范拟合指数（NFI）=0.879，Tucker-Lewis指数（TLI）=0.937，增值适配指数（IFI）=0.946，提示模型拟合整体较为理想。模型各条目因子载荷均>0.5；建构信度（CR）值均>0.7，平均方差提取值（AVE）为0.478~0.682，各维度间相关系数绝对值为0.007~0.583，聚敛效度和区分效度指标均在可接受范围内。总量表水平内容效度指数（S-CVI）为0.935，各条目水平内容效度指数（I-CVI）为0.890~1.000。总量表Cronbach's α系数为0.918，折半信度为0.962，信度良好。结论　中医健康行为量表具有良好的信度和效度，可作为健康管理者评估我国居民中医健康行为水平的本土化工具。

关键词: 中医健康行为, 健康促进, 量表, 效度, 信度