友情亲密关系经历量表(ECR)的制定——保留内容效度的初高中学生版

Establishment of Experiences in Close Relationships Inventory (ECR)—Retaining the Junior and Senior Middle School Student Version of the Content Validity

摘要: 目的：针对青少年群体中检出率较高的对人不安、对人回避的心理问题，本研究旨在研发测量这些心理问题的量表。方法：对13~18岁的初高中群体，使用本研究制定的友情亲密关系经历量表(ECR)及外部研究的总体自我价值感量表(TOTAL-S)、中学生孤独感评定量表(ESLS)，通过线上与线下扫码的方式展开调查。采用配额抽样、滚雪球抽样方法共收集样本234例。对3个量表分别使用探索性因子分析、信度分析确定其因子构造与信度。主要研究的ECR使用验证性因子分析检定模型拟合度，并检测上述3量表的相关性以评定效度。结果：探索性因子分析中，ECR确认为2因子(对人不安：α = 0.85，对人回避：α = 0.86)共17题，其余量表分别为单因子构造(TOTAL-S: α = 0.90, ESLS: α = 0.87)。验证性因子分析中，ECR表现良好(χ² = 153.38, df = 98, p < 0.00, GFI = 0.93, AGFI = 0.89, RMSEA = 0.04, AIC = 263.38)。效度方面，对人不安与对人回避对TOTAL-S呈现负的、ESLS呈现正的相关(r = −0.44, r = −0.40; r = 0.63, r = 0.53)。结论：本研究所开发的ECR-17保留了较广的测试范围(内容效度)。同时，在信度、拟合度、建构效度上均表现出色。

Abstract: Objective: This study aims to develop scales for measuring psychological problems, specifically anxiety towards others and avoidance of others, among adolescents with a high detection rate of these issues. Method: The Experiences in Close Relationships Inventory (ECR) developed for this study, alongside the Total Self-Worth (TOTAL-S) and the Emotional and Social Loneliness Scale (ESLS) from external studies, were utilized to conduct a survey among middle and high school students aged 13~18 through online and offline QR code scanning methods. A total of 234 samples were collected using quota sampling and snowball sampling techniques. Exploratory factor analysis and reliability analysis were employed to determine the factor structure and reliability of the three scales. Confirmatory factor analysis was conducted on the ECR to test the model’s fit, and its validity was assessed by examining correlations between the three scales. Results: The exploratory factor analysis confirmed that the ECR consists of two factors: anxiety towards others: α = 0.85 and avoidance of others: α = 0.86, with a total of 17 questions. The other scales are of single-factor construction (TOTAL-S: α = 0.90, ESLS: α = 0.87). The confirmatory factor analysis indicated a good fit for the ECR model (χ² = 153.38, df = 98, p < 0.00, GFI = 0.93, AGFI = 0.89, RMSEA = 0.04, AIC = 263.38). In terms of validity, anxiety towards others and avoidance of others showed a negative correlation with TOTAL-S and a positive correlation with ESLS (r = −0.44, r = −0.40; r = 0.63, r = 0.53). Conclusion: The ECR-17 developed in this study demonstrates a broad testing scope (content validity) and performs well in terms of reliability, model fit, and construct validity.