Lo Meng-Ting, Chen Ssu-Kuang, O'Connell Ann A
Meng-Ting Lo, Department of Educational Studies, College of Education and Human Ecology, The Ohio State University, 173 Arps Hall, 1945 North High Street, Columbus, OH 43210, USA,
J Appl Meas. 2018;19(4):413-427.
The present study used the Rasch rating scale model (RSM) to reassess the psychometric properties of the Chinese version of Rosenberg self-esteem scale (RSES) among 501 Grade 10 students in Taiwan. The reliability, dimensionality, and differential item functioning were examined. The dimensionality assumption was met after excluding item 8 ("I wish I could have more respect for myself."). The successive response categories for item 7 ("I feel that I am a person of worth, at least on an equal plane with others.") were not located in an expected order. After eliminating items 7 and 8 from analysis, the remaining 8-item RSES had acceptable fit statistics, good content coverage and high categorical omega, Rasch person and item reliability. The five response categories performed well; evidence for convergent validity was established through the high correlation between RSES and psychological being scores. Implications and recommendations for instrument users are discussed.
本研究采用拉施评分量表模型（RSM）对台湾501名十年级学生的中文版罗森伯格自尊量表（RSES）的心理测量特性进行重新评估。检验了信度、维度性和项目功能差异。排除第8项（“我希望我能更尊重自己。”）后，维度假设得到满足。第7项（“我觉得自己是一个有价值的人，至少与其他人处于平等地位。”）的连续反应类别未按预期顺序排列。从分析中剔除第7项和第8项后，剩余的8项RSES具有可接受的拟合统计量、良好的内容覆盖范围以及较高的分类ω系数、拉施人及项目信度。五个反应类别表现良好；通过RSES与心理幸福感得分之间的高度相关性建立了收敛效度的证据。讨论了对量表使用者的启示和建议。
