【作者】 刘凯； 张龙飞； 涂冬波； 蔡艳；

【Author】 Kai Liu;Longfei Zhang;Dongbo Tu;Yan Cai;School of Psychology,Jiangxi Normal University;

【机构】 江西师范大学心理学院；

【摘要】 [背景]进食障碍(EDs)是一组以各种异常饮食行为和态度为特征的精神障碍,是由环境事件与个体生物学和发育特征相互作用而引起的。在EDs的筛查和诊断方面,众多学者们在经典测量理论的框架下开发了大量的进食障碍量表并在临床中得到了广泛的应用。然而,由于采用了经典测量理论,这些量表会存在一些不足,例如使用均值的标准误差使得测量精度不理想等,因此有必要寻找一种新的测量模式来克服经典测量理论的缺点。[目的]:本研究的核心目标是开发一个用于筛查进食障碍的计算机化自适应测验(CAT-ED),以此来提高测量精度,减轻受测者需要作答大量题目的负担。[方法]:一共有1025名大学生(平均年龄为19.7岁,标准差为2.2岁)参与了这项研究并且回答了131道题目,这些题目选自4个被广泛使用的进食障碍筛查量表。研究主要包括两个部分:CAT-ED的题库建设和心理测量学特性分析。第一部分中通过分析题库的单维性、模型拟合度、局部独立性、项目拟合、区分度以及是否存在项目功能差异以确保CAT-ED的科学性和高质量性。第二部分中分别进行了基于模拟被试和真实被试的模拟研究以检验CAT-ED的有效性和合理性,并且分析了其信度、效标关联效度以及敏感性和特异性。[结果]:最终的题库包含77道题目,这些题目满足了计算机自适应测验在单维性、局部独立性、项目拟合性、高区分度和无项目功能差异(DIF)等方面的基本心理测量学要求。在标准误为0.3的情况下,平均每个被试作答题库中的11道题目时,CAT-ED便能够对其进食障碍的严重程度进行较为准确的估计,并且与作答题库所有题目的测验总分保持r=0.942的相关性。此外,CAT-ED效标关联效度也是可接受的。同时,在不同终止规则下所计算的灵敏度和特异度也很好地体现了CAT-ED对于不同进食障碍严重程度的大学生的有效区分。[结论]:CAT-ED可作为一种在心理门诊中用于筛查进食障碍高危人群的有效检测工具,同时也为使用心理量表诊断进食障碍(EDs)提供了一个全新的视角。更多还原

【Abstract】 Background: Eating Disorders(EDs) are a group of mental disorders characterized by various abnormal eating behaviors and attitudes, which are caused by the interaction of environmental events with an individual’s biological and developmental characteristics. In the screening or diagnosis of EDs, scholars have developed plentiful EDs scales which are widely used in clinic under the framework of Classical Test Theory(CTT). However, these scales will have some shortcomings due to the use of the Classical Test Theory, for instance, the usage of the mean’s standard error makes the measurement accuracy is not ideal. Therefore, it is necessary to find a new measuring mode. Objective: The core goal of this research is to develop a computerized adaptive testing for Eating Disorders(CAT-ED) to increase measurement precision and decrease the respondents’ burden of answering a battery of items. Methods: A total of 1,025 college students with a mean age of 19.7 years(SD = 2.2) joined in this study and completed 131 items which were selected from 4 widely-used psychological scales about Eating Disorders. This research mainly consisted of two parts: constructing the item bank of CAT-ED and analyzing its psychometric properties. In the first part, unidimensionality, model fit, local independence, item fit, discrimination and differential item functioning(DIF) were all analyzed to make sure the CAT-ED is scientific and of high quality. In the second part, two simulation studies based on simulated and real subjects were applied to test the effectivity and rationality as well as the reliability, Criterion-Related Validity, and the sensitivity and specificity of CAT-ED were also checked. Results: The final item bank contained 77 items which meets the fundamental psychometric requirements of unidimensionality, local independence, item fit, high discrimination and no differential item functioning(DIF) in computerized adaptive testing. An average of 11 items per subject was only required to achieve a 0.3 standard error in the EDs severity estimate and maintain a correlation of 0.942 with the total 77-item test score. Besides, the CAT-ED also had acceptable criterion-related validity. At the same time, the calculated sensitivity and specificity of CAT-ED in different stopping rules manifested that it ideally and effectively discriminated college students with and without EDs. Conclusions: The CAT-ED could be used as an effective method for the detection of patients at risk for Eating Disorders in psychological clinic and it may offer a bran-new perspective for diagnosis of Eating Disorders(EDs) with psychological scales.更多还原