The Rasch measurement model in rheumatology: What is it and why use it? When should it be applied, and what should one look for in a Rasch paper?
The Rasch measurement model in rheumatology: What is it and why use it? When should it be applied, and what should one look for in a Rasch paper?
复制标题
DOI：
10.1002/art.23108
复制
发表时间：
2007-12-15
期刊：
影响因子：
--
通讯作者：
Conaghan, Philip G.
中科院分区：
文献类型：
--
作者：
Tennant, Alan;Conaghan, Philip G.
在评估结果工具时，风湿病研究人员熟悉测量科学的传统标准：有效性、可靠性和响应性(1)。在这些方法中，结构效度可能通过因素分析技术得到支持，这些技术证实存在一个或多个有效的一维标尺(测量单个结构的标尺)。可靠性通常用Cronbach‘s Alpha(2)来报告。还注意到响应性的概念，即量表通过效果大小或标准化反应平均值等衡量标准发现变化的能力(3)。现代心理测量学方法已被采用来补充这一传统方法(4)。这些新方法大多与Rasch测量模型(5)的应用有关，通常被称为Rasch分析。尽管在过去的40年里，RASCH在教育中得到了广泛的应用，但在过去的十年里，RASCH分析在健康科学中的各种应用已经发表。从早期对广泛使用的患者报告的量表(如健康评估问卷)的早期工作，到修订观察员评估的评分系统，如拉森放射学评分(7)，到最近引入题库的概念(8)。然而，对这一分析技术及其应用的理解仍然有一定的局限性。使用不同的软件包产生不同的RaschRelated统计数据也没有帮助到这一点。本文试图提供关于RASCH分析使用的最新情况；解释它是什么，为什么应该使用它，什么时候应用它；并就当前最先进的分析应该包括什么，以及因此应该在使用RASCH分析的论文中期望什么提供指导。
When evaluating outcome tools, rheumatology researchers are familiar with the traditional standards of measurement science: validity, reliability, and responsiveness (1). In these approaches construct validity may have been supported through factor analytic techniques that confirmed the presence of 1 or more valid unidimensional scales (a scale measuring a single construct). Reliability is usually reported as Cronbach’s alpha (2). Attention has also been given to the concept of responsiveness, the ability of the scale to detect change through measures such as the effect size or standardized response mean (3). Modern psychometric approaches have been adopted to supplement this traditional approach (4). These new approaches have been mostly associated with the application of the Rasch measurement model (5), usually referred to as Rasch analysis. Although Rasch has been widely used in education for the last 40 years, over the last decade a wide variety of applications of Rasch analysis have been published in the health sciences. These range from early work on widely used patient-reported scales such as the Health Assessment Questionnaire (6) through the revision of observer-evaluated scoring systems such as Larsen radiographic scoring (7) to the recent introduction of the concept of item banking (8). However, understanding of this analytical technique and its applications remains somewhat limited. This has not been helped by the use of different software packages producing different Raschrelated statistics. This article seeks to provide an update on the use of Rasch analysis; explain what it is, why it should be used, when to apply it; and provide guidance on what the current state-of-the-art analysis should comprise, and consequently what should be expected in a paper that uses Rasch analysis.