医学量表编制中的统计学方法进展

量表广泛运用在心理学和教育学中,而在医学研究中,许多疾病状态和不健康行为等,只能通过量表测量和间接地测评。但量表编制在医学领域的运用中还显得有些滞后和不足。本文将已在心理学和教育学量表经过验证的量表编制中经常使用的测量学理论,即经典测量理论(classical test theory, CTT)、概化理论(generalizability theory,GT)和项目反应理论(item response theory,IRT)三大类,加以介绍。CTT理论广泛运用在医学量表编制中,具有简便易行、计算准确等优点,但其只适用于简单线性模型,限制了其在医学中的应用,而在教育学和心理学试卷及量表编制中,已开发引入了更为前沿的测量理论。GT理论作为CTT理论的补充,在同时考虑多个误差来源的基础上进行信度估计,信度估计比CTT更为细致和准确。IRT理论引入非线性模型及相关参数,可对量表条目进行更为细致和深入的研究分析,提升量表条目质量。但因以上两套理论实施复杂和学科之间壁垒存在等原因,一直未出现在医学量表编制中,如果使其在将来的医学量表的编制中得到充分广泛的运用和验证,将有重要的理论和现实意义。

Advancement on statistical methods in the development of medical scale

Scales are widely used in psychology and pedagogy. In medical researches, many diseases and unhealthy behavior can only be measured and evaluated indirectly by scales; however, the application of scales to the medical field is still a little backward and inadequate. This paper introduces the measurement theories which include test theory (CTT), generalizability theory (GT) and item response theory (IRT) and have been verified and commonly used in the psychology and pedagogy scales. CTT is widely used in the development of medical scales and has the advantages of simple operation and accurate calculation, but it only applies to simple linear models and limits its application in medicine. In pedagogy and psychology, a more cutting-edge measurement theory has been developed. As a complement to CTT, GT is more detailed and accurate than CTT in the reliability estimation based on multiple error sources. IRT introduces nonlinear models and related parameters, which make more detailed and in-depth analysis on the scale items and improve the quality of the scale. However, the above-mentioned two theories have not appeared in the development of medical scales due to their complexity and the existence of barriers between disciplines. It will have important theoretical and practical significance if the above-mentioned two theories are fully and widely applied and verified in the future medical scales.