基于项目反应理论修订中文版VoiSS量表

Modify the Chinese Version of the Voice Symptom Scale by Item Response Theory

目的应用项目反应理论对中文版VoiSS量表进行条目筛选,并评价修订后量表的信效度。方法在长沙地区采用分层整群抽样抽取412名中小学教师,应用中文版VoiSS量表评估其嗓音状况。利用MULTILOG 7. 03对结果进行项目功能分析,并对比筛选前后量表的信效度指标变化。结果量表条目区分度在1. 42～3. 86之间;难度范围为-1. 89～3. 01,且呈单调递增趋势;项目平均信息量区间为0. 556～2. 473,根据相应参数的筛选标准共保留21项条目。结论应用项目反应理论对中文版VoiSS量表进行筛选是科学、有效的;为了更准确地评估我国人群的嗓音状况,应对量表中的部分条目进行语言调适。

Objective To construct a shorter form of the Chinese version of the voice symptom scale(VoiSS)using Item Response Theory(IRT)and test the validity and reliability of the revised scale.Methods 412 middle school or primary teachers worked at Changsha were interviewed and assessed by VoiSS,through the stratified cluster sampling method.MULTILOG 7.03 was used for Item Function Analysis.Differences of the validity and reliability between the VoiSS and the revised scale were compared and assessed.Results Among the 30 items,the discrimination parameters of all items were between 1.42~3.86.The difficulty parameters of each item increased monotonically for all items with a range from-1.89~3.01 and the item average information parameters of all items were between 0.556~2.473.Finally,21 items were selected as good ones.Conclusion Using Item Response Theory(IRT)to select items on VoiSS is scientific and effective.In order to more accurately assess the voice of the population in our country,some of the items in the scale should be adapted to the language.