「工作設計量表」信效度衡鑑與工作結構潛在類別分析

Validation of the Work Design Questionnaire and Latent Class Analysis of Work Structure

工作設計是組織管理領域所重視的課題。隨著科技的進步與知識經濟的起飛，近年來有關工作設計在理論與實務上均有新的發展。本研究除了將Morgeson 和Humphrey（2006）的「工作設計量表」進行中文化與信效度分析之外，更進一步的運用潛在類別分析（LCA），針對不同工作內容，探討工作特徵的潛在結構。研究樣本為405 位不同職務內容的工作者，經過往覆翻譯法所完成的「工作設計量表」繁體中文版（TC-WDQ）信度良好，驗證性因素分析獲得穩定的22 個向度之因素結構，各向度與工作滿意度、組織創新氣氛、組織契合與工作契合等效標具有不同型態的顯著關聯。潛在類別分析結果發現，TC-WDQ 的22個工作特徵向度，可抽離出四種主要不同的工作特徵組型，在背景變數與效標變數上均具有顯著差異，以職務別所進行的共變模型LCA，可看出相對應的職務者有相對應的潛在類別特性，此顯示LCA 能夠有效的萃取出工作特徵背後的次母體。本研究示範了LCA 可以配合量表發展，檢測量表潛在結構分析，在測量方法學上具有高度實用價值。

The present study examined the reliability and validity of Traditional Chinese version of comprehensive WDQ; and applied latent class modeling to extract the latent classes of work characteristics. A total of 405 employees from different job contents were collected in this study. The measures of work characteristics were taken by Work Design Questionnaire proposed by Morgeson & Humphrey (2006). Results indicated that there are four major latent classes among the 22 work characteristics dimensions，entitled as regular working condition; technological working condition; low-level administration;and high-level knowledge based working condition. Employee belonged to different classes have foundsignificantly differences on background variables. In addition，covariance of latent class modeling by taking the job contents into account also shown the positive results. Methodologically， this study demonstrated that the latent class modeling is an appropriate analytic tool for organizational study. Further discussions on the method as well as practical usage are highly encouraged.