為提升工作狂研究的發展，本研究針對Spence與Robbins（1992）提出的三因子模式，檢驗其「工作狂量表」短版（WorkBAT-15）是否具有跨性別之測量恆等性（measurement equivalence/invariance, ME/I）。遵循Vandenberg與Lance（2000）所建議的檢驗程序，本研究以臺灣的519位全職員工為樣本，進行一系列多群組驗證性因素分析（multi-group confirmatory analysis, MGCFAs），依序檢驗不同性別之間是否具有因素構造恆等性、因素量尺恆等性、因素純量恆等性、測量誤差恆等性、因素變異數－共變數矩陣恆等性，以及因素平均數恆等性。依據上述指標分析之研究結果顯示，Spence與Robbins的「工作狂量表」短版在不同性別間具有高度恆等性，可進行性別間潛在變數平均數、構念變異與共變數，以及未標準化之路徑係數的比較。

To improve the research development of workaholism, the present study examined whether measures of the shorten version of Workaholism Battery (WorkBAT-15) in Spence and Robbins (1992) three factors model research are equivalent across gender. According the analysis process suggested by Vandenberg and Lance (2000), we conducted a series of multi-group confirmatory factor analyses (MGCFAs) to investigate the cross-gender measurement equivalence/invariance (ME/I). Data collected from 519 full-time employees is used in the present study. Configural invariance, metric invariance, scalar invariance, uniqueness invariance, factor variance-covariance invariance and factor mean invariance were established, suggesting that WorkBAT-15 is applicable to cross-gender comparisons in latent means, construct variance, construct covariance, and unstandardized path coefficients with the caution.