旅館的經營強調專業、熱忱及細膩的服務，因此旅館的第一線服務人員常有為了工作，勉強自己去表現出符合組織及顧客期待的臉部表情或行為之情緒勞務的現象產生。過去兩年來，COVID-19大流行給全球旅館業帶來了嚴重打擊，然而旅館業者仍試圖應用不同的服務流程為客人提供優質服務。本文旨在探討後疫情時代，改變服務方式是否會增加或減少旅館第一線員工的情緒勞務和工作倦怠，同時探討其對組織承諾的影響。本研究以工作要求－資源模型（The Job Demands-Resources model, JD-R model）為研究架構，採用便利抽樣法，透過問卷調查臺南地區旅館業第一線員工，共取得有效問卷302份，問卷有效率98.6%。本研究使用SPSS22.0及AMOS24.0軟體進行資料分析，並採用結構方程模式（SEM）、迴歸分析等方法驗證假設及中介與調節效果。結果發現，旅館第一線員工之情緒勞務對工作倦怠有顯著正向影響、工作倦怠對組織承諾有顯著負向影響、情緒勞務對組織承諾有顯著負向影響、工作倦怠在情緒勞務與組織承諾關係間具有中介效果以及情緒智力在深層演出與組織承諾關係間具有調節效果。SEM結果亦顯示，工作要求－資源模型（JD-R model）與旅館第一線員工的資料配適度良好。最後再依據結論提出建議與實務意涵。

Hotel operation is a service industry with particular emphasis on professionalism, enthusiasm, and delicate service. Therefore, the hotel front-line employees often manage their emotions to act or express only those requested by their organizations in order to satisfy hotel guests. Over the past two years the COVID-19 pandemic has hit the global hotel industry hard, but hotel operators still tried to provide customers with alternative service by applying different service procedures. This paper seeks to explore whether changing service types increase or decrease hotel frontline employees' emotional labor and job burnout and to explore the impact on organizational commitment in the post-epidemic era. The Job Demands-Resources model (JD-R) was adapted as study framework; quantitative analysis and convenience sampling were used to survey the hotel frontline employees in Tainan City. 302 valid questionnaires were obtained (98.6% valid rate). The study used structural equation modeling (SEM) and regression analysis to test hypotheses and to examine the mediating and moderating effect. The result indicated that emotional labor has a positive effect on job burnout, job burnout has a negative impact on organizational commitment, emotional labor has a negative impact on organizational commitment, job burnout mediates the relationship between emotional labor and organizational commitment, and emotional intelligence moderates the relationship between deep acting and organizational commitment. The fit indices of the Job Demands-Resources model indicated a good fit to the frontline employees data.