情緒勞務對於工作者而言，是一種在工作時面對不同狀態時，自身情緒管理的策略，工作者會在 面對不同的組織制度與要求，進行內在的情緒調節並在工作現場完成情緒的表達。本研究運用 Rosenthal(1995)統合分析法分析 2005 至 2022 年間臺灣情緒勞務期刊量化實證研究 51 篇，研究 結果指出情緒勞務與組織制度及工作表現間具有中度且正向相關，情緒勞務與正向主觀感受間具 有低度正向相關，情緒勞務與負向主觀感受間具有低度負向相關。依據研究結果提出實務建議: 針對個人增加自我覺察來感受情緒，並透過一致性的情緒管理策略來面對工作現場所面對的對象; 在組織方面透過制度與工作環境的調整、組織文化脈絡的傳承、員工支持方案提供等方式，協助 員工的凝聚力提升以及對於組織認同度的提升，使情緒勞務得以調整。

In this study, an objective meta-analysis was conducted to analyze studies published between 2005 and 2022 in Taiwanese journals on the effects of emotional labor on organizational systems, work performance, and the positive and negative subjective feelings of workers in Taiwan. Empirical research findings were collected and compiled to elucidate the scope of research on emotional labor in Taiwan. In addition to the effect size of each related variable, the correlations between emotional labor and various factors, including organizational systems, work performance, and positive and negative subjective feelings, were analyzed. On the basis of the results of a literature review, the variables and subconstructs considered in the meta-analysis were organizational systems (with organizational norm, organizational competition, and working atmosphere as subconstructs), work performance (with self-efficacy, voice behavior, and customer-oriented behavior as subconstructs), positive subjective feelings (with emotional intelligence, work satisfaction, and mental health as subconstructs), and negative subjective feelings (with work stress and burnout as subconstructs). Before the literature review was conducted, the current study defined emotional labor. A rigorous and comprehensive process for selecting relevant studies on emotional labor was used to ensure the comprehensiveness of the literature review and the validity of the data that were analyzed. A systematic search for journal articles was conducted in the Airiti Library. The ancestry approach was used to search the bibliographies and abstracts of these articles, and the descendancy approach was used to track follow-up research through the citation index, with the aim of maximizing the number of identified studies while avoiding missing key information. The selection criteria for journal articles were as follows. First, the articles were required to be quantitative studies on emotional labor included in the Airiti Library. Master’s and doctoral theses, qualitative studies, and case studies were excluded from the analysis. Second, the articles were required to provide sufficient statistical data, such as information on sample sizes, means, standard deviations (SDs), F values, and t values, to enable calculation of the effect size of each sample article. Third, the full texts of the articles were required to be publicly available. According to Rosenthal (1995), a meta-analysis is a technique that focuses on the overall results of each study rather than on its conclusions. That is, text descriptions should not contribute to calculations, and the probability value of effect size can be calculated only if selected studies provide sufficient numerical data, such as information on sample sizes, means, SDs, and correlation coefficients. Individual studies on emotional labor published up to December 2022 in the Airiti Library were searched using the Chinese keyword for “emotional labor.” A total of 88 studies were collected, and empirical quantitative studies providing comprehensive data were selected to facilitate the conversion of research data through mathematical equations. Studies were excluded if they did not provide information on sample sizes, means, t values, or F values, which were used for effect size calculation; they did not provide correlation coefficients, which were used for coding; they were qualitative or action studies that did not contain the data required to calculate effect sizes; or their dependent variables did not match the inclusion criteria of this study. Ultimately, 51 studies were selected, and the meta-analysis technique proposed by Rosenthal was implemented using Comprehensive Meta- Analysis 2.0, which is designed to analyze statistical research data for calculating weighted mean effect sizes, SDs, variances, confidence intervals (CIs), homogeneity degrees, and standardized residuals. According to the results, emotional labor moderately and positively correlated with organizational systems (mean effect size r = .403, 95% CI = .327– .473, Z = 9.596, p < .001), with a heterogeneous effect size distribution. An I2 value of 97.650 was obtained, indicating that 97.650% of the observed variances were actual variances between studies and that only approximately 3% were variances from random errors. These findings are consistent with those of Lin and Chang (2018) and Chang (2020), who have reported that emotional labor significantly and positively correlates with organizational systems, and Chen and Wang (2016), who reported that the work environment of an organization significantly and positively correlates with emotional labor. In terms of working atmosphere, Liu and Wang (2017) argued that employee competition positively correlates with emotional labor, which in turn significantly and positively correlates with organizational systems. According to Hur and Shin (2023), the daily perceptions of employees regarding employment insecurity can affect their emotional labor. During contingencies, organizations can provide their employees with various levels of support to mitigate their emotional labor and work stress, which frequently lead to emotional exhaustion (Kumar & Jin, 2022). In this study, consistent with the results of Cohen (1992), the correlation between variables was found to have a moderate effect size. That is, when either organizational systems or emotional labor affects workers, the other variable also simultaneously influences workers. Emotional labor moderately and positively correlated with work performance (r = .378, 95% CI = .339– .415, Z = 17.582, p < .001), with a heterogeneous effect size distribution. An I2 value of 91.282 was obtained, indicating that 91.282% of the observed variances were actual variances between studies and that only approximately 3% were variances from random errors. These findings are consistent with those of Chen and Guo (2020) and Chung et al. (2021), who have reported that emotional labor significantly and positively correlates with work performance and self-efficacy. According to Hou et al. (2015), Wang and Chang (2016), and Chen and Tsai (2020), emotional labor positively correlates with worker behavior. In an interview study, Kirk et al. (2023) discovered that reducing emotional labor mitigates the potential negative effect of work that workers perceive. Emotional labor also greatly affects the creativity of first-line workers (Ma et al., 2023). Hu et al. (2023) argued that organizational support positively mediates the positive effects of emotional labor on job competence and commitment. In this study, a moderate and positive correlation was observed between emotional labor and work performance. In other words, when either emotional labor or work performance affects workers, the other variable also simultaneously influences workers. Emotional labor weakly and positively correlated with positive subjective feelings (r = .170, 95% CI = .098– .239, Z = 4.630, p < .001), with a heterogeneous effect size distribution. An I2 value of 97.668 was obtained, indicating that 97.668% of the observed variances were actual variances between studies and that only approximately 3% were variances from random errors. According to the literature, positive subjective feelings include work satisfaction, emotional intelligence, and mental health. Xu and Fan (2023) argued that emotional labor positively correlates with work satisfaction, and this correlation is mediated by work relationships. Wu et al. (2024) indicated that the deep-level effects of emotional labor on employees increase service quality, whereas its surface-level effects reduce service quality. Delgado et al. (2020) reported a positive correlation between mental resilience and emotional labor. Various studies conducted in Taiwan have indicated that emotional labor positively correlates with work satisfaction (Chen et al., 2017; Chen & Wang, 2016; Chu, 2013; Hsu & Lee, 2012; Liu & Chen, 2016). For example, Hu (2013) and Hsieh (2015) have reported that emotional labor positively correlates with emotional intelligence. Furthermore, Lee et al. (2013), Lee (2013), and Chi (2015) have argued that emotional labor positively correlates with mental health. In this study, higher emotional labor was found to be associated with stronger positive subjective feelings. Accordingly, effectively implementing emotional labor strategies can promote work satisfaction, emotional intelligence, and mental health. Emotional labor weakly and negatively correlated with negative subjective feelings (r = -.175, 95% CI = -.263 to -.085, Z = -3.775, p < .001), with a heterogeneous effect size distribution. An I2 value of 98.388 was obtained, indicating that 98.388% of the observed variances were actual variances between studies and that only approximately 2% were variances from random errors. Multiple empirical studies on negative subjective feelings have indicated that work stress and burnout are associated with emotional labor. According to Chen (2012), Chen and Wu (2019), Liu and Chen (2016), Yang (2021), and Huang (2022), emotional labor significantly and positively correlates with work stress but negatively correlates with burnout. Peng et al. (2017) argued that each emotional labor strategy leads to unique outcomes associated with negative subjective feelings. Hu et al. (2023) contended that work stress and organizational support predict the positive and negative potential effects of work, and these two variables are associated with burnout. These findings are consistent with those of this study, indicating that emotional labor is associated with a reduction in negative subjective feelings. This notion suggests that workers can mitigate negative thoughts and maintain their work enthusiasm by managing their emotional labor through finding joy in their work, which can prevent emotional exhaustion resulting from the accumulation of psychological stress.

情緒勞務；組織制度；工作表現；正向主觀感受；負向主觀感受

emotional labor; organizational system; work performance; positive subjective feelings; negative subjective feelings