(2(67)= 241.813). Other goodness-of-fit indices proved that thestructural model reasonably fit the data (GFI = .914; NFI = .940;IFI = .956; CFI = .955; RMSEA = .082). Hypothesis 1 suggested that,as employees’ emotional labor increases, emotional dissonanceincreases. Hypothesis 1 was supported (ˇ-value = .657). In addition,hypothesis 2, which hypothesized a positive relationship betweenemployees’ emotional labor and job stress, was supported (ˇ-value = .206). Hypothesis 3 predicted that as employees’ emotionallabor increases, turnover intent increases. Therefore, hypothesis 3(ˇ-value = .426) was supported. However, hypothesis 4, which pre-dicted a positive relationship between emotional dissonance andturnover intent, was not supported (ˇ-value = .100). The findingdoes not support previous studies. Even if emotional dissonance isdirectly related to employees’ psychological stability, the sense ofshame and emotional exhaustion felt by employees do not directlyaffect their turnover intent. Hypothesis 5, which examined howemployees’ job stress affects turnover intent was supported (ˇ-value = .173)