본 연구는 유치원교사를 위한 정서노동 척도를 개발하고 타당도와 신뢰도를 검증하는 것을 목적으로 하였다. 먼저 문헌연구 및 현직교사 면담을 통해 수집한 자료를 토대로 1차 시안을 작성하였으며, 전문가 협의회를 통해 2차 시안을 만들고, 또 다른 전문가 협의회와 예비연구를 통해 3차 시안을 마련하였다. 최종 문항 선정을 위한 본연구 단계에서는 수도권 지역의 유치원교사가 응답한 414개의 자료를 분석대상으로 하여, 문항분석, 구인타당도, 내용타당도, 준거타당도, 신뢰도 검증의 과정을 거쳤다. 그 결과, 표면행위와 내면행위 각기 4요인[학부모, 원장(감), 동료교사, 유아] 12문항, 총 24문항의 최종 척도를 확정하였다. 확인적 요인분석을 통해 최적의 모형을 선정함으로써 구인타당도를 입증하였으며, 정서노동 관련 연구 경험이 있는 전문가들의 검토를 통해 내용타당도를 검증하였고, 기존 도구(교사용 정서노동 척도)와의 상관관계 분석을 통해 준거타당도를 검증(표면행위 r=.375, p<.001, 내면행위 r=.428, p<.001)하였다. 신뢰도의 경우, 문항내적일관성신뢰도(Cronbach’s α)는 표면행위 .867, 내면행위 .857, 검사-재검사 신뢰도(r)는 표면행위 .703(p<.001), 내면행위.522(p<.001)로 나타나 본 척도가 일관성 및 안정성을 가지는 척도임을 확인하였다.
The purpose of this study is to develop a emotional labor scale for kindergarten teachers in Korea and to test validity and reliability of the scale. The first item draft was developed based on extensive literature review and interviews with teachers. The second item draft was prepared through the experts’ review, and the third item draft was developed based on the review of another group of experts and a pilot study. To determine the final items, the data collected from 414 kindergarten teachers in the Metropolitan area were analyzed through item analysis, construct validity, content validity, criterion-related validity, and reliability test. As a result, the final scale of 24 items was confirmed with 4 factors and 12 items on both surface acting and deep acting. The construct validity was confirmed by selecting a best model via Confirmatory Factor Analysis; the content validity was verified through several reviews by experts with experience in research related to emotional labor; the criterion-related validity was confirmed through the correlation analysis with the existing scale (ELTS), resulting in r=.375 (p<.001) for surface acting and r=.428 (p<.001) for deep acting. In terms of reliability, the results of item internal consistency reliability (Cronbach’s α) were .867 for surface acting and .857 for deep acting, and the results of test-retest reliability were r=.703 (p<.001) for surface acting and r=.522 (p<.001) for deep acting; both of which were revealed to be acceptably high, confirming that the final scale had a consistency and stability.
Ⅰ. 서론 Ⅱ. 연구방법 Ⅲ. 연구결과 Ⅳ. 논의 및 결론
