온라인 기업 리뷰에서의 조직문화 유형과 이직의도 점수 간 공변 패턴에 관한 탐색적 연구: 텍스트 마이닝 접근

An Exploratory Study on the Covariation Patterns Between Organizational Culture Types and Turnover Intention Scores in Online Corporate Reviews: A Text Mining Approach

초록

This study explores the covariation patterns between organizational culture types and employees’ turnover intention scores in the Korean life insurance industry. Adopting the Competing Values Framework (CVF) as a theoretical framework rather than a hypothesis to be tested, we analyzed 2,425 online corporate reviews of three major life insurers—Samsung Life, Hanwha Life, and Kyobo Life—using a hybrid approach combining Latent Dirichlet Allocation (LDA) and KoBERT, a Korean-specific pre-trained language model. The text-based classification yielded distinct dominant culture types across firms: Samsung Life was classified as adhocracy-dominant, while Hanwha Life and Kyobo Life were classified as hierarchy-dominant. A comparison of turnover intention scores across organizational culture types showed that reviews classified as hierarchy-dominant exhibited statistically significantly higher turnover intention scores than those classified as adhocracy-dominant; however, the mean difference had only a limited statistical meaning. Although the results showed a one-to-one correspondence between a specific form of organizational culture and each company, there were statistical limitations in the initial analysis in distinguishing whether the observed covariation patterns originated from the organizational culture itself or from the unique characteristics of the respective companies. To address this, we additionally compared the differences in turnover intention scores within the single company, Hanwha Life, where the two forms of organizational culture are relatively balanced, and confirmed that the same covariation patterns as the overall sample were statistically marginal (p<0.1). This study explores the possibility of measuring and classifying organizational culture using unstructured text data and suggests that corporate reviews on open online platforms can capture perceptions of organizational culture in terms of external expression, distinct from internal surveys, thereby holding significance as a new methodological exploration for existing organizational culture research.

키워드

조직문화경쟁가치모형이직의도텍스트 마이닝머신러닝organizational culturecompeting values framework (CVF)turnover intentiontext miningmachine learning
제목
온라인 기업 리뷰에서의 조직문화 유형과 이직의도 점수 간 공변 패턴에 관한 탐색적 연구: 텍스트 마이닝 접근
제목 (타언어)
An Exploratory Study on the Covariation Patterns Between Organizational Culture Types and Turnover Intention Scores in Online Corporate Reviews: A Text Mining Approach
저자
설민준정구혁
DOI
10.55125/YBR.2026.06.63.1.93
발행일
2026-06
유형
Y
저널명
연세경영연구
63
1
페이지
93 ~ 121