The development of a shortened, Korean version of the Anticipated Turnover Scale for hospital nurses

Abstract

This study was a psychometric validation of a Korean version of the Anticipated Turnover Scale (ATS) involving translation, transcultural adaptation, and validation. We used data from a cross-sectional online survey of 287 Korean practicing nurses in hospitals. Confirmatory factor analysis (CFA) assessed the validity of the one-factor structure that was hypothesized in the original version. We removed any items that did not enhance the scale's reliability and validity. Test and retest reliability was performed on 206 of the nurses who completed the survey again two weeks later. The mean age of the 287 nurses was 33 years, and 278 (97%) of them were female. The final 6-item Korean version of the ATS demonstrated a satisfactory model fit in the CFA (χ2 = 22.67; degree of freedom = 9; goodness-of-fit index= 0.97; comparative fit index = 0.98; Tucker–Lewis Index = 0.97; root mean square error of approximation = 0.07; and standard root mean residual = 0.03). The intraclass correlation coefficient with the test–retest reliability for the 2-week interval was 0.81, with a good internal consistency (Cronbach's alpha = 0.85; McDonald's omega coefficient reliability = 0.92). The 6-item revised ATS is a reliable and valid scale to assess the turnover intention of Korean nurses who work in hospitals. It can be used in an acute setting and predict the anticipated turnover rate for monitoring purposes.