Cluster analysis of self-concept and job satisfaction in Chinese nurses with master's degree to identify their turnover intention: A cross-sectional study.

AIMS AND OBJECTIVES

This study aimed to implement cluster analysis of self-concept and job satisfaction to identify subgroups in nurses with master's degree and explore the associations of turnover intention with characteristics among these clusters.

DESIGN

A cross-sectional study adhering to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE).

METHODS

A convenience sample of 408 nurses with master's degree in China filled out the survey from 19 November 2019 to 30 December 2019. A sociodemographic questionnaire, the Nurses' Self-Concept Questionnaire (NSCQ), Job Satisfaction Scale (JSS) and Turnover Intention Questionnaire (TIQ) were adopted to collect the data. K-means cluster analysis was implemented on the R software, and data were analysed using SPSS 24.0.

RESULTS

Three subgroups were identified based on cluster analysis of NSCQ and JSS subscales in 405 nurses (99.3%) available for statistical analysis, among whom 30.9%, 17% and 48.1% were allocated to these clusters respectively. Turnover intention significantly differed among the three clusters, with cluster 2 having the highest turnover intention and cluster 1 having the lowest turnover intention. Working department, position, professional title, clinical nurse specialist and annual income were factors differentiating TIQ scores in each cluster.

CONCLUSIONS

This study identified three clusters of nurses with master's degree and showed that each cluster was associated with the level of turnover intention. The unique characteristics of the three clusters may be also helpful in identifying and providing specific managerial or social support to reduce turnover rates in nurses with master's degree.

RELEVANCE TO CLINICAL PRACTICE

Cluster analysis is s an unsupervised machine learning method to identify meaningful subgroups within heterogeneous population based on variables distributions and patterns underlying in the data set. Through clustering, nurses with multi-dimensional characteristics could be allocated into subgroups associated with turnover intention. As a result, nursing managers could provide approaches for each subgroup to reduce turnover intention.