*2.2. Measurement*

## 2.2.1. Job Engagement

To measure job engagement, this study used the tool developed by Kim [32], which was a modified translation of the Utrecht Work Engagement Scale (UWES) developed by Schaufeli and Bakker [33] accompanied by Kim's supplements. This tool consists of 17 questions in three sub-areas: energy (six questions), sacrifice (five questions), and immersion (six questions). Typical questions include "When I wake up in the morning, I want to go to work," "I can work for a long time at a time," and "I feel proud of my work." The questionnaire consisted of a five-point scale in which one point represents strongly disagree and five points indicates strongly agree. The higher score on the five-point Likert scale signifies a higher level of job engagement. In this study, Cronbach's α was 0.91, and the sub-areas were 0.87 for energy, 0.84 for sacrifice, and 0.82 for immersion. The exploratory factor analysis for Job Engagement is given in Table 1.


**Table 1.** Results of Exploratory Factor Analysis (EFA) for Job Engagement Scale.

#### 2.2.2. Job Retention Intention

As the measurement tool for job retention intention, this study used questions from the Theory of Planned Behavior (TPB) developed by Ajzen [34] that measured behavioral intention. The Theory of Planned Behavior consists of a structure with four independent factors including attitudes about behavior, subjective norm, perceived behavioral control, and behavioral intention. In this study, only the three questions related to behavioral intention were used. Job engagemen<sup>t</sup> questions consisted of "I will continue my career next month," "I will continue my career after six months," and "I will continue my career after a year." The questionnaire consisted of a five-point scale with one point indicating strongly disagree and five points meaning strongly agree. The behavioral intention factor is composed of a five-point Likert scale, and in this study, the higher the score, the greater the nurse's intention to keep the job. The reliability analysis, Cronbach's α, was found to be 0.87.

## 2.2.3. Social Support

This study organized questions based on the content and composition of social support in the studies of Shirey [14] and Choo [35], who studied nurses. Social support consisted of three questions in total: "experience of social support," "social support categories," and "types of social support." The questions were as follows: "Have you ever received or seen social support to nurses during COVID-19? Who did you receive social support from? What type of social support did you receive?" The social support questions are all composed of nominal measures.

## *2.3. Data Analysis*

To achieve the study purpose, all the analyses for this study were performed using SPSS Ver. 25.0 (IBM, Armonk, NY, USA) after encoding and data cleaning. In order to examine socio-demographic factors, frequency analysis and descriptive analysis were conducted, and to obtain reliability and validity of the measuring instrument, Cronbach's α and exploratory factor analysis were carried out. Lastly, a *t*-test and one-way ANOVA were executed to investigate di fferences in job engagemen<sup>t</sup> and job retention intention depending on the independent variables (characteristics of the subjects and types of social support).
