Sampling and data collection
This study employed a method of random and convenient sampling. The research team first identified the target population as registered nurses from six top general hospitals. After obtaining approval from the nursing departments of each hospital, the team developed a questionnaire distribution plan based on the number of nurses and work shifts in various departments to ensure diversity. Departments with high and low workload levels were selected for comparison. The questionnaire underwent initial design and expert review to ensure its validity and reliability.
In each hospital, nurses were randomly selected from different departments to ensure participant diversity. The number of nurses selected in each hospital was proportionally allocated based on the total number of nurses and the study’s needs. The research team scheduled time in multiple departments to explain the study’s purpose and the importance of the questions to the nurses. The questionnaire was distributed to each department in rotation and handed out directly by project team members to ensure that nurses understood the survey’s objectives and were encouraged to respond based on their actual experience. To ensure participants could freely express their opinions, the questionnaire included a statement outlining the study’s goals and guaranteeing the anonymity of personal information.
After completing the questionnaire, nurses placed it in a sealed envelope provided by the research team to protect their privacy. This study strictly adhered to the principles of data anonymity and confidentiality and received support from the Ethics Committee of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine. Project team members were responsible for collecting the completed questionnaires and ensuring that all questionnaires were safely returned to the research team for data collation within the designated timeframe. During the collection process, the team checked for incomplete or invalid responses and recorded disqualified questionnaires for exclusion from the final analysis.
Through this sampling and questionnaire distribution process, a total of 778 nurses completed the survey effectively, achieving a response rate of 91.5%, which ensured the validity of the study results.
Survey instruments
The study utilized the following tools for data collection and analysis: Demographic Questionnaire, Conditions for Work Effectiveness Questionnaire II (CWEQ-II), Psychological Conditions Questionnaire (PCQ), and Utrecht Work Engagement Scale-9 (UWES-9).
The Demographic Questionnaire, designed by the researchers, collected basic information from participants, including gender, age, education, workplace, and years of experience.
CWEQ-II consists of 20 items and aims to measure nurses’ perceptions of structural empowerment. The scale includes 12 items rated on a 6-point Likert scale (from 1 = none to 6 = very much). The total score is obtained by averaging four subscales (information, support, resources, and opportunity), with lower scores indicating a lack of perceived empowerment in the work environment.
PCQ contains 21 items to assess nurses’ psychological capital, covering self-efficacy, optimism, hope, and resilience. All items are rated on a 6-point Likert scale (from 1 = strongly disagree to 6 = strongly agree), with higher scores indicating a higher level of psychological capital.
UWES-9 consists of 22 items to measure nurses’ work engagement, encompassing vigor, dedication, and absorption. This scale also uses a 6-point Likert scale (from 0 = never to 5 = always), with higher scores indicating higher work engagement.
Regarding reliability and validity, the Cronbach’s alpha coefficients for these scales ranged from 0.94 to 0.96, with factor loadings exceeding 0.5 and adjusted fit indices over 0.9, indicating that the scales have good reliability and validity, providing a solid foundation for subsequent data analysis.
Overall, this study selected CWEQ-II, PCQ, and UWES-9 as assessment tools based on their wide application in relevant fields and their significant impact on nurses’ professional performance. Structural empowerment is measured through factors such as support, information, resources, and opportunities in the workplace, while psychological capital focuses on individual self-efficacy, optimism, hope, and resilience. Together, these two factors influence nurses’ work engagement, ultimately reflected in the dimensions of vigor, dedication, and absorption.
Data analysis
This study utilized SPSS version 27.0 for data analysis. Initially, descriptive analysis was conducted on demographic data as categorical variables, with distribution presented in percentages. Descriptive analysis of the main study variables (structural empowerment, psychological capital, and work engagement) was also performed, with mean (M) and standard deviation (S.D.) used to describe their distributions. Pearson correlation analysis was applied to examine the relationships between variables. In the regression analysis, each dimension of the CWEQ-II and PCQ was included as an independent variable to analyze structural empowerment and psychological capital, while demographic data were used as independent variables to analyze work engagement. Finally, in the complete model, demographic data and each dimension of the CWEQ-II and PCQ were included as independent variables, showing that gender and all dimensions of structural empowerment and psychological capital entered the regression model.
This study also employed structural equation modeling (SEM) to analyze the relationships among variables. Based on the literature and theoretical framework, an initial model was constructed, including independent and dependent variables. The collection of questionnaire data ensured the representativeness of the sample, with 778 valid samples obtained. Prior to SEM analysis in AMOS, data preprocessing was conducted, including handling missing values, normality testing, and outlier detection, to meet SEM analysis requirements. Maximum likelihood estimation (MLE) was used for parameter estimation. A path diagram was created in AMOS, and standardized path coefficients were calculated. Model fit was evaluated using chi-square (χ²), Comparative Fit Index (CFI), Adjusted Goodness-of-Fit Index (AGFI), and Root Mean Square Error of Approximation (RMSEA). Subsequently, the significance of the path coefficients was tested to verify the hypotheses, and the direct and indirect effects among variables were analyzed. Ultimately, the relationships among latent variables were explored in depth based on the model results, and the practical implications and theoretical contributions of the study findings were discussed.