Design and Participants
Nurses from 37 hospitals in Anhui Province, China, participated in this survey that was carried out in December 2021. The economic, social, and medical levels of Anhui Province are at the middle level of China. The results of the survey on the level of nurses’ caring behavior in this region could provide insight into the general caring behavior of Chinese nurses. The inclusion criteria were as follows: (a) having worked as a registered nurse for longer than six months; (b) being presently employed in a hospital as a registered nurse; and (c) being willing to participate in the investigation. The exclusion criteria were as follows: nurses who are on sick leave, pregnant, or have physical or mental illnesses. The sample size was determined by considering the variables’ influencing factors and multiplying that number by five to ten to get the sample size[31]. In this study, the general data included 11 items, and three questionnaires contained 60 items. Hence, we should collect 710 questionnaires from nurses and expand the sample size by 20% to offset the inadequate response. Therefore, the minimum sample size for this study should be at least 888 nurses.
Ethical Consideration
This study was an online survey. The preface of the questionnaire clearly introduced the purpose, significance, informed consent instructions, and general guidelines for participation in this study. If a nurse did not want to fill out the questionnaire, she could directly exit the QR code link. The voluntary completion of the questionnaire by a nurse indicated informed consent, and they were not required to sign the informed consent document. This study was approved by the Institutional Review Board of the X Hospital (Approval No. 20XX7).
Measures
A four-part questionnaire, comprising the General Information Questionnaire, the Authentic Leadership Questionnaire (ALQ), the Psychological Capital Questionnaire (PCQ), and the Caring Behaviour Inventory (CBI), was utilized to collect data.
General Information Questionnaire
It reflected the nurses’ demographic and sociological characteristics, such as gender, age, and years of employment.
Authentic Leadership Questionnaire
The 16-item ALQ, created by Walumbwa[13] and refined by Lin [32], was used to investigate authentic leadership. Kuo assessed the validity and reliability of the Chinese version of the ALQ[33]. The Chinese version of the ALQ includes four dimensions: self-awareness (four items), relational transparency (five items), morals and ethics (four items), and balanced processing (three items). In this questionnaire, answers are evaluated on a Likert scale from 1 (strongly disagree) to 5 (strongly agree). The higher the total score, the stronger the agreement. The Cronbach’s alpha coefficients for the ALQ in this survey were 0.96 overall, as well as 0.91, 0.90, 0.85, and 0.86 for each of the four dimensions.
Psychological Capital Questionnaire
The psychological capital of nurses was assessed using the Chinese version of the PCQ[25, 34]. The questionnaire comprises four dimensions: self-efficacy (six items), hope (six items), resilience (five items), and optimism (three items). Each item was scored from 1 (strongly disagree) to 6 (strongly agree). The total scores ranged from 20 to 120 points; the higher the score, the more psychological capital is represented. In this survey, Cronbach’s alpha coefficients for the PCQ and its four dimensions ranged from 0.72 to 0.92.
Caring Behavior Inventory
The Chinese version of the CBI was utilized to evaluate the caring behavior of nurses[35]. The inventory comprises 24 items and four dimensions: a) assurance, which is instantly available to respond to patients’ concerns and security; b) knowledge and skill, which reveal diligence and competence; c) respectfulness, which is related to the dignity of the individual; and d) connectedness, which enables patients to receive support every time [35]. Each item was scored from 1 (never) to 6 (always). The total score ranged from 24 to 144, the higher the score, the more caring the nurse. Cronbach’s alpha coefficients for the CBI and its four dimensions in our research range from 0.89 to 0.96.
Data collection
In December 2021, the researchers used the Questionnaire Star website to make this study questionnaire and formed a QR code of survey. Then researchers contacted the nursing management departments of 37 hospitals, trained 37 assistant researchers form 37 hospitals. Next, 37 assistant researchers enrolled the nurses in 37 hospital and issued the QR code to the nurses to fill out. The preface of the questionnaire introduced the purpose, significance, informed consent, instructions, and guidance for filling it out and the contact information of the researchers. Furthermore, only one questionnaire could be filled out using a mobile phone number. The QR code on this study questionnaire would automatically expire after seven days. Finally, we received online responses from 3,662 nurses. All collected responses were checked by the investigators, and 167 questionnaires whose answers had missing data or showed obvious regularity, were eliminated. In the end, 3,495 valid surveys were collected, yielding a 95.4% efficiency rate.
Statistical analysis
SPSS 24.0 and AMOS 21.0 were used for data analysis. The standard deviation and frequency percentages were used to describe the score status of each analysis variable and the basic characteristics of the sample. The data were normally distributed in the study, wherein Pearson’s correlation coefficient (r) was used to conduct correlation analysis of study variables.
By constructing Structural Equation Modeling (SEM), it is possible to examine a series of interactions between one or more independent variables that may be either continuous or discrete and one or more dependent variables that can also be either continuous or discrete. SEM was used to confirm if psychological capital mediates the relationship between authentic leadership and the caring behavior of nurses. Caring behavior, authentic leadership, and psychological capital acted as the dependent variable, independent variable, and mediator in the SEM, respectively. According to Hu and Bentler’s recommendation[36], the following goodness-of-fit statistics were used to evaluate the fit degree of model and data: Root Mean Square Error of Approximation (RMSEA) with an acceptance level of <0.08, Comparative Fit Index (CFI) with an acceptance level of >0.90, Goodness of Fit Index (GFI) with an acceptance level of >0.90, Normal Fit Index (NFI) with an acceptance level of >0.90, Tucker-Lewis Index (TLI) with an acceptance level of >0.90 and Standardized Root Means Square Residuals (SRMR) with an acceptance level of < .05. Because the sample size is larger than 500, it is not considered chi-square value (χ2) and χ2/df ratio.
A 5,000-sample bootstrapping procedure with a 95% confidence interval (CI) was used to estimate and test the mediating effects[37]. The p-value was less than 0.05, which was considered as significant. All significant tests were two-tailed with a 5% level of significance.