Common method deviation test
Due to the fact that all variables in this study were in the form of scales and self-reported by participants, a common method bias test is required. A common factor was established for confirmatory factor analysis[29], and it was found that the model fit was poor with c2/df=42.72, RMSEA=0.19, CFI=0.56, TLI=0.51, GFI=0.51, NFI=0.55. Therefore, there was no significant common method bias in this study.
The demographic characteristics of the participants
Among 1127 participants, 4.082% were males and 95.918% were females, with an average age of 32.783 ± 6.401. Clinical nurses have worked for 5 to 10 years (30.524%), are married (72.671%), hold a bachelor's degree (88.287%), and are not only children (84.206%) .(Table 1)
TABLE 1 Demographic characteristics of the participants (N = 1127)
Variables
|
N)
|
%
|
Gender
|
|
|
Female
|
1081
|
95.918
|
Male
|
46
|
4.082
|
Years of service(Years)
|
|
|
1~5
|
249
|
22.095
|
6~10
|
344
|
30.524
|
11~15
|
304
|
26.974
|
16~20
|
116
|
10.292
|
≥21
|
114
|
10.115
|
Education
|
|
|
Specialty
|
127
|
11.269
|
Undergraduate college
|
995
|
88.287
|
Graduate or above
|
5
|
0.444
|
Marital status
|
|
|
Married
|
819
|
72.671
|
Unmarried
|
289
|
25.643
|
Single
|
19
|
1.686
|
Are you an only child?
|
|
|
Yes
|
178
|
15.794
|
No
|
949
|
84.206
|
Pearson’s correlation analysis
The results of the mean, standard deviation, and correlation coefficient of the variables in this study are shown in Table 2. The clinical nurses' presenteeism scores were 4.623±1.953,while the frustration, job burnout,turnover intention scores were 9.128±2.869,43.294±14.686 and 8.113±3.341, respectively.Correlation analysis found a positive relation between presenteeism and frustration (r=0.239, P< 0.001), job burnout (r=0.276, P< 0.001) and turnover intention (r=0.305, P<0.001). In addition,there was an association between frustration and job burnout(r=0.480, P< 0.001) and turnover intention(r=0.386, P<0.001). Moreover, we also found a relation between job burnout and turnover intention (r=0.582, P< 0.001).
TABLE 2 Descriptive statistics and correlations of the study variables (N=1127)
|
Variables
|
Mean
|
Standard deviation
|
1
|
2
|
3
|
4
|
1. presenteeism
|
4.623
|
1.953
|
-
|
|
|
|
2. frustration
|
9.128
|
2.869
|
0.239***
|
-
|
|
|
3. job burnout
|
43.294
|
14.686
|
0.276***
|
0.480***
|
-
|
|
4. turnover intention
|
8.113
|
3.341
|
0.305***
|
0.386***
|
0.582***
|
-
|
***:P < 0.001,**:P < 0.01,*:P < 0.05
|
Mediating effect analysis
Establish a structural equation model with presenteeism as the independent variable, frustration and job burnout as the mediating variables, and turnover intention as the dependent variable. The results showed that the fitting indices of the model were good, with c2/df=5.24, RMSEA=0.06, CFI=0.98, TLI=0.97, GFI=0.97, NFI=0.97. According to the model results, it was found that after adding frustration and job burnout, the predictive effect of presenteeism on turnover intention is still significant(β= 0.107, P<0.001); Presenteeism has a significant predictive effect on frustration and job burnout, respectively(β= 0.275, 0.199, P<0.001); Frustration has a significant predictive effect on job burnout and turnover intention, respectively(β=0.129, 0.413, P<0.001); Job burnout has a predictive effect on turnover intention(β=0.547, P<0.001); The above results indicate that presenteeism not only directly affects the turnover intention of clinical nurses, but also indirectly affects the turnover intention of clinical nurses through frustration and job burnout. In addition, presenteeism can also affect the turnover intention of clinical nurses through frustration and job burnout. The model is shown in Figure 2.
Using the Bootstrap method, 5000 repeated samples were taken and a 95% confidence interval was calculated to verify the possible chain mediating effect of frustration and job burnout between presenteeism and clinical nurses' turnover intention. The results are shown in Table 2. The test results of the mediating effect show that the confidence interval of the overall mediating effect between frustration and job burnout does not include 0, indicating that the mediating effect of the two mediating variables between presenteeism and clinical nurses' turnover intention is significant. The mediating effect is composed of three indirect effects: firstly, the indirect effect of presenteeism on clinical nurses' turnover intention through frustration is 1, and its confidence interval does not include 0, indicating that the mediating effect of frustration between presenteeism and clinical nurses' turnover intention is significant, with an effect value of 0.036; Secondly, presenteeism has an indirect effect on the turnover intention of clinical nurses through job burnout, with a confidence interval not including 0, indicating a significant mediating effect of job burnout between presenteeism and clinical nurses' job burnout, with an effect value of 0.109; Thirdly, the indirect effect of presenteeism on clinical nurses' turnover intention through frustration and job burnout is 3. The confidence interval does not include 0, indicating a significant chain mediating effect of frustration and job burnout between presenteeism and clinical nurses' turnover intention, with an effect value of 0.062. In this study, the total mediating effect was 0.207, accounting for 65.92% of the total effect.
Table2:The mediating effect of frustration and job burnout on the relationship between presenteeism and turnover intention
|
|
intermediary
Effect value
|
95% confidence interval
|
Proportion to total effect
|
LLCI
|
ULCI
|
Presenteeism→Frustration→Turnover intention
|
0.036
|
0.014
|
0.062
|
11.46%
|
Presenteeism→Job burnout→Turnover intention
|
0.109
|
0.072
|
0.154
|
34.71%
|
Presenteeism→Frustration→Job burnout→Turnover intention
|
0.062
|
0.043
|
0.087
|
19.75%
|
Total mediating effect
|
0.207
|
0.165
|
0.260
|
65.92%
|
Direct effect
|
0.107
|
0.051
|
0.167
|
34.08%
|
Total effect
|
0.314
|
0.248
|
0.379
|
100%
|