Nursing interns who rated their internship environment more positively also reported higher levels of voice behavior, according to a survey of 498 interns at three tertiary teaching hospitals in Hunan Province, China. The researchers examined overall voice behavior alongside two forms they called promotive and prohibitive voice.
The study also found statistical indirect pathways involving self-efficacy and professional identity. For overall voice behavior, the analysis examined each measure separately and the two together. That wording matters: an indirect effect in this analysis describes a statistical relationship among the measures, not proof that a better internship environment caused interns to speak up.
Two statistical pathways for voice
In the overall model, the estimated total association between internship environment and voice behavior was B = 0.228, with a 95% confidence interval of 0.199 to 0.257. The direct estimate in the same model was B = 0.100, with an interval of 0.071 to 0.130. The indirect estimates were 0.050 through self-efficacy, 0.036 through professional identity and 0.042 through both together; their 95% bootstrap intervals were 0.027 to 0.071, 0.023 to 0.051 and 0.027 to 0.058, respectively. The combined indirect-to-total ratio was reported as 55.98%, a descriptive figure rather than a causal share.
For promotive voice, the reported indirect estimates were 0.019 through self-efficacy, 0.017 through professional identity and 0.020 through their sequential pathway. The corresponding intervals were 0.006 to 0.032, 0.010 to 0.026 and 0.011 to 0.030. The total indirect estimate was 0.056 against a total effect of 0.114, producing a descriptive ratio of 49.47%.
The same analysis for prohibitive voice produced indirect estimates of 0.031 through self-efficacy, 0.019 through professional identity and 0.022 through their sequential pathway. The intervals were 0.020 to 0.042, 0.012 to 0.026 and 0.014 to 0.030. Here, the total indirect estimate was 0.071, compared with a total effect of 0.114, and the descriptive ratio was 62.48%.
The measures rose together
The three main measures also moved together in the pairwise analysis. Internship environment correlated with overall voice behavior at 0.553, with self-efficacy at 0.498 and with professional identity at 0.538. Self-efficacy correlated with professional identity at 0.732 and with voice behavior at 0.639; professional identity correlated with voice behavior at 0.678. Each reported correlation had P < 0.001.
A survey, not an experiment
Researchers distributed 520 questionnaires and included 498 valid responses, a 95.8% response rate. The participants had a mean age of 21.09 years, 85.1% were female, and 56.8% reported moderate internship stress. They completed questionnaires covering internship environment, self-efficacy, professional identity and voice behavior.
Data were collected from January to February 2026 through voluntary anonymous paper or online questionnaires. Each participant could respond only once, and incomplete responses were excluded. The study used convenience sampling.
The analysis used Pearson correlations, regression and PROCESS Model 6, a mediation procedure, with 5,000 bootstrap resamples to produce 95% confidence intervals. The researchers treated an indirect pathway as statistically significant when its interval excluded zero and used P < 0.05 as the significance threshold. They also reported that the first factor in Harman's test explained 31.72% of variance, below the stated 40% threshold, while variance inflation factors ranged from 1.037 to 2.479, indicating no serious multicollinearity.
What the study cannot establish
The cross-sectional design means the study can show associations among the measures, but it cannot establish that the internship environment came first or caused changes in voice behavior. The sample also came from three hospitals through convenience sampling, so the results should be read as applying to the studied group rather than as evidence of a universal pattern. Finally, the indirect-to-total percentages are descriptive and should not be read as the share of voice behavior caused by self-efficacy or professional identity.
Study disclosures
The study received institutional approval, and participants gave written informed consent. The authors declared that no financial support was received for the work or its publication, and they reported no commercial or financial relationships that could be construed as a potential conflict of interest. They also stated that raw data supporting the conclusions will be made available without undue reservation.
Paper data and sources
Original title: The mediating roles of self-efficacy and professional identity in the association between internship environment and voice behavior among nursing interns: a multicenter cross-sectional study
Authors: Xiao T, Cui W, He M et al.
Journal/Repository: Frontiers in psychology
Status: Peer-reviewed
First online: 2026-10-19
DOI: Not available
Original paper · Full text