Nursing students’ risk attitudes were the strongest signal in the safety climate model
A path analysis of 591 undergraduate nursing students at Gaziantep University in Türkiye draws a sharp line between emotional strain and practical safety culture. Occupational anxiety was common, but it was not the variable that moved perceptions of the clinical learning safety climate. Instead, students with more positive workplace risk attitudes consistently reported a more favorable safety climate, a relationship that held in the final path model and explained 27% of the variance.
The finding matters because nursing education lives at the intersection of stress, supervision, and exposure to real-world hazards. Clinical placements train students to recognize biological, chemical, ergonomic, and procedural risks while also asking them to function under pressure in unfamiliar environments. When the environment feels psychologically unsafe or poorly organized, students may learn to normalize hazard rather than manage it. Here, the data suggest that attitude toward risk is the more immediate lens through which students judge whether a clinical setting feels safe enough for learning.
Anxiety was high, but it did not drive the safety climate score
Participants reported moderate-to-high occupational anxiety, with a mean score of 69.6 ± 17.3 on the occupational anxiety scale for healthcare students. Their attitudes toward workplace risks were generally positive, with a mean of 144.5 ± 22.5 on the nurses’ attitudes toward workplace risks scale, and their perceptions of the clinical learning safety climate were moderately favorable, with a mean of 71.9 ± 11.3 on the safety climate questionnaire.
These are not abstract survey totals. In the context of nursing education, they map onto how students behave in wards, how they interpret supervision, and whether they feel able to speak up when something looks unsafe. A student may feel anxious about making an error and still believe a unit takes risk management seriously. Another may feel less anxious yet remain poorly prepared to recognize hazards. The path model separated those constructs cleanly.
Occupational anxiety showed no meaningful association with safety climate perceptions, either directly or indirectly through workplace risk attitudes. The direct path coefficient was essentially null, with β < 0.001 and p = 0.997. That abrupt lack of association is the study’s central surprise. Emotional pressure alone did not predict whether students viewed their clinical environment as safe; the more important variable was whether they had internalized practical risk awareness and a more constructive orientation toward workplace hazards.
Gender, nationality, and academic standing tracked different anxiety patterns
Group comparisons showed clear differences in how anxiety and risk attitudes clustered across student subgroups. Occupational anxiety was significantly higher among female students, Turkish students, and students with a grade point average of 2.00 or above, with all comparisons reaching p < 0.001. More positive workplace risk attitudes, by contrast, were seen among male students and international students, with p = 0.029 and p = 0.011, respectively.
Students who reported no trust in healthcare institutions or said they had inadequate knowledge of occupational risks had higher anxiety and less favorable safety climate perceptions, both at p < 0.001. Those patterns fit a straightforward clinical education logic: if students do not trust the institutions where they train, and if they do not feel equipped to identify hazards, the placement becomes a place of uncertainty rather than supervised practice. Anxiety in that setting is not just a personal emotional response; it can become a marker of unresolved educational gaps.
Internal consistency was strong, and the path model was stable
The study used three established instruments: the occupational anxiety scale for healthcare students, the nurses’ attitudes toward workplace risks scale, and the safety climate questionnaire. Internal consistency was high across the total scales, with Cronbach’s alpha ranging from 0.94 to 0.97. That level of reliability strengthens confidence that the measures captured coherent constructs rather than noise.
Statistical testing included independent-samples t tests, one-way ANOVA with Bonferroni correction, and Welch or Games–Howell procedures when variance assumptions were not met. The path analysis was performed using AMOS 29.0. Within that model, workplace risk attitudes emerged as the dominant explanatory factor for safety climate perceptions. The standardized beta of 0.52 reflects a substantial positive association in the model, not a minor statistical flourish.
A 27% explanation of variance leaves room for other influences, of course, but it is enough to matter in day-to-day training. Students who learn to recognize hazards, communicate concerns, and interpret safety rules as part of professional identity appear more likely to experience their placement as safer and more coherent.
Clinical placement design may matter more than stress reduction alone
The practical implication is not that anxiety should be ignored. It should not. High anxiety can impair attention, reduce confidence, and make novice learners less willing to ask for clarification. But this dataset suggests that lowering anxiety by itself may not be sufficient to improve how students experience the safety climate of clinical learning sites.
A more direct lever may be curriculum design. Risk management content, supervised exposure to workplace hazards, and explicit instruction on reporting and prevention may shape how students assess the environments in which they train. Psychological safety in placements also matters, because students who feel dismissed or excluded are less likely to translate safety knowledge into action. Professional motivation plays a role too; when students see safety as part of nursing identity rather than an administrative burden, they may engage with hazards more actively and more intelligently.
There is a clear bedside consequence here. Students who do not trust the system, or who cannot read the system’s risks, are less prepared to protect themselves and their patients.
The study’s limits are tied to design, setting, and measurement
This was an analytical cross-sectional study, so the data capture association rather than causation. The sample came from a single university, which limits how far the findings can be extended to other institutions, health systems, or national contexts. The survey design also depends on self-report, leaving room for response bias and differences in how students interpret the scales.
Even with those constraints, the study offers a useful signal for nursing education. Safety climate during clinical learning appears to be shaped less by anxiety itself than by how students think about risk, institutional trust, and the practical norms of the placement environment. Further work in multi-site cohorts would need to test whether those relationships hold across other nursing programs and whether targeted educational changes can shift both workplace risk attitudes and safety climate perceptions over time.