Latent profiles of career-choice anxiety and associated factors among undergraduate nursing interns: a cross-sectional study

Technical Summary: Latent Profiles of Career-Choice Anxiety and Associated Factors Among Undergraduate Nursing Interns

Problem Statement
Undergraduate nursing interns face a critical transition from clinical training to employment, a period characterized by significant vulnerability to career-choice anxiety. Existing research has predominantly relied on variable-centered approaches using aggregate scores, which may obscure meaningful heterogeneity among individuals. Furthermore, the specific role of core self-evaluations in shaping distinct anxiety patterns within this population remains underexplored. This study addresses the need to identify unobserved subgroups (latent profiles) of career-choice anxiety and to examine the demographic, familial, and psychological factors associated with membership in these profiles.

Methodology

• Design and Setting: A cross-sectional study was conducted between May and August 2026 at three tertiary grade-A hospitals in a single province of China.
• Participants: The study included 421 valid respondents (out of 450 distributed questionnaires) who were full-time fourth-year students in a four-year bachelor's nursing program. Inclusion criteria required at least three months of cumulative clinical internship experience to ensure exposure to professional role demands and employment planning. Exclusion criteria included interrupted internships, suspension, or pre-existing psychiatric disorders.
• Measures:
  • Sociodemographic Questionnaire: Developed by the research team, covering demographics, family background (e.g., parental education, household income, only-child status), and employment-related variables (e.g., route into the major, career guidance received).
  • Core Self-Evaluations Scale (CSES): A 10-item Chinese revision assessing broad appraisals of competence and worth (Cronbach's $\alpha$ = 0.810).
  • Career-Choice Anxiety Symptom Questionnaire for College Graduates: A 26-item instrument measuring four dimensions: employment competition pressure, lack of employment support, insufficient self-confidence, and concern about employment prospects (Cronbach's $\alpha$ = 0.861).
• Statistical Analysis:
  • Latent Profile Analysis (LPA): Performed using Mplus 8.3 with the four anxiety dimensions as continuous indicators. Models with one to five profiles were compared using the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), sample-size-adjusted BIC (aBIC), entropy, Lo–Mendell–Rubin likelihood ratio test (LMR-LRT), and bootstrap likelihood ratio test (BLRT).
  • Regression Analysis: Multinomial logistic regression (SPSS 27.0) was used to examine factors associated with profile membership, using the "lower-anxiety" profile as the reference category.

Key Results

• Latent Profiles: The four-profile model was selected as optimal based on fit indices (Entropy = 0.921; LMR-LRT $P < 0.05$ for 2–4 profiles; BLRT significant for all). The four identified profiles were:
  1. Moderate Anxiety: 35.15% ($n=148$).
  2. Lower Anxiety: 11.40% ($n=48$).
  3. Low Anxiety: 8.55% ($n=36$).
  4. High Anxiety: 44.89% ($n=189$), the largest subgroup.
• Profile Characteristics: The high-anxiety profile exhibited the highest scores across all four dimensions (competition pressure, lack of support, low confidence, and concern about prospects). The low-anxiety profile showed the lowest scores across all dimensions.
• Associated Factors:
  • Low-Anxiety Profile Membership: Significantly associated with being an only child (OR = 6.06) and higher core self-evaluation scores (OR = 1.07).
  • High-Anxiety Profile Membership: Associated with lower parental educational attainment (specifically, higher parental educational attainment was associated with lower odds of membership in the high-anxiety profile), higher monthly household income (OR = 15.58 for >6,000 CNY), and entering the nursing major through program reassignment (OR = 5.43) compared to voluntary choice.
  • Moderate-Anxiety Profile: No statistically significant factors were identified in the multivariable model for this group relative to the reference.

Key Contributions
The study contributes a person-centered perspective to nursing education research by demonstrating that career-choice anxiety is not a uniform construct but exists in heterogeneous patterns. By identifying that nearly 80% of interns fall into moderate or high-anxiety profiles, the study highlights the prevalence of psychological burden during the internship period. Furthermore, it elucidates the complex relationship between socioeconomic factors and anxiety; specifically, it challenges the assumption that higher household income is universally protective, suggesting instead that it may correlate with higher expectations and anxiety in this specific context. The integration of core self-evaluations as a psychological resource associated with lower anxiety profiles offers a target for intervention.

Significance and Claims
The authors claim that these findings support the implementation of "profile-informed" career guidance and psychological support. They argue that nursing schools and clinical training hospitals should move beyond one-size-fits-all approaches. Instead, interventions should be tailored:

• High-anxiety interns may require individual assessment, structured career planning, and practical employment resources.
• Moderate-anxiety interns may benefit from preventive career-readiness programs and monitoring.
• Lower-anxiety interns may benefit from developmentally tailored opportunities to sustain confidence.

The study emphasizes that support strategies should consider core self-evaluations, family background, and the student's pathway into nursing. However, the authors maintain a modest stance regarding causality and generalizability. They explicitly note that the cross-sectional design precludes causal inference, the convenience sampling from a single province limits generalizability, and the reliance on self-report data introduces potential biases. Consequently, they conclude that while the findings are valuable for targeted support, longitudinal and multicenter studies are necessary before broader causal conclusions can be drawn.