Artificial Intelligence (AI), Educational Technology, Employment Anxiety, Midwifery Education, Nursing Education Research, Perceived Employability, Simulation Training
Conditions
Keywords
ChatGPT, Artificial Intelligence, Job Interview Simulation, Nursing Students, Midwifery Students, Employability, Interview Anxiety, Self-Efficacy, Simulation-Based Learning
Brief summary
This randomized controlled trial aims to evaluate the effectiveness of a ChatGPT-based job interview simulation on the employment perceptions and interview-related anxiety of senior nursing and midwifery students. Transitioning from education to professional practice in healthcare is a critical phase that directly influences employability and career readiness. Particularly for nursing and midwifery students, the ability to navigate job interviews with confidence plays a pivotal role in shaping their future career paths. As such, innovative and digital interventions are needed to better prepare students for this process. Grounded in Bandura's Social Cognitive Theory and the Technology Acceptance Model (TAM), the study explores how AI-driven simulations affect students' self-efficacy, perceived utility, and usability, and ultimately their career-related outlook. The intervention involves a structured, text-based job interview simulation powered by ChatGPT-4o, during which students respond to a series of nine professionally tailored questions. These questions are aligned with international competency frameworks such as those from ICN (2008) and ICM (2024), focusing on themes like professionalism, teamwork, evidence-based care, communication, and leadership. At the end of the simulation, the chatbot provides brief, constructive feedback to the participant. A total of 102 final-year students from Koç University and Istanbul University-Cerrahpaşa will be recruited using stratified randomization. Participants will be assigned to either an intervention group, which will complete the ChatGPT simulation, or a control group, which will not receive any interview intervention but will complete the same pre- and post-test questionnaires. Key outcome measures include the Perceived Future Employability Scale (PFE), the Interview Anxiety Scale (MASI-T), and a simulation experience form for the intervention group. Quantitative data will be analyzed using SPSS with appropriate parametric and non-parametric tests based on data distribution, and an intention-to-treat (ITT) approach will be adopted. To ensure the integrity of the experiment, blinding procedures, strict confidentiality, and group separation protocols will be applied. The simulation will be conducted individually on research-owned devices in private rooms, and no personal or textual data will be saved from the AI interactions. Ethical approval has been obtained from Koç University Social and Behavioral Ethics Committee. Participation is voluntary, informed consent will be collected, and all processes will comply with the Helsinki Declaration and Turkish Personal Data Protection Law. Ultimately, this study seeks to offer evidence on the pedagogical utility of AI-based simulation tools in preparing healthcare students for employment, while also contributing to the broader field of digital transformation in health education.
Detailed description
This study introduces an innovative AI-based intervention aimed at enhancing employability preparation among senior nursing and midwifery students through a simulated job interview experience. The intervention is structured around a generative artificial intelligence model (ChatGPT-4o), which delivers a dynamic, text-based interaction mimicking realistic interview settings. The chatbot prompts are informed by international professional competency frameworks and were developed with input from nursing educators, digital learning experts, and HR professionals. The conceptual basis of the study combines Social Cognitive Theory and the Technology Acceptance Model, allowing exploration of how simulated, interactive environments contribute to professional readiness. In particular, the study investigates the influence of self-efficacy beliefs and user perception of AI technology on employment-related attitudes. To ensure validity and usability, a pilot implementation was conducted prior to full deployment. This included iterative testing of the chatbot dialogue flow, question clarity, and interview timing, with input from students outside the main sample. Minor refinements were applied to optimize realism and reduce cognitive overload during simulation. The AI interaction does not involve natural language learning or data retention; its sole function is to facilitate reflective engagement in a structured scenario. All responses are deleted after the session to comply with data protection standards. The chatbot-generated feedback is non-evaluative and designed to reinforce confidence. Quantitative analysis will be supplemented with descriptive insights from a post-simulation feedback form that captures students' subjective experience, perceived preparedness, and usability impressions. This mixed-methods integration supports a comprehensive understanding of the intervention's practical and pedagogical value. Ultimately, the project aims to contribute to the responsible implementation of AI tools in health professions education, particularly in the context of employability, professional identity formation, and digital literacy.
Interventions
OTHERAI-Based Interview Simulation
The intervention group participated in a structured, text-based job interview simulation supported by ChatGPT-4o, using nine standardized prompts based on ICN (2008) and ICM (2024) core competencies. Developed by experts in simulation and nursing education, the intervention aimed to enhance self-efficacy and reduce interview-related anxiety. Each session lasted 15-20 minutes and concluded with structured feedback provided by ChatGPT. Participants were also offered an optional preparation guide with 40 reflective questions across themes such as communication, anxiety, and self-awareness. This preparation was not mandatory and not included in the simulation time.
Sponsors
Koç University
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
OTHER
Masking
DOUBLE (Subject, Investigator)
Masking description
Group assignment was concealed using a sealed-envelope method. Randomization was conducted by an independent researcher who was not involved in data collection or intervention delivery. Both participants and investigators were unaware of group allocation until the day of implementation.
Intervention model description
Interventional Study Model: Parallel Assignment Model Description: Participants will be randomly assigned to one of two parallel groups. The intervention group will receive a ChatGPT-based job interview simulation, while the control group will receive no intervention. Both groups will complete identical pre- and post-intervention assessments.
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
Yes
Inclusion criteria
Must be 18 years of age or older Must be a senior-year student enrolled in a nursing or midwifery program Must voluntarily agree to participate in the study Must provide written informed consent
Exclusion criteria
Unable to participate in the intervention or assessments due to extended absence or medical leave Fails to complete either the pre-test or post-test assessments Previous participation in the pilot phase of the study Prior exposure to the ChatGPT-based interview simulation used in this study
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Change in Perceived Future Employability Score | From baseline to post-intervention (approximately 2 weeks) | Assessed using the Perceived Future Employability Scale (PFE). The scale includes 24 items across 6 subdomains and is rated on a 6-point Likert scale. Higher scores indicate more positive perceptions of future employability. Change in total score from pre-test to post-test will be used to assess the intervention's effect. |
| Change in Interview Anxiety Score | From baseline to post-intervention (approximately 2 weeks) | Assessed using the Turkish version of the Measure of Anxiety in Selection Interviews (MASI), originally developed by McCarthy and Goffin (2004) and adapted into Turkish by Turgut, Kümbül Güler, and Vural Yüzbaşı (2024). The scale consists of 30 items across four subscales: communication anxiety, social appearance anxiety, performance anxiety, and behavioral anxiety. Each item is rated on a 5-point Likert scale (1-5). Higher scores indicate higher levels of job interview-related anxiety. Change in total score will be analyzed to evaluate the effectiveness of the intervention. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Simulation Experience Evaluation Score | Immediately after the simulation session | Participants in the intervention group will complete a post-simulation feedback form designed by the researchers. The form includes 8 closed-ended items rated on a 10-point Visual Analog Scale (VAS) and 3 open-ended questions. Dimensions include realism of the simulation, clarity of chatbot prompts, comfort level, perceived usefulness, performance perception, and readiness for real interviews. |
Countries
Turkey (Türkiye)
Outcome results
None listed