PnB-W15 - Using a Hiring Simulation to Teach Statistical Reasoning and Responsible AI Use 

Information

In a Statistics II course, students engaged in a structured, scaffolded project simulating a job interview for a Data Analyst position. The scenario centers on a church board that previously relied solely on generative AI for data analysis but now seeks candidates who can add value beyond automation. Students may use ChatGPT (free version) as a support tool but must: critique and fact-check AI outputs, independently reproduce all analyses in SPSS or StatCrunch, identify limitations of AI responses, and demonstrate how human reasoning improves the final results. Students assume two roles: 1. Data Analyst Candidate – Conduct and defend analyses. 2. Church Board Member – Evaluate peers and recommend the strongest candidate. The project is introduced midway through the semester and scaffolded across the remaining weeks. Key components include: 1. Developing and refining a research question. 2. Prompting ChatGPT to generate potential research questions. 3. Critiquing AI-generated suggestions. 4. Conducting statistical analyses (e.g., ANOVA, regression, and other course concepts). 5. Designing peer evaluation rubrics. 6. Participating in a live defense/interview with peers and instructor. 7. Submitting a self-reflection on AI use and learning. The oral defense/interview ensures students demonstrate conceptual understanding rather than relying passively on AI output. By integrating AI critically, students experience the full cycle of applied data analysis: from question formulation to interpretation and ethical reflection on AI’s role in decision-making. This project was implemented in two face-to-face Statistics II courses, each enrolling 9–12 students, primarily mathematics, data science, and engineering majors, at a Lutheran university that integrates faith and ethical reflection into learning.