By Samantha Robinson (University of Arkansas)
Information
This poster will present details about a previously (and currently) implemented framework for designing human centered assessments in an agricultural statistics course. While the implementation described will focus on agricultural statistics, the ideas are applicable to a broad range of applied statistics courses e.g., educational statistics, psychology statistics, business statistics, etc. As AI tools become increasingly capable of generating analytical results, creating data visualizations, drafting written interpretations, and identifying common limitations in traditional take-home or project-based statistics and data science assignments, students (especially non-majors that are often taking such courses primarily for degree requirements rather than out of genuine interest) are at risk of undermining their own learning by taking shortcuts. The proposed poster will review several different assignment types that intentionally cultivate discovery, curiosity, and authentic engagement by asking students to draw upon their own individualized knowledge and research interests. Moreover, by integrating career readiness into the course experiences via our university's Career Everywhere Catalyst Program, the assessments aim to provide students opportunities for knowledge building that AI cannot. Using examples from a graduate agricultural statistics course, the poster will showcase assessments that have evolved alongside Generative AI improvements over the past two years. These assessments require personal data collection, emphasize critical thinking, involve domain specific reasoning, and include communication across a variety of settings - areas where student insight is essential and AI can serve only as a supplemental tool, not a replacement, for learning. This information comes from an R1 institution located in the Southern United States that serves as a flagship university with approximately 35 to 40 graduate-level (MS or PhD) students pursuing degrees in a diverse College of Agricultural, Food, and Life Sciences. This first, graduate-level statistics course is a prerequisite course for all of the other graduate-level statistics course offerings introducing students to applied statistical methods, basic experimental design, and the use of statistical software (R/RStudio). Evidence of efficacy for the ideas provided in the poster will consist of institutional course evaluations, informal qualitative course surveys and student emails, and longer-term metrics e.g., conference presentations, inclusion of assessment feedback into theses, publications, future statistics course enrollments, and statistical micro-certificates earned.