Development and Implementation of an Online OSPE Test Bank Graded by Artificial Intelligence
Bibliographic record
Abstract
The development of online anatomy education has rapidly accelerated during the COVID‐19 pandemic. This shift to the online world has mainly been focused on the delivery of content, while testing anatomical knowledge has proven to be more challenging, particularly for objective structured practical exams (OSPEs, also known as “spot tests” or “practical exams”). Online resources for OSPEs are uncommon compared to the extensive banks of multiple‐choice questions available. Another issue is that whether virtual or in‐person, grading OSPEs is challenging and time consuming. Recent research in our laboratory has suggested that machine learning algorithms, using decision trees, can be trained to mark OSPEs with a >95% accuracy. Building on these findings, the goal of this project is to create a virtual OSPE bank, train the AI to grade OSPEs, and develop an application with automated AI grading to act as a resource for students studying anatomy and physiology. Currently, we have written over 120 OSPE question sets using images from the Bassett Collection, the UBC Neuroanatomy collection, and images developed at the Education Program in Anatomy at McMaster University. The questions and answers were initially developed by senior undergraduate students, with coaching from faculty and staff familiar with OSPE generation acting as experts. The questions were then collectively reviewed by the students before undergoing two independent reviews by the experts. After revision, a final blind review of the questions was undertaken by a third expert to ensure validity and accuracy. These questions are being made available on the undergraduate anatomy and physiology course learning management system (LMS), Avenue to Learn. On this LMS, students will be able to use the questions for OSPE practice and answers will be available in the traditional manner. The answers given by the students and graded by the faculty will be collected and analyzed to determine the difficulty and discrimination of the questions. The data and analyses will be used to refine additional OSPEs, as well as refine the AI marking tool for the virtual OSPE application. Our hypothesis is that by providing instant and accurate feedback on valid and accurate questions, course evaluations and performance on OSPEs will improve. The research group is actively searching for collaborators willing to generate and review additional OSPE questions to expand the question bank and improve the AI marking tool.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.014 | 0.029 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.004 |
| Open science | 0.004 | 0.003 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.016 | 0.011 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".