Moderate evidence
Published recordEducationGood Space
More
Beyond the tool: what a course-aligned AI assistant reveals about student learning in microbiology
- Truvace record
- TRV-2026-3034raw data (JSON)
- Version
- v1Published
- First published
- Evidence type
- Observational study
- Sources
- 1 sourcePeer-reviewed
- sha256
- 0d9195edb870…check it
Course-aligned artificial intelligence (AI) tutors are increasingly offered as learning support in higher education. Beyond providing practice and feedback, they may also help teachers identify gaps in understanding across a cohort. In a second-year medical microbiology and immunology unit of 500 to 800 students, a course-aligned AI assistant built on The University of Sydney’s Cogniti platform gave students scenario-based short-answer practice with rubric-referenced feedback. Its conversation logs turned out to…
The gain
In a University of Sydney microbiology and immunology unit of 500 to 800 students, logs from more than 1300 conversations with a Cogniti-based AI assistant revealed cohort-level knowledge gaps and recurring terminology errors mid-semester, which teachers fed back to the class.
What this doesn’t fix
The link between practice-oriented use of the assistant and final short-answer marks was exploratory, small and correlational, so the authors say it should be interpreted cautiously.
Sources
1 on file- Microbiology Australiadoi:10.1071/ma26055
Cite this record
v1Truvace Impact Record TRV-2026-3034, v1: “Beyond the tool: what a course-aligned AI assistant reveals about student learning in microbiology.” Truvace, 2026-11-18. /record/TRV-2026-3034 (accessed on the day you copy it). sha256 0d9195edb8701f4f…
Change history
1 versionEvery change made to this record since it was first published, and why. Nothing has changed since it went up.
Certified into the record
Check this record hasn’t been altered
How to check it yourself, without trusting this page
Download any version from /api/record/TRV-2026-3034 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.
ace