Welcome to ASMCUE in Seattle, Washington! Registered attendees are able to download the mobile app onto their phone, create a log-in, and make a personalized schedule and chat with attendees. Questions? Email [email protected]
Log in to add sessions to your schedule and sync them to your phone or calendar.
Background Science communication and information literacy are essential graduate competencies in STEM, yet opportunities to develop these skills through authentic assessment vary considerably (Brownell et al., 2013). Developed with Wikimedia Australia, a Wikipedia-based group assessment requires microbiology students to compare a Wikipedia article and an AI-generated article on the same topic for accuracy, organisation, verifiability, depth, and suitability for a general audience before producing a comparative report and reflection. As generative AI has become increasingly embedded in undergraduate learning environments (Black & Tomlinson, 2025), this assessment has provided an unexpected opportunity to examine how students perceive their skill development and the use of Wikipedia and AI-generated articles as information sources across consecutive cohorts. Research question What are students' perceptions of skill development and of Wikipedia and AI-generated articles as information sources across three consecutive cohorts following a Wikipedia-based science communication assessment? Methods An anonymous post-assessment survey was administered across three consecutive cohorts (2024–2026; n = 113) of final-year microbiology students. Closed-ended items captured perceptions of article accuracy, readability, future use intentions, and learning outcome achievement; open-ended items explored perceived skills developed. Quantitative data were analysed descriptively and qualitative responses underwent thematic analysis. Results Across all cohorts, students most frequently identified critical thinking and source evaluation as skills developed through the assessment (~44%). Students consistently rated AI-generated articles as more readable than Wikipedia; however, confidence in AI accuracy declined after the first cohort and remained low. Despite this, intended AI use increased (31%→42%→45%), while intended Wikipedia use declined (60%→42%→40%), revealing a persistent adoption paradox. Conclusion Students consistently reported perceived gains in critical evaluation skills, while data revealed a persistent tension between AI readability and trustworthiness. This transferable assessment offers a practical approach to fostering science communication and information literacy through critical evaluation of public-facing information sources.
