This study examines whether virtual internships serve as a viable alternative to in-person work-based learning opportunities, and, if so, for whom and in which specific dimensions. Drawing on a large-scale survey ( N =11,009) and administrative data collected from 17 U.S. postsecondary institutions during the 2020-2021 academic year, the research compares the experiences of historically marginalized students in virtual and in-person internships to those of their non-marginalized counterparts across access, program quality, and outcomes. Findings from a Latent Profile Analysis indicate that while most virtual interns (86%) participate in programs of comparable quality to in-person formats, this parity breaks down at the extremes, with virtual interns overrepresented in under-resourced experiences and underrepresented in high-quality, long-term placements. Regression analyses further reveal that racial minorities and students with moderate work commitments experience smaller gaps in either participation, learning quality, or outcomes in virtual internships. However, rural students, other racial minority groups, women, and those with extensive work commitments face equivalent or amplified barriers in either access to or quality of virtual learning experiences. These asymmetrical distribution patterns highlight important nuances concerning technology-fueled internships as universal equity solutions, offering implications for stakeholders designing inclusive work-based learning ecosystems. • This study examines virtual internships as alternatives to in-person opportunities across access, quality, and outcomes. • 86 % of virtual interns participate in educationally structured programs at rates comparable to in-person counterparts. • However, 14 % are overrepresented in lower-quality roles or underrepresented in long-term, high-quality programs. • Racial minorities and moderately working students experience smaller gaps in virtual internships. • Rural, women, and extensively working students face sustained or amplified barriers, requiring targeted equity efforts.

The use of Bayesian Knowledge Tracing (BKT) models in predicting student learning and mastery, especially in mathematics, is a well-established and proven approach in learning analytics. In this work, we report on our analysis examining the generalizability of BKT models across academic years attributed to "detector rot." We compare the generalizability of Knowledge Training (KT) models by comparing model performance in predicting student knowledge within the academic year and across academic years. Models were trained on data from two popular open-source curricula available through Open Educational Resources. We observed that the models generally were highly performant in predicting student learning within an academic year, whereas certain academic years were more generalizable than other academic years. We posit that the Knowledge Tracing models are relatively stable in terms of performance across academic years yet can still be susceptible to systemic changes and underlying learner behavior. As indicated by the evidence in this paper, we posit that learning platforms leveraging KT models need to be mindful of systemic changes or drastic changes in certain user demographics.