EUVIMEDEUVIMED entdecken
Scoping Review

A scoping review of early detection of at-risk programming students using AI-driven predictive analytics

Mashite Tshidi

Multidisciplinary Reviews · 22.05.2026 · 9 (12) · 2026563

DOI: 10.31893/multirev.2026563 ↗

Kurz gefasst

Scoping Review in der Zeitschrift Multidisciplinary Reviews, online erschienen am 22.05.2026. Verfasst von Mashite Tshidi. Die Arbeit nennt 8 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introductory programming is a foundational component of computer science education, yet it remains associated with high rates of failure, disengagement, and attrition. These challenges are commonly attributed to the cognitive and metacognitive demands of learning to program, compounded by uneven prior exposure among increasingly diverse student cohorts. In response, artificial intelligence (AI)-driven predictive analytics has been widely explored as a means of identifying students at risk of academic difficulty early in introductory programming courses. However, research in this area is fragmented across disciplines and varies substantially in data sources, modelling approaches, definitions of “at-risk,” and links to pedagogical practice. This scoping review systematically maps and synthesises peer-reviewed research on AI-driven predictive analytics for the early detection of at-risk students in introductory programming. Guided by the Arksey and O’Malley framework and reported in accordance with PRISMA-ScR guidelines, the review analysed 37 studies published between 2015 and 2025. The synthesis examines the learning data and predictive features employed, the modelling and evaluation techniques applied, how risk is conceptualised and operationalised, the extent of alignment with pedagogical interventions, and the treatment of contextual, ethical, and equity-related considerations. The findings indicate that early detection in introductory programming is technically feasible using a wide range of behavioural, assessment, and programming-process data. However, predictive capability has advanced more rapidly than conceptual coherence, pedagogical integration, and ethical grounding. By consolidating this fragmented landscape, the review provides a structured foundation for the development of context-aware, pedagogically meaningful, and equitable predictive analytics in introductory programming.

Zitieren

APA

Tshidi, M. (2026). A scoping review of early detection of at-risk programming students using AI-driven predictive analytics. Multidisciplinary Reviews, 9(12), 2026563. https://doi.org/10.31893/multirev.2026563

Vancouver

Tshidi M. A scoping review of early detection of at-risk programming students using AI-driven predictive analytics. Multidisciplinary Reviews. 2026;9(12):2026563. doi:10.31893/multirev.2026563

BibTeX
@article{Tshidi2026,
  author = {Tshidi, Mashite},
  title = {A scoping review of early detection of at-risk programming students using AI-driven predictive analytics},
  journal = {Multidisciplinary Reviews},
  year = {2026},
  volume = {9},
  number = {12},
  pages = {2026563},
  doi = {10.31893/multirev.2026563}
}
Kostenlos starten →