Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

From Pilots to Practices: A Scoping Review of GenAI-Enabled Personalization in Computer Science Education

Iman Reihanian, Yunfei Hou, Qingquan Sun

TL;DR

The paper addresses the challenge of delivering scalable, personalized CS education using GenAI. It adopts a scoping-review approach to map mechanisms, signals, and governance considerations across 32 studies (2023–2025), identifying core patterns such as explanation-first tutoring, graduated hint ladders, artifact grounding, and tests/rubrics alignment. It provides a design-focused taxonomy, analyzes effectiveness signals and risks, and offers an implementation roadmap for departments emphasizing exploration-first adoption with human-in-the-loop oversight. The work informs practical policy, tooling, and curriculum decisions to preserve productive struggle, maintain integrity, and achieve durable learning in GenAI-enabled CS education.

Abstract

Generative AI enables personalized computer science education at scale, yet questions remain about whether such personalization supports or undermines learning. This scoping review synthesizes 32 studies (2023-2025) purposively sampled from 259 records to map personalization mechanisms and effectiveness signals in higher-education computer science contexts. We identify five application domains: intelligent tutoring, personalized materials, formative feedback, AI-augmented assessment, and code review, and analyze how design choices shape learning outcomes. Designs incorporating explanation-first guidance, solution withholding, graduated hint ladders, and artifact grounding (student code, tests, and rubrics) consistently show more positive learning processes than unconstrained chat interfaces. Successful implementations share four patterns: context-aware tutoring anchored in student artifacts, multi-level hint structures requiring reflection, composition with traditional CS infrastructure (autograders and rubrics), and human-in-the-loop quality assurance. We propose an exploration-first adoption framework emphasizing piloting, instrumentation, learning-preserving defaults, and evidence-based scaling. Recurrent risks include academic integrity, privacy, bias and equity, and over-reliance, and we pair these with operational mitigation. The evidence supports generative AI as a mechanism for precision scaffolding when embedded in audit-ready workflows that preserve productive struggle while scaling personalized support.

From Pilots to Practices: A Scoping Review of GenAI-Enabled Personalization in Computer Science Education

TL;DR

The paper addresses the challenge of delivering scalable, personalized CS education using GenAI. It adopts a scoping-review approach to map mechanisms, signals, and governance considerations across 32 studies (2023–2025), identifying core patterns such as explanation-first tutoring, graduated hint ladders, artifact grounding, and tests/rubrics alignment. It provides a design-focused taxonomy, analyzes effectiveness signals and risks, and offers an implementation roadmap for departments emphasizing exploration-first adoption with human-in-the-loop oversight. The work informs practical policy, tooling, and curriculum decisions to preserve productive struggle, maintain integrity, and achieve durable learning in GenAI-enabled CS education.

Abstract

Generative AI enables personalized computer science education at scale, yet questions remain about whether such personalization supports or undermines learning. This scoping review synthesizes 32 studies (2023-2025) purposively sampled from 259 records to map personalization mechanisms and effectiveness signals in higher-education computer science contexts. We identify five application domains: intelligent tutoring, personalized materials, formative feedback, AI-augmented assessment, and code review, and analyze how design choices shape learning outcomes. Designs incorporating explanation-first guidance, solution withholding, graduated hint ladders, and artifact grounding (student code, tests, and rubrics) consistently show more positive learning processes than unconstrained chat interfaces. Successful implementations share four patterns: context-aware tutoring anchored in student artifacts, multi-level hint structures requiring reflection, composition with traditional CS infrastructure (autograders and rubrics), and human-in-the-loop quality assurance. We propose an exploration-first adoption framework emphasizing piloting, instrumentation, learning-preserving defaults, and evidence-based scaling. Recurrent risks include academic integrity, privacy, bias and equity, and over-reliance, and we pair these with operational mitigation. The evidence supports generative AI as a mechanism for precision scaffolding when embedded in audit-ready workflows that preserve productive struggle while scaling personalized support.
Paper Structure (40 sections, 3 figures, 6 tables)

This paper contains 40 sections, 3 figures, 6 tables.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. From ITSs to LLMs
  4. Clarifying Terms
  5. Affordances Across CS Subdomains
  6. Pre-GenAI Baselines
  7. GenAI-Enabled Patterns in CS Education
  8. Methods: Scoping Approach
  9. Eligibility and Selection
  10. Charting and Synthesis
  11. Results
  12. Corpus Characteristics
  13. Application Areas and Mechanisms
  14. Measures and Constructs
  15. Descriptive Outcome Signals
  16. ...and 25 more sections

Figures (3)

  • Figure S1: PRISMA-style flow for the scoping review. Counts reflect purposive sampling for mechanism-rich, representative coverage.
  • Figure S2: Distribution of studies by application area in the included corpus ($N~{=}~32$). Colors distinguish application types for visual clarity.
  • Figure S3: Conceptual relationships between design mechanisms, implementation conditions, and outcome patterns. Blue boxes denote design mechanisms; yellow boxes denote implementation conditions; green boxes denote patterns associated with positive learning outcomes; red boxes denote risk and negative outcome patterns. Solid arrows indicate patterns associated with positive outcomes; dashed arrows indicate patterns associated with risks and mixed outcomes.