一句话：访谈七名有执行功能困难的工程学生，揭示 LMS 支持不足并提出 EF–SRL–LMS 对齐框架与实践清单。
执行功能（EF）是工程教育与软件工程实践的基础，但学习管理系统（LMS）对有 EF 困难学生的支持尚不明确。本文访谈七名自报存在 EF 困难的工程学生（含神经多样与神经典型），以自我调节学习（SRL）与 EF 为视角做主题分析。结果显示：指令不清导致反复重读与外部检索，学生偏好简短示例而非密集文本；求助优先发生在 LMS 之外的同伴；日历与清单有用但使用不稳定；邮件缓慢；对 AI 即时反馈与社区网站依赖日增。学生认为 LMS 对组织必要，却难以提供及时低成本的澄清支持，因而产生平台外补偿行为与跨系统切换。作者提出 EF–SRL–LMS 对齐框架与实践清单，主张在平台内加入个性化、情境感知的自动化支持层以降低 EF 负荷。
Executive functions (EF)—including planning/organization, task initiation, working-memory management, cognitive flexibility, and help-seeking regulation—are foundational to engineering education and core software engineering practices. Yet little is known about how learning management systems (LMSs) support students who experience EF challenges in day-to-day coursework. We conducted an exploratory qualitative study using semi-structured interviews with seven engineering students both neurodivergent (ND) and neurotypical—who self-reported EF difficulties, and analyzed the data thematically through a self-regulated learning (SRL) and EF lens. Participants reported recurring EF-linked difficulties: (i) unclear or context-poor instructions prompting repeated re-reading and searches for external examples; (ii) a strong preference for brief visual/worked examples over dense text; (iii) peer-first help-seeking outside the LMS and limited use of discussion boards perceived as inconvenient; (iv) calendars valued and checklists helpful when used consistently (but usage was uneven); (v) email as the primary yet slow channel, with unease about formal messaging; (vi) growing reliance on AI for instant feedback and community sites (e.g., Stack Overflow) for coding questions. Overall, students regarded LMS tools as necessary for organization but insufficient for timely, low-effort clarification and support, leading to compensatory practices beyond the platform and increased cross-system switching. We synthesize an EF–SRL–LMS alignment that locates where EF load accumulates and derive a practice-oriented checklist. We argue for a more personalized, automated layer of in-platform support—adaptive, context-aware, and designed to reduce switching between multiple tools and platforms, to better sustain SRL across forethought, performance, and reflection while lowering EF load.
@inproceedings{paperbot11993,
title = {Why Students Leave the LMS: Executive Function Demands and Self-Regulated Learning in Software Engineering Courses},
author = {Khawla Shnaikat and Ann Barcomb and Emily Marasco},
booktitle = {ICSE},
year = {2026},
doi = {10.1145/3786580.3786967}
}