critical-thinking
Filtering by topic critical-thinking(2)Clear all filters
- PaperarXiv — AI in Education (cs.CY)15 Jul 2026
When Rubrics Change: Cross-Rubric Generalization for Critical Thinking Essay Scoring
Nischal Ashok Kumar, Payu Wittawatolarn, Sana Kang, Marisa C. Peczuh et al.
Automated essay scoring research extends to cross-rubric generalization, where models trained on essays labeled under one rubric are evaluated on essays scored with different rubrics. Using a Large Language Model fine-tuning framework with rubric-agnostic intermediate 'traits', the study achieves up to 5.0% improvement in macro F1 over baselines. Results show that trait-based structure and controlled supervision enhance generalization to unseen rubrics.
Original abstract
Automated essay scoring (AES) research has largely focused on cross-prompt generalization, where essays from unseen prompts are scored while the scoring criteria are typically held constant. In practice, however, educators may revise or even introduce new rubrics in their scoring task, to evaluate different aspects of essays. We study cross-rubric generalization: training on essays labeled under one set of rubrics and evaluating on previously unseen rubrics, which target different aspects of the essay. We use a Large Language Model (LLM) fine-tuning framework with two components: rubric-agnostic intermediate representations, called traits, and target-essay supervision under seen rubrics during training. On an AES dataset augmented with multiple rubric-defined labels of student critical thinking skills, we find that traits improve macro F1 by 5.0% over a baseline without traits in the hardest setting, where both target rubrics and target essays are unseen during training. We further find that increasing target-essay supervision improves performance, with our best fine-tuned open-source Llama-based model outperforming GPT-5-mini prompting by 2.1% macro F1 and trailing GPT-5 by 1.9%. These results show that trait-based intermediate structure and controlled supervision improve generalization to unseen rubrics.
- NewsEdSurge17 Jun 2026
Before Schools Write the AI Policy, Should They Have a Conversation First?
Only one in three students report having a school-wide AI policy, yet many use AI for assignments and studying. Aleta Margolis argues that schools should prioritize open conversations with students over top-down AI policies, aiming to co-create guidelines that address both the promise and pitfalls of AI in education. The article cites polls showing teachers' concerns about AI harming critical thinking and the need for responsible use instruction.
Original abstract
<p>Only about one in three students say their school has a school-wide AI policy, yet students across the country are already using AI to complete assignments, study for exams, and make decisions about their daily lives. Aleta Margolis, founder and president of the Center for Inspired Teaching, has spent three decades coaching teachers to ask better questions and listen more carefully to students. She argues in <a href="https://www.edsurge.com/news/2026-06-03-what-to-do-about-ai-begin-by-talking-about-it"><u>a recent </u><strong><u>EdSurge</u></strong><u> piece</u></a> that the policy conversation cannot come first: the human conversation has to.</p><h2><strong>What the Classroom Conversation Looks Like Right Now</strong></h2><p>Margolis describes a growing dynamic in schools where teachers are using AI to generate assignments and students are using AI to complete them, raising a deeper question about what teachers and students are actually there to do together. She says co-creating AI guidelines with students, rather than handing down a list of rules, puts everyone on the same team and draws on knowledge that young people already have.</p><h2><strong>When the Policy Gap Comes Home</strong></h2><p>A new <a href="https://www.npr.org/2026/06/05/nx-s1-5779757/school-ai-education-students-teachers-poll-critical-thinking"><u>NPR and Ipsos poll</u></a> found that 54 percent of teachers worry AI is hurting students’ critical thinking skills, and nearly three in four believe its impact will be larger than the internet or computers. Sarah McKibben, editor-in-chief of <strong>EdSurge</strong> and the parent of two middle schoolers, is watching that tension play out at her own kitchen table. She sees both the genuine promise of AI as a learning tool and the corners students are cutting when no one has taught them what responsible use actually looks like.</p><h2><strong>Stories Mentioned in This Episode</strong></h2><p><a href="https://www.edsurge.com/news/2026-06-03-what-to-do