Automated assessment outputs must retain human monitoring and never serve as official grades or sole decision bases¶

Strategy · All strategies
Evidence · no claims cited

Description¶

An ethics and privacy requirement the article states for any future use of the dataset in automated assessment. The article specifies that "Any future use involving automated assessment must retain human monitoring" and that outputs from research models must not be treated as official grades or used as the sole basis for a decision affecting a student. This follows the article's separation of dataset research activities from official grading decisions for participating students.

Design Implications¶

Context¶

Requirements¶

• Human monitoring retained in any automated assessment use
• Recipients must avoid re-identification and public redistribution

Constraints¶

• Some scanned pages retain student names, student IDs, university names, institutional logos, or related identifiers within the page image

Target Learners¶

• higher education students whose examination scripts are included in assessment research

Target Learning Goals¶

• ethical automated grading research
• privacy-preserving use of educational records

Related Strategies¶

• Train teachers in prompt engineering, evaluating AI-generated responses, ethical and privacy considerations, and integrating AI outputs into pedagogical decision-making
• Deploy agentic AI in education with transparency about agentic decisions, human oversight and institutional responsibility
• Achieve high-precision automated constructive assessment through prompt design alone—explicit role assignment plus human-scored hint examples—without fine-tuning
• Be cautious with LLM-based TTX assessment and have a human instructor review automated outputs
• Address core ethical challenges—data privacy, algorithmic fairness, and transparency—when deploying AI-driven personalized learning

Examples¶

-

Key Sources¶

• Jahangir Alam SM, Md Khalid Syfullaha, Saad Ahmed, Munira Akter Mou, A K Z Rasel Rahman, A.K.M. Masudur Rahman, Mohammed Sowket Ali. (2026). Multimodal examination answer data with expert-designed Outcome-Based Education rubrics for criterion-level assessment. arXiv. https://arxiv.org/abs/2608.22346