Confidence-Aware Automated Assessment of Student-Drawn Scientific Models
Researchers have developed a confidence-aware automated assessment system for student-drawn scientific models. This system utilizes a Vision Transformer (ViT) with parameter-efficient adaptation to analyze drawings aligned with Next Generation Science Standards (NGSS). By deriving confidence scores from predictive distributions, the system can automatically score high-confidence responses and flag uncertain cases for human review, offering a practical balance between automation and accuracy in educational settings. AI
IMPACT This approach could enhance the efficiency and reliability of educational assessments by automating the evaluation of complex student work.