Skip to main content Skip to navigation

Connecting Feedback to Choice: Understanding Educator Preferences in GenAI vs. Human-Created Lesson Plans in K-12 Education -- A Comparative Analysis

Project Overview

The document examines the integration of generative AI (GenAI) in K-12 education, with a specific focus on its application in lesson planning. A comparative analysis of AI-generated lesson plans—produced by customized GPT-4 and fine-tuned LLaMA-2-13b models—versus those created by human educators reveals a nuanced landscape. While human-designed plans are still favored, particularly in elementary education, AI-generated plans are gaining traction in high school settings. The research indicates that using GenAI as a collaborative tool can improve the efficiency of lesson planning without compromising educational quality. Overall, the findings underscore the potential of generative AI to support teachers in creating effective lesson plans, suggesting a future where AI serves as a valuable partner in the educational process.

Key Applications

Comparative analysis of lesson plans generated by AI models and human curriculum designers.

Context: K-12 education, particularly for mathematics lesson planning.

Implementation: An exploratory study with mathematics educators evaluating lesson plans from AI models and human designers.

Outcomes: Identified that AI-generated lesson plans can rival human-created plans in certain contexts, especially at the high school level.

Challenges: AI-generated plans often lack domain-specific knowledge, leading to potential inaccuracies and pedagogical misalignments.

Implementation Barriers

Technological Barrier

Generative AI tools often lack domain-specific knowledge, leading to inaccuracies in content and pedagogy.

Proposed Solutions: Fine-tuning AI models with domain-specific datasets and incorporating educator feedback into model training.

Perceptual Barrier

Educators exhibit a preference for human-authored plans due to perceived higher quality and engagement levels.

Proposed Solutions: Highlighting effective use cases of AI-generated content and demonstrating improvements in AI-generated plans.

Project Team

Shawon Sarkar

Researcher

Min Sun

Researcher

Alex Liu

Researcher

Zewei Tian

Researcher

Lief Esbenshade

Researcher

Jian He

Researcher

Zachary Zhang

Researcher

Contact Information

For information about the paper, please contact the authors.

Authors: Shawon Sarkar, Min Sun, Alex Liu, Zewei Tian, Lief Esbenshade, Jian He, Zachary Zhang

Source Publication: View Original PaperLink opens in a new window

Project Contact: Dr. Jianhua Yang

LLM Model Version: gpt-4o-mini-2024-07-18

Analysis Provider: Openai

Let us know you agree to cookies