Both STEM and non-STEM teachers require specialized AI features for educational platforms, with important differences in assessment approaches, resource development needs, and student monitoring tools based on their subject domains.

Objective: This study aimed to identify and compare the AI-enhanced features needed by STEM and non-STEM teachers when using block-based programming (BBP) platforms, focusing on three key areas: assessment, course development and resource expansion, and student monitoring.

Methods: The researchers conducted semi-structured interviews with eight K-12 teachers - four from STEM disciplines (computer science, mathematics) and four from non-STEM disciplines (art, English as a Second Language, music, dance). The participants were selected based on their experience incorporating coding activities into their teaching. The interviews lasted 40-60 minutes and explored teachers' approaches to assessment, resource expansion, and student monitoring, along with their experiences using educational platforms and desires for AI integration. Thematic analysis was used to analyze the interview transcripts.

Key Findings:

• While STEM teachers typically use structured formative assessments like quizzes and tests, non-STEM teachers favor more qualitative and creative assessment methods like portfolios and project visualizations.
• Both groups emphasized the need for AI features that provide integrity and plagiarism checks, customized rubrics, and detailed feedback in assessments.
• Non-STEM teachers specifically requested AI tools that preserve student work originality and authenticity, and support creative assignments with simulation capabilities.
• For resource development, both groups wanted AI tools to assist with curriculum updates, tutoring libraries, and generative AI features, but non-STEM teachers were particularly interested in supporting creative endeavors like art simulations.
• In student monitoring, both groups prioritized desktop control, daily tracking, behavior monitoring, and distraction prevention tools.
• Teachers highlighted the importance of integration and adaptability with existing Learning Management Systems (LMS) and single sign-on (SSO) functionality.
• Non-STEM teachers emphasized the need for more individualized and accessible tools for English Language Learners and tools that can evaluate creative assignments.

Implications: The study identifies specific AI-enhanced features needed by K-12 teachers across various disciplines, providing a foundation for developing more effective and personalized educational platforms. The findings suggest that educational platforms need to incorporate different AI features based on the domain-specific needs of teachers, rather than using a one-size-fits-all approach. Educational technology developers should consider the different assessment strategies, resource needs, and monitoring approaches required by STEM versus non-STEM educators.

Limitations: The study has a relatively small sample size of eight teachers, which may limit the generalizability of the findings. The research focuses on K-12 education in North and South Carolina, so the results may not be applicable to other geographic regions or education systems. Additionally, the study relies on self-reported needs and preferences, which may not fully capture all aspects of how teachers would use AI-enhanced features in practice.

Future Directions: Future research could expand the sample size and geographic diversity to validate these findings across different educational contexts. Researchers could also develop and test prototypes of AI-enhanced educational platforms that incorporate the suggested features to evaluate their effectiveness in real classroom settings. Additionally, exploring specific AI tools and features for different age groups within K-12 education would provide more targeted insights for educational technology development.

Title and Authors: "Comparative Analysis of STEM and non-STEM Teachers' Needs for Integrating AI into Educational Environments" by Bahare Riahi and Veronica Cateté.

Published On: March 2025

Published By: This is a preprint posted on ResearchGate with the DOI: 10.13140/RG.2.2.15444.92809.