Writing
Analytics Toolkit

WAT (Writing Analytics Toolkit) was developed by researchers at Arizona State University’s Learning Engineering Institute (LEI) and the Science of Learning and Educational Technology (SoLET) Lab. Our interdisciplinary team brings expertise in:

• Writing instruction
• Linguistics
• Natural language processing
• Learning sciences
• Educational technology

The tool was created in collaboration with educators to support revision-based writing instruction and to provide actionable, linguistically grounded feedback for students, teachers, and researchers. WAT is freely available and built to serve instructional, research, and assessment needs across diverse learning environments.

WAT was supported by the Institute of Education Sciences (IES), U.S. Department of Education, through Grant #R305A180261.

The content and tools developed reflect the work and views of the research team and do not necessarily represent the views of IES or the U.S. Department of Education.

When contacting technical support, provide:

Essential Information:

• System Configuration: Operating system, version, RAM.
• Debug Logs: Export and attach relevant logs.
• Reproduction Steps: Detailed steps to replicate the issue.
• Corpus Information: Type of content, file formats, corpus size.
• Error Messages: Exact error text or screenshots.
• Expected vs. Actual Results: Clear description of what should happen vs. what occurred.

This information helps technical support diagnose issues quickly and provide effective solutions.

Reach us via the LEI website.

WAT Researcher is built using the Electron framework with Java and Python dependencies packaged as integrated resources, eliminating the need for manual runtime installation by users.

System Requirements:

macOS:

• macOS 64-bit architecture (M-series chips required)
• Minimum 8GB RAM (works for smaller corpus analysis)
• Recommended 16GB+ RAM for optimal performance with larger corpora
• 3.5GB disk space after application unpacking

Windows:

• Windows 10 and later versions
• 64-bit architecture
• Minimum 8GB RAM (works for smaller corpus analysis)
• Recommended 16GB+ RAM for optimal performance with larger corpora
• 3.5GB disk space after application unpacking

Important Installation Notes:

• Allow permissions when macOS or Windows prompts for application access
• As a newly independently developed application, macOS/Windows security systems may initially show trust warnings — grant necessary permissions for proper functionality
• All Java and Python dependencies are pre-packaged within the application — no manual installation required