TADR in ELT | Chapter 10: AI Transcription & Verbatim Cleaning: Fixing AI Hallucinations in Research

Discover how to leverage AI speech-to-text tools for qualitative research without compromising academic integrity! While tools like OpenAI Whisper and Otter.ai save hours of manual typing, AI outputs cannot be trusted blindly. Learn how to clean AI draft transcripts and fix critical errors before coding. 📌 Key Topics Covered: Average accuracy rates of AI speech-to-text tools (~85%) and why human verification is non-negotiable. Understanding and identifying AI "hallucinations" (fabricated speech) to protect research integrity. Step-by-step workflow for verbatim cleaning against original audio. Standardizing speaker labels and inserting consistent timestamps. 🎓 Course: Technology-Assisted Data Research (TADR) 📖 Chapter: 10 (Sub-CPMK: SUB-PBING-04 - Part 2)