Review
Aug 2026
LLMs for Survey Text Analysis - A Performance Comparison Between Humans and GPT-5 on Inductive Content Analysis
Overall, the findings suggest that LLMs can approximate human coding in this case-specific setting, particularly at the coding level, and may serve as a scalable support tool for inductive qualitative analysis.
Leonardo Bergmann, Renata Gheorghiu, Ana Gvritishvili et al.
· 0 citations