AI analysis ≠ qualitative analysis

"AI analysis ≠ qualitative analysis" is one of most misunderstood things I teach about AI for UX research.

I'm not saying that AI analysis is *worse quality* than human qual analysis.

I'm saying that if you look under the hood, they're fundamentally different processes.

Human qualitative analysis is about understanding and interpreting data.

Generative AI analysis is not about sensemaking but statistics: navigating the text fragments that are most likely to co-occur across a large corpus of text.

Each approach has its own strengths and weaknesses. One is not inherently worse than the other. But we run into problems when we treat them as interchangeable, because we misunderstand what errors are most likely to occur.

📊 Leveraging research effectively means understanding your tools.