The article examines artificial intelligence applications in qualitative focus group data analysis. An experiment involved three researchers: the first conducted manual analysis of focus group transcripts, the second used the Deep-Seek model, the third compared results. The empirical base includes materials from two focus groups with teenagers collected in Perm region in 2022. The theoretical basis was the participatory approach of childhood sociology with R. Olsen’s participation model operationalization. LLM advantages in data structuring and routine task automation were identified, however, significant limitations were found quote distortions, generation of non-existent information and one-sided analysis. It was concluded that LLM cannot replace humans in deep qualitative data analysis but are effective as primary processing tools. An algorithm for working with LLM for qualitative data analysis was developed.
Stored only when both papers are in the corpus — 2 citations exist upstream.