Innovations in Data Analysis: Specialists from the Republican Center for the Study of Public Opinion “Ijtimoiy Fikr” at a Master Class of the Grushin Conference

Staff members of the Republican Center for the Study of Public Opinion “Ijtimoiy Fikr” took part in the “Future of the Research Industry” track within the framework of the Grushin Conference, one of the largest sociological forums.

Amid the rapid digitalization of research processes, the Republican Center for the Study of Public Opinion “Ijtimoiy Fikr” continues to actively introduce advanced technologies into the practice of public opinion research. An important stage in professional development was the participation of Timur Muratov, Head of the IT and Information Security Department, and Azim Nabiev, department specialist, in an online master class devoted to a systematic approach to the use of artificial intelligence in qualitative research.

From a “Naive” Approach to an Architectural Pipeline

The main topic of the master class, conducted by well-known sociologist and expert in qualitative methods Sergey Protsenko, was the transformation of work with transcripts of focus groups and in-depth interviews through the use of large language models (LLMs).

During the training, the Center’s specialists studied the difference between two approaches to the use of AI:

• “Naive” approach: one-off prompts to a neural network, which often lead to loss of context, model hallucinations, and the inability to reproduce results.
• Architectural approach (Pipeline): building a clear sequence of operations in which the analytical process is divided into controlled stages — from extracting verbatim quotations to cross-analysis and the synthesis of final conclusions.

“Human in the Loop”: Safety and Accuracy

Special attention during the master class was given to the Human-in-the-loop principle. Participants discussed how AI does not replace the researcher, but rather acts as a powerful tool for automating mechanical operations (60–70% of routine work), while leaving critically important functions to humans: interpretation of meanings, data verification, and quality control at every stage.

For the IT specialists of “Ijtimoiy Fikr,” the most relevant issues were:

1. Data traceability: the ability to track the path from a final conclusion in a report to a specific respondent quotation in the transcript.
2. Scalability: applying refined algorithms (prompts) to analyze large volumes of qualitative data without loss of quality.
3. Security: working in closed environments and using local neural networks to ensure full confidentiality of respondents’ data.

The Future of the Research Industry

Participation in the “Future of the Research Industry” track allowed specialists of the Republican Center for the Study of Public Opinion “Ijtimoiy Fikr” to become acquainted with new tools for structuring data (Markdown markup, automated tables) and methods for minimizing typical AI errors, such as the loss of irony or subtext in respondents’ speech.