LLM Engineering and AI Agents (RSiII)

2026/2027

Учебный год

ENG

Обучение ведется на английском языке

5

Кредиты

Кто читает:

Кафедра информационных систем и технологий (Нижний Новгород) (Факультет информатики, математики и компьютерных наук (Нижний Новгород))

Статус:

Курс обязательный

Когда читается:

3-й курс, 3, 4 модуль

Course Syllabus

Abstract

This course develops practical and methodological competence in using Large Language Models (LLMs) to support qualitative data analysis. Students learn to integrate LLMs across several analytic traditions: thematic analysis (including the AI-supported frameworks GAITA, NITA, and CAAI), grounded theory, narrative analysis, and frame analysis. The course grounds all applied work in established qualitative methodology and in deliberate prompt design, so that the analysis stays transparent, reproducible, and defensible for reporting and peer-reviewed publication.

Learning Objectives

• ground LLM-assisted analysis in established qualitative methodology
• select and apply an appropriate approach (thematic, grounded theory, narrative, or frame) for a given research question and dataset
• design, document, and calibrate prompts so that the analysis can be reproduced and scrutinised
• critically evaluate the outputs of different models and tools, and report the use of AI in line with academic and ethical standards

Expected Learning Outcomes

• Explain what LLMs are and why their architecture suits qualitative methodology, language-centred inquiry
• Design, document, and calibrate prompts (including system prompts) for qualitative analytic goals
• Conduct LLM-assisted thematic coding (inductive and deductive) and build a transparent codebook for research
• Compare the behaviour of different models and AI-enabled CAQDAS tools, and assess reproducibility and reliability across them
• Apply at least three AI-supported thematic frameworks (GAITA, NITA, CAAI) and compare their assumptions and outputs
• Maintain the centrality of the researcher, critically appraising AI outputs rather than naturalizing them
• Apply grounded-theory coding (open, axial, selective) with LLM support, using constant comparison and memo-writing, and assess where the model aids versus limits theory-building
• Apply narrative analysis with LLMs, coding stories holistically and operationalizing theory as a codebook
• Apply frame analysis with LLMs, operationalizing a frame scheme as a codebook and distinguishing framing from emotional language

Course Contents

• Prompt engineering foundations for qualitative analysis
• Coding with LLMs: levels of abstraction, calibration, and reliability
• Three frameworks for AI-supported thematic analysis: GAITA, NITA, CAAI
• Grounded theory with LLMs
• Narrative analysis with LLMs
• Frame analysis with LLMs

Assessment Elements

• Лабораторные работы
• Экзамен

Interim Assessment

• 2026/2027 4th module
  0.35 * Лабораторные работы + 0.3 * Экзамен

Bibliography

Recommended Core Bibliography

• Readme first for a user's guide to qualitative methods, Richards, L., Morse, J.M., 2013

Recommended Additional Bibliography

• Kathy Charmaz. (n.d.). Advances in Qualitative Methods Conference Premises, Principles, and Practices in Qualitative Research: Revisiting the Foundations. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsbas&AN=edsbas.5A292B7C

Authors

• ULITIN BORIS IGOREVICH