Can AI Supercharge Your Thesis? Large Language Models in Social Science Research 4219-SH0050(KC)-OG
This project-based workshop introduces Bachelor’s and Master’s students to the practical, ethical, and critical use of Large Language Models (LLMs) in social science research. During the course, participants will have a chance to discuss their individual projects and learn how to use AI in their own research endeavours. We will learn to leverage modern AI tools, such as Gemini and NotebookLM, to streamline key stages of research—from project design to the final write-up of the thesis.
This will include using LLMs to:
- assist in coding in qualitative research,
- summarize key concepts for writing abstracts and summaries,
- search for relevant quotes and page numbers in primary sources,
- transcribe interviews,
- create questionnaires,
- perform translations,
- verify hypotheses, and edit academic texts.
Each session will be highly interactive and centered around individual student research. Two or three students will be selected each time to present their current projects (short oral presentations). The group will debate and analyze how specific AI tools (like Gemini or NotebookLM) can best support the presented projects.
The course will also include discussions about ethics, plagiarism, and appropriate referencing for using AI as an editing assistant in preparing manuscripts. We will intensively discuss the limitations of LLMs, including the risk of algorithmic bias in data analysis, and learn how to distinguish between facts and AI "hallucinations" (generating false information), as well as how to critically verify information provided by LLMs.
Type of course
Course coordinators
Learning outcomes
Knowledge:
• The student possesses basic knowledge of the functioning and structure of Large Language Models (LLMs), such as Gemini and NotebookLM.
• The student understands how AI tools can streamline the key stages of the research process in social sciences (from project design to thesis editing).
• The student is aware of the ethical and legal issues related to using AI in academic work, including the problems of plagiarism, appropriate referencing, and the risk of algorithmic bias.
• The student knows the limitations of LLM models, including the phenomenon of "hallucinations" (generating false information) and the need for critical data verification.
Skills:
• The student is able to effectively use Gemini and NotebookLM for research purposes (e.g., transcription, summarization, qualitative coding, questionnaire creation).
• The student is able to develop precise and effective prompts (prompt engineering) to obtain desired results from the LLM.
• The student is able to critically verify and evaluate information generated by AI, distinguishing facts from errors or hallucinations.
• The student is able to integrate AI tools into their individual research project (Bachelor's/Master's), increasing its efficiency.
Competences:
• The student is responsible for an ethical and critical approach to utilizing AI technology in scientific research.
• The student develops the ability to continuously improve and adapt to rapidly changing digital tools in the academic environment.
• The course develops discussion skills on complex issues related to technology, ethics, and bias in data analysis.
Assessment criteria
- Attendance at a minimum of 80% of sessions
- Presentation of one's own research project (30%)
- Active participation in exercises and discussions during class (35%)
- Final Assignment (35%)
To pass the course, the student must complete all four assessment components.
0-55% = 2
56-69% =3
70-83% =4
84-97%=5
98-100%=5!
Bibliography
Igor Grossmann et al. "AI and the transformation of social science research". Science 380, 1108-1109 (2023).DOI:10.1126/science.adi1778
Rossi, L., Harrison, K., Shklovski, I. (2024). “The Problems of LLM-generated Data in Social Science Research.” Sociologica, 18(2), 145–168. https://doi.org/10.6092/issn.1971-8853/19576
John Berryman, Albert Ziegler “Prompt Engineering for LLMs: The Art and Science of Building Large Language Model-Based Applications.” O'Reilly Media 2024.
Additional information
Additional information (registration calendar, class conductors, localization and schedules of classes), might be available in the USOSweb system: