AI for Supporting Analysis of Qualitative Data

What will AI mean for the future of social research? Claims about the opportunities and risks abound. Meanwhile, there is evidence that government and funding bodies are becoming increasingly confident that this will mark a step change in the value, scale and replicability of social research. Are they ‘hallucinating’ (or worse) and, if so, with what consequences? Critical social researcher interrogation of these claims, the affordances, constraints and risks of AI is essential.

In this one-day course, Prof Les Carr (Web Science Institute, University of Southampton and Visiting Professor at CenSoF) will support researchers in critical exploration of generative AI (in particular large language models and chatbots) and in considering potential uses in participants’ own research. The session will start with a short introduction to how LLMs work and then we will open up to a hands-on-session where you explore your own research questions with some trial ‘data’. 

This course outlines the use of AI tools for the analysis of text data, which may have been generated through qualitative data collection (e.g. through interviews, focus groups etc) or through quantitative data collection, such as open-ended question(s) in surveys. 

The course covers:

• Use of Generative AI Chatbots (e.g. ChatGPT, Co-Pilot etc)

• Large Language Models (LLMs)

• Ethical and Responsible AI Issues

• Applying Chatbots (e.g. ChatGPT, Co-Pilot etc) to qualitative data for analysis (this can contain also text data from quantitative data collection methods)

• Prompt Engineering (i.e. the process of refining instructions and input prompts to guide LLMs, including tips on how to write prompts)

• Critical Evaluation of Research-Supported AI

Learning Outcomes:

By the end of the course participants will:

• Be confident users of Microsoft CoPilot or similar AI platform
• Understand the capabilities and limitations of LLM products
• Have experience in applying a Chatbot with qualitative data / text data
• Be able to critically analyse the costs/benefits/risks/opportunities of deploying AI in their research

Schedule:

This one day course will run between 10:00 - 17:00 with one hour for lunch. It will consist of a plenary morning workshop (teaching and discussion) and a practical individual session in the afternoon where each participant individually experiments with using CoPilot/ChatGPT on some of their research data (or relevant data of interest) with support and feedback from the course tutor.

• Introduction to the course
• Explain how generative AI and LLM models work
• Talk through the motivations for using AI and LLMs
• Research vignettes/case studies
• Discussion
• Hands on introduction to Copilot with simple worked examples.
• Break for lunch
• Open session where trainees individually experiment with their research questions and research data on Copilot, supported by trainers as necessary.
• Report back
• Planning for future work

Payment using the Online Store can only be completed via Visa and Mastercard Credit/Debit Card or PayPal.  AMEX is not accepted
If you have not previously created an account for the Online Store, you will need to create an account to make a booking