Productivity / UK / Healthcare
Qualitative data analysis using AI
"Using the latest generative AI algorithm, the Ipsos team generated transcriptions of group recordings, using Focus Group Data (FGD), to uncover, and synthesise, the insights generated by group discussions."
Business issue
The client wanted to improve productivity by leveraging artificial intelligence to summarise data and insights generated through focus group discussions.
Our solution
Using the latest generative AI algorithm (OpenAI’s speech-to-text Whisper AI), the Ipsos team generated transcriptions of group recordings using Focus Group Data (FGD). Following this, they evaluated how AI (GPT4) performs in its ability to summarise and synthesise the data and uncover the insights generated by the group discussions.
As a result, this study generated a variety of prompts which analysed the data in a variety of contexts using two prompt engineers: a researcher involved in project and one that was not. Lastly, the team compared the results of AI-enabled analysis to the original report conducted entirely by a human researcher.
For more information on this case study, get in touch here.
Impact
This study demonstrated the need for detailed background, context and clearly defined targets, as well as adjustments for moderation, in order to help AI get more out of the transcripts. Whilst this produced insightful results, more testing is needed before it can be replicated on more complex projects.
Ultimately, AI is a powerful tool for qualitative researchers, and massive productivity can be gained as it augments capabilities and speeds up the research process. Although, there is a need for this to be merged with research frameworks, and to be led by a researcher, to better understand KPIs.
