Truth Tips; where we share quick advice on how to create meaningful and commercially valuable truth for our clients.
Humanising Numbers With AI Tips for AI led open-end analysis
Open-ended questions in surveys help delve deeper and uncover what lies behind closed-ended responses, adding colour and texture to our analysis. Often, time and cost constraints mean they're overlooked or not analysed – until now! At Truth, we've been experimenting with AI tools to analyse open-ended questions to effectively blend quantitative and qualitative methodologies. The result? More human data points and happier clients!
Hot off the heels of analysing thousands of data entries from gaming and soft drink datasets, Victor Reynoso, Associate Truth Director shares learnings and tips from using AI tools for content, linguistic, and semiotic analysis of open-end responses.
Pick the right tool – Codeit, for example, has data protection built-in and is GDPR compliant, allowing us to efficiently analyse thousands of verbatims without risking privacy and data protection.
A human sandwich - start with human consultant input and end with human analysis but use AI to speed things up in the middle.
Look through the cupboard – don’t let data go to waste! Most questionnaires have open-ended questions, make the most of it!
Share your workings and AI processes with the team and clients - it reduces nervousness around the automation.
Use real-time analysis in brainstorms and workshops – AI can help speed up analysis and make the process more collaborative.
Paint a picture - Share quotes and vignettes to humanise your data. Not every data point needs numbers to tell your story.
Think outside the box – AI coding tools can be utilised with qualitative methodologies as well. If there is text, it can be coded and analysed!
Put the work in and use frameworks – AI on its own is not going to help you structure your data. Use semiotic principles, and strategy frameworks to help come up with more insightful and actionable insights.
To find out more about Truth and how we can help email hello@truth.ms
Comments