We’ve all heard about ChatGPT by now, but can we use it to analyze user research interviews? This article will explain what it is, and how to do it.
What is a ChatGPT analysis of a research interview?
A ChatGPT analysis of a research interview involves inputting your qualitative research interview transcription into ChatGPT and asking it to return a summary of the interview, as well as a list of the important takeaways.
ChatGPT can use natural language processing techniques to analyze and extract insights from your interviews, saving you hours of work of qualitative analysis. Depending on what you’re looking for, ChatGPT’s analysis can include pre-processing, keyword extraction, sentiment analysis, topic modelling, text analysis and summarization, comparison, thematic analysis, and clustering.
![A woman in a casual outfit taking pictures with a disposable camera.](https://ux-everything.com/wp-content/uploads/2023/02/chatGPT_interviews-main.jpg)
What do you need for a ChatGPT analysis of a research interview?
TIME
MATERIALS
How do you perform a ChatGPT analysis of a research interview?
![Several polaroid photos, ribbons and crafting supplies scattered on a table.](https://ux-everything.com/wp-content/uploads/2023/02/chatGPT_interviews-second.jpg)
Step 1: Export your transcription
Export the transcription of your research interview into a file that you can easily edit and copy the text within.
Step 2: Clean your transcription
Clean and format the text data from the transcription of the interview, making sure to remove any sensitive or identifying information that you don’t want uploaded to ChatGPT. You should also check for spelling mistakes here too.
Step 3: Add instructions for ChatGPT to analyze your transcription
Enter the instructions for ChatGPT. You may have to experiment a little so get creative, but you should be able to get some good results.
Here are some examples of instructions for ChatGPT to analyze your research interviews:
- Take this interview transcript and list out the main takeaways from every question asked:
- Take this interview transcription and create a list of all of the questions and provide a summary for each response:
- Take this interview transcript, and create a list of all of the questions asked by (interviewer name), and add a summary of (participant’s name)’s responses below each question. The format should be question, response, question, response:
Step 4: Add your transcription to ChatGPT in small batches
It seems that the free version of ChatGPT has limits on how much text you can input. I’ve found I can paste about 6-8 pages of transcription at a time.
Below the instructions you’ve entered, paste a few pages of your research interview transcription into ChatGPT and submit it.
Step 5: Consolidate the results
If you had to break the transcription up into smaller chunks due to ChatGPT’s size limitations, then paste all of the results into a separate text editor. Give the the whole thing a read and make sure it’s coherent and accurately reflects your research interview.
Step 6: Repeat for any other interviews
If you have multiple interviews, repeat the process for each one and make sure you consolidate all of the results into one master summary and list of important insights.
Tips for a great ChatGPT Analysis of a Research Interview
![A woman sitting on a bed writing in her journal and staring off into the distance.](https://ux-everything.com/wp-content/uploads/2023/02/chatGPT_interviews-third.jpg)