TECHNOLOGY FOR HEALTHCARE WORKFORCE FOR OLDER ADULTS: CHATGPT-ASSISTED QUALITATIVE ANALYSIS
2024

Using ChatGPT for Analyzing Qualitative Research in Elder Care

Sample size: 47 publication Evidence: moderate

Author Information

Author(s): Han Soojeong, Alexander Gregory

Primary Institution: Columbia University

Hypothesis

How does ChatGPT's qualitative analysis compare to human analysis in the context of elder care?

Conclusion

ChatGPT can analyze qualitative data much faster than humans but misses some important details.

Supporting Evidence

  • ChatGPT completed qualitative analysis in minutes, while humans took days.
  • ChatGPT provided clear definitions and explanations.
  • ChatGPT calculated convergence and divergence rates efficiently.
  • 14 out of 79 sub-themes were missing in ChatGPT analysis.

Takeaway

This study looked at how well ChatGPT can help analyze discussions about taking care of older people. It can do it really fast, but sometimes it misses important things.

Methodology

The study analyzed 10 roundtable discussions using ChatGPT and compared it to human analysis based on established qualitative research methods.

Potential Biases

ChatGPT's responses are based on its training data, which may not be current.

Limitations

ChatGPT missed 18% of sub-themes and may not capture all nuances from discussions.

Participant Demographics

47 experts in care of older adults participated.

Digital Object Identifier (DOI)

10.1093/geroni/igae098.4274

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication