Beskrivning

Designing a question that clearly, quickly, and consistently communicates its intention is an extremely difficult task that requires great amounts of time, money, and expertise. Given that these resources are often in short supply, we designed an experiment to assess if generative AI can identify points of confusion in question wording in a time- and cost-efficient manner.

Join us to learn why it’s so hard to ask good questions and how tools like ChatGPT might be able to help us identify weaknesses in question wording.

Medverkande

Erica Metheney, Statistician, Political Science, University of Gothenburg

Lauren Yehle, Associate Researcher, Political Science, University of Gothenburg

Chuwei Chen, Data Analyst, Political Science, University of Gothenburg