In any research study, it's important to make sure your participants feel at ease and know what to expect from your study before they participate. Introducing your study well will lead to a better participant experience, and help you gather more in-depth and thoughtful responses from your participants.
For studies that are longer than a few questions, we recommend introducing your study in a similar way to how you would introduce a traditional (non-AI) moderated user interview, by explaining:
A brief intro to who you are, what research you are doing and what you plan to do with it.
Set clear expectations of the study. For example, if you're collecting voice responses as part of your study, let your participants know in advance so that they can prepare.
Here's an example of how we often introduce our own studies:
Hey there! 👋 My name is Tom, and I'm a researcher interested in learning more about your experience with [TOPIC]. To do this I have set up this AI-led interview in which you’ll be asked a series of questions about your experience with [TOPIC].
⏱️ The interview will take about 15 minutes to complete.
🗣️ You’ll need to record your voice to answer some of the questions question. Please make sure to answer each question asked as fully as possible.
💭 Remember there are no right or wrong answers. I work independently and I want to hear your honest thoughts, whether positive or negative.
🌐 Make sure you have a reliable internet connection to complete the study.
We make clear to participants that some questions in my study are moderated by AI
We'll clearly display to participants in the user interface that AI is being used to ask questions in the study. We do this with the "AI" badge:
We've done extensive testing on this experience to make sure participants generally feel comfortable with it, and have found that most participants prefer an AI-moderated experience to participating in a human-moderated session due to:
Flexibility: Participants can take the study when they want to, and don't have to complete the study at a specific time (which is often required for human-moderated user interviews)
Not being observed: Many participants reported feeling comfortable participating in AI-moderated studies than human-moderated studies, as they were not being observed by another human as they were thinking about and giving their answers.