Wondering uses generative AI to help you automate steps in your research process, such as generating good research studies and analyzing the responses you collect to identify findings.
To get the best results when using Wondering's AI to generate a research study or analyze your study responses, you can describe your research questions to guide Wondering's AI.
Where do I write my research questions?
If you're creating your Wondering study using the AI-powered study generator, you'll need to describe what you would like to learn with your study on the Brief step of the study builder:
If you're generating the AI analysis for your study, you can describe your research questions by clicking the Research goal button in the top right corner of the study results page:
How should I formulate my research questions?
Understand the difference: between research questions and interview questions
A major source of confusion in user research is mixing up research questions with interview questions (or survey questions). In short, research questions define what you want to learn, discover, or confirm in order to make informed, evidence-based decisions. They tend to be broader in scope and focus on the knowledge gap you are trying to fill. They should be specific, actionable, and feasible.
Interview questions on the other hand, are the actual questions you ask participants during a study. They seek to uncover the data needed to answer your research questions. Often need to be phrased carefully to avoid biased or unanswerable queries.
In Wondering, you should use your research questions as prompts for generating your studies and guiding your AI analysis — not interview questions. By inputting clear, well-defined research questions, the AI-powered study builder will design a study to help you find the answers you need, while the AI analysis will be better equipped to interpret your data.
Examples: research questions vs interview questions
Research question: “How do families with school-age children decide how to spend money on holidays?”
Possible interview question: “Walk me through your last holiday—from planning and budgeting right through to returning home.”
Research question: “How do our current customers select and view television programmes?”
Possible interview question: “Describe the last time you searched for something new to watch. Where did you look and what influenced your choice?”
Research question: “What incentives and information sources influence product decisions within our organisation?”
Possible interview question: “Who on your team usually signs off on new product features, and how do they decide what to prioritise?”
Notice how the interview questions prompt concrete examples or stories from participants, whereas research questions provide the overarching goal of the inquiry.
Tips for writing good research questions in Wondering
Here are a number of tips for how to write research questions in Wondering:
We've found the formats of research questions in this article work well for generating studies and AI analysis in Wondering, but you should feel free to explore prompts that aren't recommended in this guide.
Be clear, concise and direct
When writing your research questions, imagine Wondering as a capable research assistant with no prior context. Your questions should spell out precisely what you aim to learn. The clearer you are, the more precisely the AI-powered study builder and AI analysis can address your objectives.
Example: “How do first-time homebuyers perceive our online mortgage calculator, and which features do they value most?”
Avoid imprecise descriptions
If you have particular aims, frameworks, or areas of focus, spell them out in your research questions. Specificity helps Wondering create a study design tailored to your requirements and guides the AI analysis to deliver more relevant insights.
Vague language leads to guesswork in both generating your study and guiding the analysis. The more precise your research questions, the more accurate and valuable your insights will be.
Less precise ❌ | More precise ✅ |
Are our products easy to use? | What accessibility barriers do users with visual or hearing impairments encounter in our educational software, and how does this affect their overall satisfaction? |
Ask yourself if a colleague would understand what you want to learn based on your research questions
When writing your research questions, ask yourself whether a colleague with no context would know what you are trying to discover. If they find your question confusing, refine it until the intent is clear. Share your research questions with a teammate before finalising them. If they can envisage a suitable study design and the kind of analysis you want, you’re on the right track.
Don't be afraid to test and iterate
The best way to find out whether your research questions produce the studies and AI analysis you seek is to try them out. Generate a study using the AI-powered builder and see whether it meets your expectations. Once you have results, review the AI analysis that’s been generated based on your research questions. If you need more detail or discover new aspects to explore, refine your research questions and try again.
Examples of good research questions
Measure user satisfaction
Less effective ❌ | Better ✅ |
Do people like our latest product? | How satisfied are our existing customers with the latest product version, and which features drive their overall satisfaction or dissatisfaction? |
Assess product usability
Less effective ❌ | Better ✅ |
Is our product accessible? | What usability challenges and accessibility barriers do users with visual and hearing impairments experience when using our educational software, and how do these impact their willingness to continue using it? |
New product concept testing
Less effective ❌ | Better ✅ |
Would runners like this new idea? | To what extent are frequent runners on Strava interested in a ‘find a running mate’ feature, and how do their motivations and past efforts to find running buddies influence their openness to this concept? |
Evaluate homepage copywriting
Less effective ❌ | Better ✅ |
Is our product copy effective? | How do consumers respond to three different marketing copy variations (emotional, informative, and testimonial) for our new line of athletic wear, and which style most strongly influences purchase intent? |
Understand brand perception
Less effective ❌ | Better ✅ |
How do people see our brand? | What are consumers’ perceptions of our fashion brand compared with key competitors, and how do factors like brand awareness, loyalty, and recognition of competitor products shape their overall impressions? |