Como conduzir entrevistas com usuários sem liderar ou enviesar os participantes?

Pro tip: Create a quick mockup of dummy results, then work backwards. If needed, you can also use the mockup to verify stakeholder buy-in of your methodology and results template. Example A: Say your goal is to select the best of two competing designs for a game controller. Your dummy mockup results might show comparison graphs of (a) task time to completion, (b) user errors, and (c) user-reported satisfaction. Seeing this dummy data will naturally bring up a discussion about the user’s task and how errors are measured. Depending on your team and workflow, it may also be appropriate to verify the target audience and agree on the volume of testing required to give high confidence in the results.

"How much did you love the new screen layout?" When conducting user interviews, it can often seem harmless to ask a question in this manner, but you may be unknowingly steering the user in your preferred direction. What if the user was confused or did not like the new screen layout? By asking how much they "loved" it we are inherently steering them to have a positive reaction regardless of what they actually think. We want to point the user to the topic of interest we are looking to explore without applying any emotional element within our question. A better way to ask the above question could be: "Was it easy or difficult to navigate through the new screen layout?"

Try to avoid yes/no questions. Asking "did you think [x] was easy/hard" only gives the user 2 options to respond. But asking "how was the experience of [x]"? allows them to put into their own words what they were feeling - frustrated, annoyed, quick, seamless. Instead of "Was [x] confusing," ask: "Was there anything that seemed confusing to you? Which part and why?" This gives more context than just if it was confusing or not. Another question you may ask: "I noticed [observation] - can you explain what you were thinking or feeling there?" This gives users a chance to explain why they hesitated or clicked on something unexpected. ^These questions get into the why of your users behaviors and feelings, not just the what.

Once you’ve perfected your wording and you’re conducting your research, a huge bias source is anticipation. For example, if your first 3 participants respond identically to a given question, it’s human nature to anticipate that your next participant will respond the same. You might even feel tempted to cut their answer short and fill in the blank for them. At the very least, your body language and facial tics will likely bias them in a certain direction. However, if you treat that next person with the same open curiosity as the first, you might be surprised to learn they have a very different answer. When you remain in a state of open curiosity with *every* participant, your results will be cleaner. Turns out this a great life skill, too!

In my experience, I find packaging results with a visual artifact and a Loom to be helpful. While it’s common to present work in slides, it’s important that your results tell a clear story and is “easy on the eyes”. As a designer, I like to prioritize visuals over copy as that tends to keep folks more engaged. Additionally, a video collaboration tool like Loom is great to use. Recording a short video of your presentation ahead of time, or you can use it to give a full walkthrough encourages your audience to give rapid feedback, and reactions. The important thing is to give your audience members flexible options to consume your results.