Sign up to follow
Sign up to save
Sign up to like
Sign up to respond
Sign up to follow
Sign up to follow
- Suggest a goal"{{search.query}}"
{{search_results.questions.label}}
-
Follow
{{search_results.people.label}}
-
A little context goes a long way...
-
Only you can see your suggestions
Vicki Tan
Product Designer at Headspace
Product designer. Currently 🙏🏻 @Headspace. Previously @Lyft + @Google.
6 action items
Setup user testing for unbiased feedback
-
-
Clarify the goals of your user research, also known as the research question.
What are you trying to understand about your product, a specific feature, or your users, and why?
-
Create an interview script that directly addresses that question.
It sounds simple, but relying on your script (which has been reviewed and agreed upon by a diverse set of stakeholders) can help remove bias.
-
Keep simple rules in mind when writing the script:
a. Ask open-ended questions that begin with "why", "what" and "how". This helps draw out information and personal stories. b. Conversely, avoid yes/no questions. They aren't great conversation starters and don't help you understand motivation. c. Avoid questions that you think you already know the answer to. This can help with confirmation bias later on (when we remember information that confirms our beliefs). d. Write out probing, follow-up questions. This helps clarify anything that could be ambiguous or misinterpreted. This also helps us avoid an anchoring bias, when we rely too heavily on the first piece of information we get.... See more
-
Whenever possible, the interviewer should not be the designer/PM on the project.
These teammates are inherently biased and invested. than someone else who didn't work on the feature. If this is unavoidable, you may consider telling a white lie to let users know that you didn't work on the feature. This helps users feel more comfortable being candid in their feedback.... See more
-
Enlist a dedicated note-taker, taking notes transcript style.
Highlight things that stick out as wins or red flags, but save the interpretation for post-session debriefs. This can help prevent a whole load of cognitive biases, such as a clustering illusion (overestimating importance of small clusters within a bigger dataset) and the framing effect (reacting to information differently based on how it's presented).... See more
-
Make sure you're mostly listening and the user is mostly talking.
Like a first date, right?
Be intentional about your process and take the time to prepare. Create clear research goals, write an interview script, and watch out for common pitfalls.
{{ answer_likes.includes(95) ? 6 : "5" }}{{ answer_favs.includes(95) ? "Saved" : "Save" }}Share -