Most UX surveys are administered during the final Listen phase of the design cycle. However, there are many different types of surveys, which can be used effectively throughout the design cycle.
Designing a good survey is harder than it may seem. Ensure accurate and reliable data by writing questions that are appropriate for the method and worded to minimize bias.
Individuals often modify their behavior if they know they are being observed. That phenomenon became known as the Hawthorne effect or the observer bias. We can mitigate this effect by building rapport, designing natural tasks, and spending more time with study participants.
Confounding variables can affect the validity of data collected during research studies. It's important for researchers to know what they are and how to identify them.
The Net Promoter Score (NPS) can be gamed, and its definition loses information and precision by treating fairly dissimilar responses in the same way. It should be used together with other UX metrics rather than in isolation.
The SUS is a well-established 10-question survey administered at the end of a user test; it gives you a measure of the perceived usability of your product and enables you to compare it with others.
Surveys asking users to give feedback during or after an interaction should not interrupt the users' task and should be sent to the appropriate channel. They need to be short, easy to complete, and give the user the opportunity to provide details about their experience.
Response biases make it difficult to create good surveys. Follow these tips to counteract 10 of the major survey response biases and improve your survey data.
Well-designed questions related to age, gender, race, income and other demographic characteristics help UX researchers screen participants, recruit a diverse participant pool, and segment data. These questions are sensitive and should put research participants at ease.
To gain a holistic picture of your users, exchange data with the non-UX teams in your company who are collecting other forms of customer data, besides the user research you do yourself. You gain; they gain.
For most teams, approaching persona creation qualitatively is the right balance of effort vs. value, but very large or very small organizations might benefit from statistical or lightweight approaches, respectively.
Likert and semantic differential are instruments used to determine attitudes to products, services, and experiences, but depending on your situation, one may work better than the other.
Users' answers to survey questions are often biased and not the literal truth. Examples include acquiescence bias, social desirability bias, and recency bias. Knowing about response biases will help you interpret survey data with more validity for any design decisions based on the findings.
Through 4 rounds of revisions, we made a survey more specific and usable. Running pilot studies before conducting a full-scale survey ensures you’ll reach your research goal.
