UX Survey
A survey is one of the most commonly used quantitative research methods. It helps you to gather data by asking relevant questions to a large number of participants that represent your target audience.
Conducting a survey is very tempting as it is very quick and easy to implement but it has its own challenges too. You should be very careful while creating the survey questions as it is the most crucial part. If the questions are biased and not properly crafted, you may end up with misleading results.
When surveys are good to use
Here are a few examples where surveys can help you gather data:
- To gather feedback about the existing or new features.
- Collect quantitative data to validate your qualitative research findings since qualitative research can be done only with a limited set of people.
- To know the overall customer satisfaction.
- Gather feedback from the actual users while they are using the app. This is beneficial because people tend to forget things easily and they might give a different opinion when asked later.
However, surveys are not suitable for the below cases:
- When you want to study user behavior and why they have chosen a particular answer. For that, you will have to use any of the qualitative research methods.
- When you want to understand the usability problems of your product.
- When you want to understand the mental model of the user.
Survey benefits & limitations
Types of survey questions
Below are a few types of questions to get more responses:
- Multiple choice questions: These are the most used questions in a survey. A list of options is provided, from which the participants need to select one or more options. They are intuitive, easy to use, and produce easy-to-analyze data.
2. Rating questions: NPS question is a good example of rating scale questions. It can help to know the overall experience of an app or service.
3. Likert scale questions: These questions can help to know the feeling or opinions of the user. The question consists majorly of 5,7 and 9 scale options and the participants are expected to select the option that is most accurate to their response. It is important to value the numbers for these questions.
4. Picture choice questions: Using pictures in a survey makes it more interactive.
5. Demographic questions: Demographic questions can be single or a group of questions. You can know the age, gender, location, etc of a person. Singe choice options or dropdowns can be used here based on the type of question.
6. Open-ended: These questions are used to collect qualitative data to understand the user’s behavior and the way they are thinking. It takes a lot longer to analyze the qualitative data from a survey. A free text field is given with these questions for the participants to explain.
It is recommended to use open-ended questions at the end of the survey as there are chances that the participants might not be interested in writing long answers.
Tips to write good survey questions
Survey bias
People have different opinions and biases based on their past experiences, which affects the accuracy of the responses in the survey.
Below are a few biases that might affect the survey response:
Biases are very common and obvious in humans. You cannot completely eliminate them but should try to minimize it.
A quick guide to design a survey
Step 1: Define your goal
You should be very clear about the type of understanding you want to get at the end of the survey. Also, how will that information help you in achieving your project goal.
Step 2: Target Participants
Based on your survey goal you will have to decide on the target participants. And, the number of responses you want in a certain time frame.
Step 3: Write your questions
Write the survey questions keeping in mind the pointers that we discussed above. You can use the online survey tools available. Make sure that the questions for similar topics are grouped and logically sequenced.
You can add some incentive for the participants to get more responses and keep them motivated throughout the survey.
Step 4: Test with pilot users
After preparing the questions, test it with a small number of participants to know if the survey is understandable and good to launch.
Re-iterate the questions if there is any feedback and then test again.
Step 5: Launch the survey
Once your survey is finalized, you can roll it out to the real participants.
Step 6: Analyse the data
After all that effort, it is very exciting to see the response to the surveys. You can use the data to validate your initial hypothesis.
For open-ended questions, start by grouping similar types of responses.
Conclusion
Surveys are in itself not good or bad. Like any other tool, it also has its own challenges and limitations. It is true that surveys cannot be used to collect qualitative or behavioral data, but if used correctly it is very scalable as it can help you to validate new or existing features with a large number of participants.
