Table of Contents
Guidance and tips for pre-testing an online survey
Never invite participants to a survey that you have not tested first. But how should you test it and what should your testers look out for?
Before sending hundreds or thousands of invitation emails, test the questionnaire first. Send the survey link to a few colleagues or test participants.
- Clone the survey so that you always have a backup.
- Reset the survey after testing to delete all test results.
- Allow multiple participation during testing so that testers can open and complete the questionnaire more than once.
- Set a password so that unauthorised people do not accidentally see the unfinished questionnaire.
- Add extra questions or comment fields to receive feedback immediately in real time.
- Set the survey to “not anonymous” during testing so that you receive more test data, such as the time spent on each page.
Even before launching the survey, you can see simulated results under “Analysis”. This lets you check in advance whether you have used the right question types.
Online survey pre-test: what your testers should check
- Does the survey display correctly on all devices?
- Do all dynamics, visibility rules and skips work?
- Can the survey be interrupted and resumed?
- How does the survey look in different browsers? Very old browsers may not display certain questions, such as drag-and-drop ranking questions.
- Does the invitation email arrive correctly?
- Is the invitation link displayed correctly in the email?
Walk through the entire participant journey: invitation, click the link, enter a password if needed, respond, interrupt, resume, respond, finish.
Monitoring during the live fieldwork period
You can still optimise the survey after its actual launch: our survey tool offers extensive options for checking data and results while the survey is running.
Our survey tool gives you all responses in real time, allowing you to follow everything immediately and intervene where necessary. For example, if an important question is often skipped, make it mandatory. If many respondents abandon the survey at a question, revise it.
- Time spent allows you to identify implausible responses quickly in non-anonymous surveys.
- The sample view shows response tendencies visually. They can be seen even more clearly after export to Excel.
- Closed surveys can effectively prevent multiple participation.
- Filters can help identify incorrect responses immediately, provided you have included control questions.
