Skip to main content

Randomising questions and answers

Reduce order bias by varying how questions, answers and groups are shown

Updated yesterday

Randomisation helps reduce order bias by changing the order in which respondents see questions, answers or groups of content. This makes your results more reliable, especially when testing concepts, creative or preferences.

Attest allows you to randomise answers, individual questions and groups of questions, depending on what best suits your research.


Randomising answer order

Answer randomisation shuffles answer options so each respondent sees them in a different order. This helps prevent respondents from favouring options at the top of a list.

By default, answer options are randomised for single choice, multiple choice and ranked questions. You will see a randomisation icon next to answer options when this is enabled.

When to turn answer randomisation off

If you are using a scale, such as a 1–5 scale or “very likely” to “very unlikely”, you should turn answer randomisation off. Keeping scales in a fixed order avoids confusing respondents. You can manage this setting in Answer settings.

Pinning answers

If answers are randomised, you can pin individual answer options to a fixed position. This is useful when you want options such as “None of the above” or “Don’t know” to always appear at the bottom.


Randomising subjects in grid questions

In grid questions, you can randomise both:

  • The order of the subjects (items being rated)

  • The order of the answer options

Subject randomisation is switched on by default. You can turn it off if you want subjects to appear in a specific order.

As with answers, you can pin individual subjects to a fixed position if needed.


Randomising question order

You can randomise the order in which questions or text cards are shown to respondents.

To do this, open the card list on the left-hand side, select Randomise, choose the cards you want to randomise, and confirm. Randomised cards are marked with a randomisation icon and can be previewed before launch.

Some cards cannot be randomised, such as qualifying questions or cards that use routing or piping. If a card cannot be randomised, you will see an explanation in the card list.


Randomising groups of questions

Groups allow you to randomise sets of questions together, while keeping the order within each group fixed.

This is particularly useful when:

  • Showing one creative followed by its follow-up questions

  • Comparing multiple concepts with the same question structure

To use groups, add questions and text cards to a group, then randomise between groups using the card list.

Limitations to be aware of

When using groups:

  • Questions cannot be moved into or out of a group after creation

  • Questions within a group cannot be randomised

  • Routing, piping and display logic cannot be used within or across groups


Check randomisation in preview and results

Use preview mode to test how randomisation behaves for different respondents.

In the results dashboard, randomised questions and answers are marked with a randomisation icon in the question details. This helps you understand how content was shown when interpreting results.

If you have any other questions about when and how to use the randomisation features, or how these features affect ongoing trackers, get in touch via the in-platform live chat!

Did this answer your question?