In Attest you have the option to randomise cards (text and questions), groups of cards, and/or answers, to suit the research you’re carrying out.
Randomising question cards
If you’re testing creatives, concepts or competitors (to name just a few!) you can remove order bias in your survey by randomising the order in which certain questions are presented.
To randomise questions, open the card list and select Randomise. Select the questions you want to randomise, and then select Done. You’ll see a new blue dot icon next to the cards that will be randomised, and you'll also see which questions are randomised in the Preview. You can edit the questions to be randomised as much as you like by returning to the card list and clicking Edit.
If cards in the card list turn red as you select them, that’s because they can’t be randomised. You can’t randomise qualifying questions, for instance, or cards containing routing or piping. You’ll see errors at the bottom of the card list explaining which of your cards can’t be randomised and suggesting ways to fix the error.
Randomising groups of questions
If you want to randomise one set of questions with another set (the questions within the group will not be randomised with each other), you can do this by creating question groups. This is especially useful if you’re displaying one creative followed by a set of follow-up questions, and then a second creative and set of questions (and so on).
Click Question and select Group. Add the static list of questions and text cards to the group. Then create a second group, and so on for as many groups as you require.
Groups are randomised in the same way as individual questions, using the card list on the left hand side. You can choose to randomise any combination of individual questions and question groups.
There are some limitations of the randomising groups feature, please note:
You cannot move questions or text cards into or out of groups, nor can you re-order questions and text cards within a group.
You can randomise between groups and individual (ungrouped) questions to your heart's content, but you cannot also randomise questions within the groups.
You cannot pipe within, into or out of a group.
You cannot create routes within groups.
Spotting randomisation in results
You can see which questions and answers are randomised in the results of your survey (or any survey shared with you). Randomised questions are displayed with a randomisation icon on the bottom of the card.
Randomising answer and question order minimises order bias. Be aware that if your research tracks metrics over time, introducing randomisation to questions and answers that haven't previously been randomised may have a small impact on results as data quality will be improved.
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!