It has been said many times that randomness is next to godliness in the world of experimental design. I would extend this thinking to the structured world of survey research as well. In this post we will look at two methods for incorporating randomness into your online survey design.
Cvent offers the researcher two forms of randomization. Both can be employed in a survey or assessment, but there are some caveats.
Within a question the categories offered to the respondent can be randomized. This is a perfect fit for long lists in a multiple response formatted question. Randomization of categories minimizes order bias. This form of bias relates to the greater likelihood of items that are in the beginning or ending of the list being selected over items in the middle of the list. If you are creating assessments then randomizing categories is also called for. Categories can be set for randomization while in the edit mode (see below).
The other option available is to randomize the question order on a page. A use case for this would be to randomize Likert type questions where multiple items form one or more scales. The researcher can employ question randomization on any or all of the pages in a survey, unless there is logic involved. If skip logic or other forms of advanced logic (e.g. masking) are being used then question randomization cannot be employed on the affected pages.
Randomization is a critical tool for survey design in consumer and B2B market research. It’s primary use is to mitigate for bias for which it is very effective.
Want to learn more about Cvent Web Survesy? Register for a demo and we'll find a time that's perfect for you!