A known methodological issue in survey research is structural response bias in case enumeration answers are presented in a fixed order. Respondents tend to select the first answers that are presented in an answer list. This occurs in self interviews such as internet surveys, but also in face to face surveys and telephone surveys. To miminize the amount of structural bias due to this undesirable response behaviour, answers may be presented in a random order.

There have been several attempts in the past to realize a random presentation order of answer categories in a Blaise questionnaire. Until recently, in Blaise this could only be done in a relatively complex way. The methods used in the old examples are not generic. They require modification for every new question. More generic solutions have been proposed but they require using the Blaise API, which is not available to everybody.

A new approach with Manipula in Blaise 4.8 makes it possible to realize this in a less elaborate way without using the Blaise API. This newly proposed method uses a Manipula setup and a generic include file for the Blaise data entry program. The Manipula procedures can be used in the data entry program relatively easy. For using the procedure a minimal amount of additional fields and instructions is required in the Blaise rules. This paper explains this generic approach to present answer actegories in random order. A practical example will be added in the appendix.