One of the leading issues restricting the automation of questionnaire instrument construction is the quality of metadata available describing question flow within an instrument. There has been ample research into the wording, ordering and presentation of questions. However, there has not been the same level of investigation into how questions link together in a logical manner.

Historically, paper-based forms have relied on the "skip statement" to control the sequencing of question modules for different respondents. Unfortunately the implementation of this approach in modern metadata standards and computer aided interview tools has not been entirely successful. By examining the overlaps between question sequencing logic and fundamental computer programming this paper examines the elimination of the skip-based methods at the metadata level. It also examines how this change in methodology will impact questionnaire designers.

The proposed solution is the use of structured conditional blocks and population-based state variables as a way of describing question flow. This approach, inspired by early techniques in computer programming, hopes to improve the reusability of instrument metadata and the predictability and efficiency of automated instrument creation – for both paper and online surveys.