Raw data contain errors that must be corrected before datasets are published or analyzed. In a census, this is done during data editing. Data editing is the process of detecting and correcting errors or logical inconsistencies. Data editing also helps to identify problems concerning missing or invalid responses. These problems are typically addressed using data imputation. Developing detailed computer assisted personal interviewing (CAPI) specifications is fundamental to preserving the original data and minimizing the amount of edited and imputed responses in a census dataset. This Select Topics in International Censuses (STIC) technical note provides guidance to National Statistical Offices (NSO) on how to develop specifications for an electronic questionnaire.