This poster illustrates how survey para-data can be utilized to improve/correct the design of an online questionnaire, specifically duration items that were implemented in GGS-II, wave 2 in Czechia as part of a country-specific module on first reproductive experiences. Inappropriate splitting of duration questions leads to an increased number of backward clicks and a high number of edits to previously provided responses as well as higher response times. After a correction was applied, the occurrence of these events/activities was in line with other items in the same questionnaire. The poster concludes that we should continuously monitor para-data to improve questionnaire design and data quality.