The category-partition method and classification-tree method both help construct test cases from specifications. To achieve this end, a set of categories and the associated choices (also known as classifications and the associated classes) have to be identified. However, the identification process is often done in an ad hoc manner. We have conducted a case study to examine the common mistakes made by software testers during this process. The result of our study will facilitate researchers and practitioners in the development of systematic identification methods.