Two studies investigated how people interpret open-ended categorical questions. The observed findings of both studies show that people do in fact misinterpret category titles and that they do so in systematic ways. The results of Study 1 indicate that people were most likely to give a false positive when they interpret a category as including items that serve a particular goal. People were more likely to give correct (desired) responses when they thought in terms of varieties of items literally belonging to a category. Study 2 confirmed these findings. The use of supplemental instructions is recommended as a means to improve data quality in these questions.