Do responses gained using category rating accurately reflect respondents' true evaluations of an item? "True" in this sense means that they have a real opinion about the issue, rather than being compelled by the survey to speculate an opinion, and that the strength of that opinion is faithfully captured. This article describes some common issues that suggest that it should not be simply assumed that a response gained using category rating reflects a true evaluation. That assumption requires an experiment to have been carefully designed and interpreted, and examples are shown where this is not the case. The article offers recommendations for good practice.