The purposes of this study were to test for uniform differential item functioning (DIF) between male and female students on end-of-semester class evaluations in an introductory university science course and to demonstrate DIF detection methods that take into account the complex sampling mechanism that often underlies data used in such assessments. Analyses were conducted using the multiple indicator multiple cause (MIMIC) model, which has been shown to be a flexible and effective tool for DIF detection. Because the data were collected in a hierarchical framework, with students nested in course section, a multilevel modeling approach was used. Results highlight certain aspects of the course that suggest differences in item response patterns between males and females while controlling for overall course satisfaction, as well as demonstrate the differences in DIF detection when the analytic strategy matches the data structure. (Contains 2 tables and 3 figures.)