Background/Question/Methods College students come to the freshman classroom with preconceptions relating to course material. Sometimes these ideas are alternate or misconceptions that interfere with their understanding or learning of new content. For four semesters, we used Diagnostic Question Clusters (DQCs) to define student reasoning about matter cycling and energy flow in a freshman biology class at Hampton University, a historically Black institution (HBCU) with predominantly African-American students. First DQCs were administered as pre-tests to diagnose understanding of the two concepts; next we focused on correcting problematic student thinking using active learning activities or homework followed by discussion; then DQCs on the same topics were administered as post-tests to determine whether learning had taken place. Student responses from the pre- and post-tests were coded using rubrics, developed for use with the DQCs, that categorized responses into five levels of reasoning ranging from informal to scientific thinking. 
Results/Conclusions The majority of student responses from both pre- and post-tests included a mixture of informal and scientific ideas, and very few student responses indicated purely scientific reasoning, which was the goal of instruction. Student reasoning improved after focused instruction, but not as well as we had hoped. Some lessons learned from this process include: (1) Do not make assumptions about what your students know or don’t know, because frequently you will be wrong; (2) Do not emphasize “covering the material” over discussing concepts in depth, because students need a lot of reinforcement of difficult or unfamiliar ideas; (3) Use precise language and insist that your students do the same, because many students think that being close to correct is good enough; (4) The best way to include formative assessment in your course is not to just add it in, but to first revise the course to focus on major concepts and then diagnose student understanding of those topics; (5) Diagnostic Question Clusters work very well to diagnose misunderstanding, but it is difficult to fit reinforcement directed at correcting misunderstandings into a general biology course if one has not first pared the course down to essential concepts.