Fisk University and Tennessee State University are using the Intrusive Learning Community approach to test a theoretical model of factors that influence students to switch from STEM to non-STEM disciplines. Researchers propose to use the Performance Pyramid to develop measures that examine the effectiveness of two interventions - a combined section of college algebra and general biology that infuses math into the science curriculum and a learning community to complete mathematics projects related to biology facilitated by undergraduate peer mentors called Peer Partnership Learning (PPL) leaders. The objectives of the learning community are to promote interdisciplinary thinking and problem solving, decrease student mathematics anxiety, encourage students to pursue STEM coursework, and increase student self-efficacy and persistence in STEM. The PPL leaders in the learning community will form supportive networks and enhance learning and personal development of student participants. The proposed study is expected to provide important insights to guide STEM education and increase the pipeline of African American students who pursue STEM graduate education and careers.<br/><br/>Researchers will use a quasi-experimental non-equivalent control group design to test a multi-faceted hypothesis. The hypothesis states that students who participate in the PPL leaders-focused learning community will demonstrate greater knowledge and skills in their STEM courses; that the conceptual elements of the performance pyramid can be quantified; that PPLs will endorse more performance pyramid support systems; that students will evaluate the learning community positively; and that participants will show higher show higher STEM retention and persistence. Conversely, researchers postulate that the number of STEM students who switch to non-STEM disciplines will be reduced among participants in the small learning communities, and the level of math anxiety will significantly decrease upon completion of the integrated course. The researchers will develop the course materials for intrusive learning and the survey to measure the support system in the Performance Pyramid model. They will use a combination of multivariate analyses of variance, Poisson regressions, structural equation modeling, Pearson's correlations, and descriptive statistics to analyze the research data. They also intend to organize a faculty development workshop to encourage more STEM faculty to integrate learning communities in teaching.