Increasing success rates in calculus is of critical importance for improving the pipeline to careers in STEM. The Arlington Undergraduate Research-based Achievement in STEM (AURAS) project, partially funded by the National Science Foundation, implemented the Arlington-Emerging Scholars Program (A-ESP) to address these pipeline issues. The A-ESP intervention involves engineering, mathematics, chemistry, and physics intended majors in an intensive supplemental workshop setting in which students encounter calculus tasks at a deeper more conceptual level. We compare the mathematical experiences of A-ESP students versus non-A-ESP students in the same lecture (n=93) in the context of their performance on departmental exams. In general, A-ESP students outperform non-A-ESP students on departmental exams, but in an effort to further improve curriculum, items in which A-ESP students drastically outperformed their counterparts or vice versa were back-mapped to the homework, labs, and A-ESP tasks. Findings suggest that on items where A-ESP students outperformed the non-A-ESP students the tasks students worked in the A-ESP problem sets tended to be more conceptual or abstract. However, on the items where non-A-ESP students outperformed the A-ESP students, the A-ESP tasks tended to resemble ordinary homework tasks. Thus, on typical primarily procedurally based calculus exams, the conceptual and abstract tasks appear to be boosting student performance and ultimately helping them successfully complete the course. The implications of this work will be discussed as well as ways to follow up on the findings.