In the biosciences, quantitative skills are an essential graduate learning outcome. Efforts to evidence student attainment at the whole of degree programme level are rare and making sense of such data is complex. We draw on assessment theories from Sadler (evaluative expertise) and Boud (sustainable assessment) to interpret final-year bioscience students' responses to an assessment task comprised of quantitative reasoning questions across 10 mathematical and statistical topics. The question guiding the study was: "do final year science students graduate knowing the quantitative skills that they have, and knowing the quantitative skills that they do not have?" Confidence indicators for the 10 topics gathered students' perceptions of their quantitative skills. Students were assigned to one of four categories: high performance-high confidence; low performance-low confidence; high performance-low confidence; or low performance-high confidence--with those in the first two categories demonstrating evaluative expertise. Results showed the majority of students effectively evaluated their quantitative skills as low performance-low confidence. We argue that the application of evaluative expertise to make sense of this graduate learning outcome can further the debate on how assuring graduate learning outcomes can enhance student learning.