An enhanced understanding of the nature of scientific knowledge-what counts as a scientific argument and how scientists justify their claims with evidence-has been central in Korean science instruction. However, despite its importance, scholars are generally concerned about the difficulty of both addressing and improving students' epistemic understanding, especially for students of a young age. This study investigated Korean middle school students' epistemic ideas about claim, data, evidence, and argument when they engage in reading both text-based and data-inscription arguments. Compared to previous studies, Korean middle school students show a sophisticated understanding of the role of claim and evidence. Yet, these students think that there is only a single way of interpreting data. When comparing students' ideas from text-based and data-inscription arguments, the majority of Korean students barely perceive text description as evidence and recognize only measured data as evidence. Keywords Ⅰ. IntroductionThe epistemic understanding of science-what counts as scientific knowledge and how scientists come to know and warrant their ideashas been recognized as critical to promote science literacy (Archive, 2012;Duschl, 2008). However, despite its importance, scholars are generally concerned about the difficulty of both addressing and improving students' epistemic understanding, especially for students of a young age. Questionnaires developed to assess students' understanding of the professional nature of science (NOS) could be too abstract and thus difficult for the level of elementary and middle school students to understand (Sandoval, 2005). Consequently, these questions elicit only ambiguous ideas, often expressed with very short, unclear answers, making it difficult to make a reliable and useful interpretation. Often students simply answered, "I have no idea." In addition, since these questionnaires hardly situate students in relevant contexts, it could be the case that students rarely reflect and think deeply about these questions and may consider them to be tests evaluating content knowledge. If we want to understand students' ideas and thinking about evidence, it would be best to ask questions, such as "what do you think evidence is?", in situations where they look for or evaluate evidence during their science activities and tasks.Situating students in relevant tasks also allows them to use their epistemic resources, drawn from both their everyday experience and experience with science (Moje, 2007(Moje, , 2010McNeill, 2011).Alternatively, students' epistemic ideas are increasingly addressed in the context of argumentation practice, as this would offer students richer opportunities to reflect on epistemic ideas, such as what can be counted as claim, data, evidence, and argument. In Korea, while there have been growing studies regarding scientific argumentation, little research has been concerned with the connection between student argumentation practice and the epistemic understanding of science (for revie...