As a contribution to the discussion on the assessment of informal inferential reasoning (IIR) and the transition from this to formal inferential reasoning (FIR), we present a review of research on how these two types of inferential reasoning have been conceptualized and assessed. Based on our review, we discuss the need to redefine the conceptions of IIR and FIR in order to create an integrated description of inferential reasoning that includes not only ideas of IIR and FIR, but also the whole activity of argumentation, which involves the production of both statistical and contextual reasons. Current descriptions of IIR and FIR list the facts that might be brought from data analysis to the process of inferential reasoning. The approach we propose considers how the facts, both statistical and contextual, can be used as arguments, leading to assessments of students' inferential reasoning focusing on articulating the statistical and contextual reasons students present to support an inference.