This article is a written version of the Presidential Address I gave at the annual meeting of the National Council on Measurement in Education (NCME) in April 2017. It is a call to NCME members (and others who read this, of course) to rebalance their focus so that classroom assessments are seen as being at least as important as large-scale assessments for education (in fact, in my view, they are more important). The article reviews research literature about the effects of classroom assessment to establish its importance for education. Then, the roles of large-scale assessment are reviewed, and, in particular, it is noted how these can have negative results when the large-scale assessments are not well aligned with sound curriculum and instructional and assessment practices grounded in theories of learning. In the next two sections (a) the idea of a learning progression is described as a way to facilitate the coherence between classroom and large-scale assessment and (b) the idea of a 'roadmap' is described, being the assessment components of the learning progression. This is followed by a description of an example of such a roadmap, developed for the Assessing Data Modeling and Statistical Reasoning project using the BEAR Assessment System (BAS). Finally, a concluding discussion reviews the ways that the coherence between large-scale and classroom assessments can be achieved using the BAS, and hence make measurement more important for education. Author abstract