Attending to Coherence Among Research Questions, Methods, and Claims in Coding Studies

We consider a kind of study common in mathematics education research: one that allocates qualitative data to categories in a theoretical or conceptual framework. These studies sometimes lack coherence among research questions, sampling and analysis methods, and claims, which can be attributed to tensions in how these aspects are framed. We ground our discussion in examples from five published studies, focusing on the methodological and reporting decisions that increase coherence: answering research questions from the same perspective they are asked (using a variance or a process lens), using (relative) frequencies properly to warrant claims, employing a coherent sampling strategy, and making appropriate generalizations. We argue that attending to coherence can increase the quality and contribution of coding studies.