Response: why instructional communication scholars should use accurate, applicable, and inclusive methodologies

I must start by commending all forum authors for their contributions here. These five essays should provoke discussion that far exceeds their brevity, and although diverse in topic, they nevertheless cohere and intersect with each other in important ways. It is these intersections I will explore in this response, identifying themes that appear in most or all of the essays. From this, I will consider topics implied by these essays—one might say, that lie in the shadows between them—that might further elaborate the important claims in this forum. Three of the five essays primarily focus on how we use research methods, with each of those three emphasizing a specific aspect of that how. Brammer et al. focus on practical application, considering how researchers and practitioners might leverage research methods to engage in assessment. Johnson highlights accuracy and completeness, underscoring how data collection procedures and handling of missingness may bias obtained results. And Kelly et al. center inclusion, elaborating how research practices have systematically spotlighted some groups while excluding others. Although each essay places emphasis on a particular theme, all three themes cut across all three essays. For example, Brammer et al. weigh assessment biases (accuracy) that might arise from ignoring the experiences of students with disabilities (inclusion), and Johnson considers the reduction of research burden for participants (inclusion) while continually focusing on the link between communication and learning (application). This weaving of themes makes sense because, following Kelly et al., inclusive methodologies lead to more accurate/complete results, which is crucial for effective instructional practice. The remaining two essays focus on whatmethods instructional communication scholars might use. Specifically, each of these essays advances a methodological approach that instructional communication scholars have ignored. Of note, these two also repeat the themes considered in the three how essays. Goodboy and his colleagues discuss mixture modeling as an approach that can reveal groups hidden in the data, obscured by other approaches (such as the general linear model). In some sense, this recasts the argument of Kelly et al. in quantitative language, providing a tool whereby quantitative scholars can investigate groups whose experiences might be rendered invisible beneath statistics that aggregate across the majority. Without question, a clearer understanding of student groups should foster practical application of practices that enhance learning (Brammer et al.).