Relevance in numbers: accounting for interactional principles when coding and quantifying conversational phenomena

How can interactionality be accounted for when coding and quantifying conversational actions? This article tackles this question by following three methodological steps commonly pursued in mixed method approaches that combine conversation analysis (CA) with quantification. Drawing on empirical examples from a four-year research project on the CA-based quantification of actions in conversational argumentation, it illustrates the practical challenges and solutions developed to address them. The first step, coding , concerns the identification and categorization of interactional phenomena. Five coding rationales are derived from sequential analyses of selected transcripts, showing how coding can be aligned with core interactional principles such as sequentiality , co-constructivity , methodologicality , pragmaticity , and multimodality , thereby producing analytically justified numerators . In the second step, quantification , the article emphasizes the importance of matching coding, research questions, and quantitative methods. Selected results of quantitative analyses are used to discuss which kinds of questions can – and should – be asked in CA-based quantitative studies, based on which denominators are demonstrably relevant to the participants themselves. The third step, interpretation , focuses on making sense of quantitative results. Using selected examples, the article shows how indispensable qualitative post-analyses are for validating, disambiguating, and refining quantitative findings. Overall, the article demonstrates that a crucial part of achieving codability and quantifiability in CA-based research rests on identifying the particular sets of relevancies that guide the organization of the domain at hand. Moreover, the implementation of iterative cycles of hypothesizing, analyzing, and validating is shown to be essential for any quantitative expansion of CA research.