Models All the Way Down

This chapter delves into the complexities of Quantitative Ethnography (QE), focusing on the often-confusing topic of segmentation. The authors, experienced QE researchers, identify the initial formulation of segmentation in the book Quantitative Ethnography as a source of confusion and aim to rectify this by proposing a new concept: the discourse model. They begin by unpacking the original concept of segmentation, highlighting its problems, and then introduce the discourse model as a more transparent and practical approach. The chapter also discusses the Initial Letter Convention, which distinguishes between evidence and inference, and suggests removing it to reduce confusion. The authors then focus on constructing a QE data file, emphasizing the importance of defining lines and ensuring they represent a similar amount of information. They also discuss content models, or codes, and their role in interpreting the content of each line in the data. The chapter concludes by proposing a new way to conceptualize QE models, consisting of four models: structural, content, discourse, and semantic. This approach aims to provide a clearer and more theoretically-grounded way of representing an analysis, making it easier to understand and implement.