ABSTRACT
This chapter discusses how computer-mediated communication (CMC) challenges methodological assumptions in sociolinguistics and outlines data sampling criteria in the framework of computer-mediated discourse analysis. It introduces two distinctions that impact how we approach language online: viewing CMC as "text" or "place" and collecting data "on screen" or through contact with users. The chapter also discusses issues related to the modes and environments being sampled, multimodality, social identities and participation roles, units and sequences of online data, and research ethics. It summarizes a range of issues related to online data collection, based on an inclusive view of sociolinguistics that encompasses variationist, interactional, and discourse-oriented approaches to language in society. "Screen-based" data are produced by participants and collected online by the researcher, while "user-based" data are prompted by researcher activities and produced through their contact with CMC users. A limitation to screen-based data may seem the norm in language-focused CMC studies, but this is neither self-evident nor uncontested within the discipline.
