Keystroke logging and qualitative analysis

Over the past few years keystroke logging has become a standard tool for data collection and analysis of writing processes. Keystroke logging software includes a number of automatic analyses and common issue in keystroke logging research is the extreme amount of data that is generated and how to make sense of it. Most studies to date use the automatic outputs from keystroke logging software and use statistical methodologies to analyze it. In this paper I will first give an overview of keystroke logging research to date, after which I present and discuss the applicability of a qualitative analytical approach. I will exemplify how the keystroke log files can be used qualitatively for close linguistic analyses. We will explore, in detail, revisions at the leading edge and how they can be understood as idea generating points in the writing process. Revisions at the leading edge are defined as being 'open' or 'closed' for semantic elaboration.Open revisions provide more than one alternative route while closed revisions only allow one option. The notion is applied to a 15-year-old writer composing in Swedish as L1 and English as FL. Results show that a substantial amount of the writer's revisions are open and that they occur on different syntactic levels. Results also show that language competence impacts on writers 'open' and 'closed' revisions; the stronger the language the more open revisions. I argue that revision at the leading edge indicate that processing of ideas and meaning occur constantly during text production, more or less during transcription and without substantial pausing, and that the dynamics of idea generation is affected by language competence. Knowing that language impacts on different levels of writing, including idea generation, provide input to writing education that better accounts for the multilingual classroom and learners' different strengths and areas of development.