Aligning keystrokes with cognitive processes in writing

Keystroke logging provides a detailed record of the process of writing as it unfolds in time. However, by themselves, the measures it provides—pauses, bursts and revisions—do not provide an unambiguous interpretation of the cognitive processes involved, particularly when the measures are aggregated across texts and writers. We have described this as a problem of alignment (Baaijen, Galbraith & de Glopper, 2012; see also Wengelin, Frid, Johansson & Johansson, this volume, pp. 30–49). How does one map keystroke units of analysis to specific components of the writing process? Pauses may reflect not just various different levels of planning and reflection, but also re-reading and text production during revision. Bursts of language may reflect an initial formulation of thought or an attempt to improve previously formulated text. Revisions may reflect semi-automatic correction of errors or a systematic attempt to modify content. In this chapter we will illustrate this problem of alignment with the help of Hayes’ (2009) model of text production. We will then describe procedures that we have used in order to increase the alignment between keystroke log measures and cognitive models of writing. Finally, we will demonstrate that when these measures are combined into more overarching measures they show systematic relationships with text quality and the development of understanding through writing.