Academic journal article Educational Technology & Society

Analyzing Pauses in Computer-Assisted EFL Writing-A Computer Keystroke-Log Perspective

Academic journal article Educational Technology & Society

Analyzing Pauses in Computer-Assisted EFL Writing-A Computer Keystroke-Log Perspective

Article excerpt

Introduction

Written text production involves complex cognitive processes that place heavy demands on the writers' working memory (Kellogg, 1996; Olive, Kellogg, & Piolat, 2008). When writers experience cognitive overload in their working memory, some processes would be suspended or even sacrificed to accommodate the immediate call for a specific process (DeKeyser, 2001; de Larios, Manchon, & Murphy, 2006) and pausing is a strategy that writers actively or passively adopt to free up attentional resources for processes of immediate priority. Therefore, pauses may serve as windows to writers' writing activities, such as "phrasing, memory such, decision, feedback, conceptual integration, and so forth" (de Beaugrande, 1984, p. 166). In addition, although writing consists of major processes as planning, translating (i.e., putting ideas into visible language) and revising (Hayes & Flower, 1980), these processes are not activated linearly but interact with one another in a recursive fashion, with one process dominating while other processes remaining dormant until being reactivated. While Olive and Kellogg (2002) suggest skilled writers can activate transcription concurrently with translating, planning and revising are mainly activated during pauses (Alves, Castro, & Olive, 2008; Olive, Alves, & Castro, 2009). Therefore, pausing may reveal possible problems and writing strategies behind such shifts of writing processes (Wengelin, 2006). As writing performance depends on how writing processes are organized during the composition process (Van den Bergh & Rijlaarsdam, 2007), pausing, being windows to writing activities and cognitive processes during text production, merits more research attention if the writing process is to be better understood.

Many pause studies were conducted in pen-and-paper settings and mostly adopted video-recording or thinkaloud protocols to probe into the writing process. While both methods have produced enlightening results for the comprehension of writing and think-aloud protocols have, in particular, advanced the knowledge of how pauses are filled qualitatively, these methods are limited to case studies due to practical reasons. In addition, there have been wide concerns over the intimidating nature of video-recording and the reactivity issues of thinking-aloud in writing research (for a review, see Yang, Hu, & Zhang, 2014).

With computer arising to be the leading writing medium for academic purpose, writing has witnessed much difference from that in pen-and-paper settings (Alves, Castro, Sousa, & Stromqvist, 2007; Olive & Kellogg, 2002). Meanwhile, with the advent of computer science and technology, a number of keystroke logging tools have been developed and greatly promoted written production research (Leijten & Van Waes, 2006). These computer programs "log and time stamp keystroke activities to reconstruct and describe the text production processes" (Leijten & Van Waes, 2013, p. 359), thus enabling writing research to examine the online writing process both multi-dimensionally and at higher accuracy levels. More importantly, these programs, by running in the background, minimize writers' anxiety of being observed. Informed by recent development in writing research and theory, these programs also provide rather sophisticated analyses of the writing process, thus more opportunities for refined writing research (Latif, 2009; Sullivan & Lindgren, 2006; Van Waes, Leijten, Wengelin, & Lindgren, 2012).

As different writing processes draw on the same working memory pool (Kellogg, 2001; McCutchen, 2000), the activation or deactivation of certain processes provides insights into the writers' management of their cognitive processes during the text production. With the rationale behind keystroke logging that "writing fluency and flow reveal traces of the underlying cognitive processes" (Leijten & Van Waes, 2013, p. 360), this study, using Inputlog6. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.