Self-editing behavior in real-time computer-based writing: Effects of prior print exposure, writing prompt, and pre-response time

Effective communication requires that speakers or writers monitor and respond to the reactions of their listeners or readers, as well as to their own speech or writing. Speakers and writers largely rely on their experience as listeners and readers to effectively monitor their own output; they become self-listeners and self-readers. When a self-listener or self-reader detects a problem in his or her speech or writing and responds by amending the problem, he or she is self-editing. Writers rely on self-editing---more commonly called revision---to shape their expressions toward maximal effectiveness.; This study explored a new methodology for investigating self-editing (revision) in writing. In each of two experiments, 30 college students composed essays in response to 2 writing prompts. A keystroke-logging program captured various aspects of the writing and editing behavior, which then were summarized by 24 dependent measures spanning 6 categories: pauses, pause-associated edits, edits without pauses, total edits indices, quality/readability, and question reviews.; Experiment 1 examined whether the nature of the writing prompt affected writing and editing. Data confirmed that measures of writing and editing did vary as a function of writing prompt. Experiment 2 then varied the amount of time available before beginning composition (pre-response time) and found that some writing and editing behaviors varied as a function of pre-response time. Both experiments compared the writing behavior of students who scored high in print exposure to that of students who scored low in print exposure. For several writing measures, high and low print exposure writers responded differently to the experimental manipulations.; The methodology explored in this study proved promising as an approach to analyzing self-editing in writing. In particular, the behaviorally based visual analyses of distributions and rates for keystrokes and pauses add an analytical dimension otherwise left unconsidered in the field. In addition, various measures of writing and editing behavior demonstrated sensitivity to situational variables. Moreover, the consistent role of print exposure in the interactive effects found for some writing measures established a link between print exposure and writing, a relationship that, to date, has not received much attention.