09:00 - 10:45 : Individual Papers - Writing

Lead author	Mark Torrance
Institution/organisation	Staffordshire University
Co-author(s)
Institution(s)/organisations
Title	Strategy-focussed writing instruction: it works, but why?

Abstract

71 normally-functioning Spanish sixth grade students participated in classroom-based cognitive strategy training aimed at developing the quality of their written composition. Compared with controls, intervention participants showed improved text quality and a greater tendency to preplan their text. This effect was sustained at 3 months and at 28 months after the intervention. Immediately after the intervention tendency to preplan was a robust predictor of text quality. This effect was absent at 28 months. In a second study 75 students received either strategy-focussed instruction or instruction that just emphasised setting product goals. Both interventions had substantial and positive effects on students writing quality, and these effects were not reliably different. These findings suggest, perhaps, that the success of strategy-focussed interventions is associated more with students' developing understanding of what constitutes a good finished text than with regulation of their writing strategies.

---

Lead author	David Galbraith
Institution/organisation	Staffordshire University
Co-author(s)
Institution(s)/organisations
Title	Constructing knowledge objects in writing

Abstract

This paper describes the results of an experiment designed to test a dual-process model of writing, which suggests that the development of ideas in writing depends on an interaction between two distinct types of process: explicit planning and spontaneous text production.

96 undergraduates, divided into two groups of low and high self-monitors, were asked to write under one of three different conditions: (i) a rough draft condition; (ii) a planned essay condition; and (iii) a control condition. Participants were asked to generate lists of ideas before and after writing, and to compare the ideas they contained. This enabled us to calculate the extent to which they had developed new ideas through writing. They were also asked to rate the degree of relationship between ideas within the lists produced before and after writing. This enabled us to calculate a measure of conceptual coherence and to assess the effect of different forms of writing on it.

The results for new ideas replicated those of previous research, with low self-monitors producing a greater number of new ideas after writing rough drafts than high self monitors, but a similar number of new ideas after writing planned essays. Despite the similar number of new ideas produced in the planned essay condition, however, the ideas produced by low self-monitors were significantly more coherent than those produced by high self-monitors. Furthermore, there were highly significant negative correlations between number of new ideas and conceptual coherence for both the low and high self-monitor’s planned essays.

We will argue that these results (i) support the claim of the dual process model that dispositionally guided text production leads to more conceptually coherent ideas than rhetorical planning, and (ii) suggest that immediate changes in conceptual coherence as a consequence of writing are restricted by the limited capacity of working memory.

---

Lead author	Elicia Cooke
Institution/organisation	Staffordshire University
Co-author(s)	Mark Torrance, David Galbraith
Institution(s)/organisations	Staffordshire University
Title	Students' reviewing and editing practices

Abstract

The study looks at reviewing and editing text, specifically what students think it is, whether what they do is what they say they do, and whether changes students make to their writing actually improve their text. The sample consisted of ten year three (a range of abilities) and year six students (mainly top band students). Each participant completed a total of three activities. Firstly, students were asked to write a letter to their head teacher asking for more school trips, this was done on a writing tablet so writing could be viewed and stored via a program called eye and pen on a computer as they wrote. Students were then questioned about their writing with regard to any changes they may have made, on paper or in their head. Secondly, students completed an editing task, they were given a letter trying to persuade the Prime Minister to allow them three days off at the weekend instead of two, they had to go through and highlight any changes they would make to the text and explain why. Lastly, students were interviewed about reviewing and editing, asking what changes they made to their writing, whether they thought this was a small or big change and so on. All interviews were recorded via Dictaphone and transcribed. The analysis is underway and will be a mixture of RST (for the second activity, so a baseline of quality of text can be established and whether changes made actually improved text), and coded answers for the questions that deal with more metalanguage/ metacognitive questions i.e. why did you make that change? I will hope to answer questions about what students understand by reviewing and editing, and whether the changes they make improve their text.

---

Lead author	Christian Weinzierl
Institution/organisation	Heidelberg University of Education
Co-author(s)	Joachim Grabowski1, Cornelia Glaser2
Institution(s)/organisations	1Heidelberg University of Education, 2Justus-Liebig University Giessen
Title	Using keystroke logging in school research: From mere typing to text revision

Abstract

While the study of writing with keystroke-logging software is typical of laboratory research, we show examples of how keystroke logging was employed in school environments. Nowadays, writing literacy is not restricted to handwriting, because a prerequisite of many occupations (to which school is obliged to prepare) is some degree of keyboard mastery. Keystroke logging (here: ScriptLog) provides real-time records of all keyboard activities.

(1) Fluent keyboard typing does not necessarily depend on the mastery of touch-typing methods; many students acquired some individual typing style. We analysed typing patterns while performing on a basic kind of writing task: copying text. A factor analysis revealed that students' keyboard mastery comprises of three factors: speed, efficiency, and precision. Eight-graders outperform fifth-graders, and both still do better in handwriting than in typing. (Data from 32 university and 86 secondary school students.)

(2) On a much higher level of writing skills, the number and category of text revisions can predict the quality of the final text. 43 seventh-graders of different school types (and a control group of 24 university students) were to edit and revise their own previously composed fictitious narrations. Keystroke logging allowed for the classification of revisions according to multiple dimensions, while basic typing speed was controlled.

Referring to the two examples of empirical work, we will discuss the use of controlled settings of keyboard writing in school for the study of writing from the very basic level of mastering the input device to high-level measures for the improvement of text quality.