Mark Torrance

Combined eyetracking and keystroke-logging methods for studying cognitive processes in text production

Writers typically spend a certain proportion of time looking back over the text that they have written. This is likely to serve a number of different functions, which are currently poorly understood. In this article, we present two systems, ScriptLog+TimeLine and EyeWrite, that adopt different and complementary approaches to exploring this activity by collecting and analyzing combined eye movement and keystroke data from writers composing extended texts. ScriptLog+TimeLine is a system that is based on an existing keystroke-logging program and uses heuristic, pattern-matching methods to identify reading episodes within eye movement data. EyeWrite is an integrated editor and analysis system that permits identification of the words that the writer fixates and their location within the developing text. We demonstrate how the methods instantiated within these systems can be used to make sense of the large amount of data generated by eyetracking and keystroke logging in order to inform understanding of the cognitive processes that underlie written text production.

Revision in the Context of Different Drafting Strategies

Our position in this chapter is that revision cannot be understood independently of the writing strategies in which it is embedded. Some authors have promoted an interactive writing strategy (e.g., Elbow, 1998) which relies on multiple rewriting as a means of developing text content in lieu of producing a structured outline. Galbraith’s (1999) dual-process model, which describes writing processes as an interleaving of dispositional content generation and rhetorical structuring, provides some basis for understanding why this kind of strategy might be successful. We review existing research exploring the efficacy of outlining and rough-drafting strategies. This suggests a clear benefit for outline-based strategies over various forms of rough-drafting strategies, but has not, we argue, included an appropriate form of interactive writing strategy. We then discuss research exploring the writing processes of writers performing non-laboratory real-world writing tasks and find that a minority of writers, and particularly more experienced writers, appear to habitually adopt a nonutline multiple-drafting strategy similar in form to that specified by the dual-process model. Finally, we describe an experiment in which we found (a) that this form of interactive strategy was more effective than the forms of rough drafting strategy which have been investigated in previous research, (b) that, consistent with previous research, outlining also has clear text-quality advantages, and (c) that the effectiveness of revision depends on the form of initial draft to which it is applied.

Knowing What to Write

Knowing What to Write brings together new and recent research and theory exploring the cognitive processes involved in retrieving, ordering and creating knowledge during text production. Contributions from cognitive psychology, text linguistics, psycholinguistics, and computer science combine to provide a sophisticated and wide-ranging picture of the ways in which writers develop and structure their ideas. This book is recommended both to writing researchers and to any readers interested in the cognitive processes that lie behind complex language behavior.

Typing pictures: Linguistic processing cascades into finger movements

The present study investigated the effect of psycholinguistic variables on measures of response latency and mean interkeystroke interval in a typewritten picture naming task, with the aim to outline the functional organization of the stages of cognitive processing and response execution associated with typewritten word production. Onset latencies were modulated by lexical and semantic variables traditionally linked to lexical retrieval, such as word frequency, age of acquisition, and naming agreement. Orthographic variables, both at the lexical and sublexical level, appear to influence just within-word interkeystroke intervals, suggesting that orthographic information may play a relevant role in controlling actual response execution. Lexical-semantic variables also influenced speed of execution. This points towards cascaded flow of activation between stages of lexical access and response execution.

Timed written picture naming in 14 European languages

We describe the Multilanguage Written Picture Naming Dataset. This gives trial-level data and time and agreement norms for written naming of the 260 pictures of everyday objects that compose the colorized Snodgrass and Vanderwart picture set (Rossion & Pourtois in Perception, 33, 217–236, 2004). Adult participants gave keyboarded responses in their first language under controlled experimental conditions (N = 1,274, with subsamples responding in Bulgarian, Dutch, English, Finnish, French, German, Greek, Icelandic, Italian, Norwegian, Portuguese, Russian, Spanish, and Swedish). We measured the time to initiate a response (RT) and interkeypress intervals, and calculated measures of name and spelling agreement. There was a tendency across all languages for quicker RTs to pictures with higher familiarity, image agreement, and name frequency, and with higher name agreement. Effects of spelling agreement and effects on output rates after writing onset were present in some, but not all, languages. Written naming therefore shows name retrieval effects that are similar to those found in speech, but our findings suggest the need for cross-language comparisons as we seek to understand the orthographic retrieval and/or assembly processes that are specific to written output.

Adolescent weak decoders writing in a shallow orthography: process and product

ABSTRACT It has been hypothesized that students with dyslexia struggle with writing because of a word-level focus that reduces attention to higher level textual features (structure, theme development). This may result from difficulties with spelling and/or difficulties with reading. Twenty-six Norwegian upper secondary students (M = 16.9 years) with weak decoding skills and 26 age-matched controls composed expository texts by keyboard under two conditions: normally and with letters masked to prevent them reading what they were writing. Weak decoders made more spelling errors and produced poorer quality text. Their inter-key-press latencies were substantially longer preword, at word end, and within word. These findings provide some support for the word-level focus hypothesis, although we found that weak decoders were slightly less likely to engage in word-level editing. Preventing reading did not affect differences between weak decoders and controls, indicating that their reduced fluency was associated with production rather than reading difficulties.

Effects of ADHD on writing composition product and process in school-age students

Objective: This study examined the relationship between ADHD and writing performance. Method: Students in Grades 3 to 7, 84 with ADHD and 135 age- and gender-matched controls completed a writing task (including process logs), and measures of working memory and attention. Results: Students with ADHD wrote texts of similar length but with poorer structure, coherence, and ideation. In all, 6.7% of the variance in writing quality was explained by whether or not the student had an ADHD diagnosis, after control for IQ and age-within-year, with ADHD students producing text that was less coherent, well structured, and ideationally rich, and spending less time thinking about and reviewing their text. Half of the effect on text quality could be attributed to working memory and sustained attention effects. Conclusion: ADHD has some effect on writing performance, which can, in part, be explained by working memory and attentional deficits.

Effects of Direct Instruction and Strategy Modeling on Upper-Primary Students’ Writing Development

Strategy-focused instruction is one of the most effective approaches to improve writing skills. It aims to teach developing writers strategies that give them executive control over their writing processes. Programs under this kind of instruction tend to have multiple components that include direct instruction, modeling and scaffolded practice. This multi-component nature has two drawbacks: it makes implementation challenging due to the amount of time and training required to perform each stage, and it is difficult to determine the underlying mechanisms that contribute to its effectiveness. To unpack why strategy-focused instruction is effective, we explored the specific effects of two key components: direct teaching of writing strategies and modeling of strategy use. Six classes (133 students) of upper-primary education were randomly assigned to one of the two experimental conditions, in which students received instruction aimed at developing effective strategies for planning and drafting, or control group with no strategy instruction: Direct Instruction (N = 46), Modeling (N = 45), and Control (N = 42). Writing performance was assessed before the intervention and immediately after the intervention with two tasks, one collaborative and the other one individual to explore whether differential effects resulted from students writing alone or in pairs. Writing performance was assessed through reader-based and text-based measures of text quality. Results at post-test showed similar improvement in both intervention conditions, relatively to controls, in all measures and in both the collaborative and the individual task. No statistically significant differences were observed between experimental conditions. These findings suggest that both components, direct teaching and modeling, are equally effective in improving writing skills in upper primary students, and these effects are present even after a short training.

Chapter 2.02.06: The Effects of Dyslexia on the Writing Processes of Students in Higher Education

Undergraduates with dyslexia typically report that writing is the activity that they find most problematic in their studies. Twenty-eight undergraduates registered as dyslexic at a university in the United Kingdom were compared with 32 undergraduates matched for age, gender and academic discipline. Tests of intelligence and of reading and spelling skills showed that the two groups were of equal intelligence but that the dyslexic students performed significantly worse on the reading and spelling tests and on a test of working memory span. Dyslexics made more spelling mistakes than non-dyslexics. Even after texts were corrected for spelling and capitalization, their texts were rated as of much lower quality by two independent judges. There were no differences in proportions of time which dyslexics and non-dyslexics spent on the different writing processes or in the effort they put into the processes. Keywords: dyslexia; dyslexic students; United Kingdom; writing processes

List of Volumes

This section contains a list of volumes that have been cited for the book Learning to Write Effectively: Current Trends in European Research . The study of writing from a developmental perspective inquires about the particular ways in which knowledge of writing evolves over time. The term writing, however, has multiple meanings. It can be employed for referring to a cultural practice that fulfills different functions, writing can be used, for example for mnemonic, religious or poetic functions. Moreover, writing is a cultural, institutionally supported practice; children learn to write at school. Their evolving notions about the functions of writing, the way in which they handle the spelling system of their language and their grasp of the different genres of discourse are unavoidably intertwined with schooling. Keywords: mnemonic; spelling system; writing