Using structural equation modeling the authors evaluated the contribution of morphological awareness, phonological memory, and phonological decoding to reading comprehension, reading vocabulary, spelling, and accuracy and rate of decoding morphologically complex words for 182 4th-and 5th-grade students, 218 6th-and 7th-grade students, and 207 8th-and 9th-grade students in a suburban school district. Morphological awareness made a significant unique contribution to reading comprehension, reading vocabulary, and spelling for all 3 groups, to all measures of decoding rate for the 8th/9th-grade students, and to some measures of decoding accuracy for the 4th/5th-grade and 8th/9th-grade students. Morphological awareness also made a significant contribution to reading comprehension above and beyond that of reading vocabulary for all 3 groups.

The authors used multiple-group structural equation modeling to analyze structural relationships between latent factors underlying separate measures of handwriting, spelling, and composing in Grades 1-6. For compositional fluency, the paths from both handwriting and spelling were significant in the primary grades, but only the path from handwriting was significant in the intermediate grades. For compositional quality, only the path from handwriting was significant at the primary and intermediate grades. The contribution of spelling to compositional quality was indirect through its correlation with handwriting. Handwriting and spelling accounted for a sizable proportion of the variance in compositional fluency (41 % to 66%) and compositional quality (25% to 42%). These findings show that the mechanical skills of writing may exert constraints on amount and quality of composing. Theoretical and educational implications of the findings are discussed.Mechanical requirements for producing text have been hypothesized to contribute to individual differences in writing performance in several important ways (Berninger et al., 1992;Graham, 1990;Scardamalia, Bereiter, & Goleman, 1982). For persons who have not yet mastered the mechanics of writing, having to attend to the lower level skills of getting language onto paper may "tax" a writer's processing capacity in working memory, interfering with higher order skills such as planning and content generation. Having to switch attention during composing to mechanical demands, such as figuring out how to spell a word, may lead the writer to forget already developed ideas and plans. Simultaneously allocating attention to mechanical concerns while trying to plan the next unit of text may further interfere with the planning process, affecting the complexity and coherence of content integration. If attention is occupied with mechanical concerns, the writer may also have less opportunity to make expressions more precisely fit intentions at the point of translation. Finally, the writer's fluency with handwriting (or typing) may not be fast enough to keep up with his or her thoughts, interfering with content generation and recall of ideas or text already planned and held in working memory. However, the empirical research on the relationship between the mechanics of production and the composing process has yielded mixed results, depending on age and

We used multiple-group structural equation modeling to analyze structural relationships between latent factors underlying writing-related developmental skills and component writing skills in Grades 1-6. For handwriting, both motor skills and orthographic coding contributed to the model fit, but only the path from orthographic coding was significant at all grade levels. For spelling, only the path from orthographic coding was significant in the primary grades, but both that path and the path from phonological coding were significant in the intermediate grades. For compositional quality, both reading and oral language contributed in the primary grades but the model was unclear in the intermediate grades because of high covariance between those factors. Theoretical and educational implications of the findings are discussed.The purpose of our research was (a) to analyze the relationships between individual differences in developmental skills that children bring to the task of learning to write and individual differences in component writing skills and (b) to determine whether these relationships change over the course of development. Before discussing the rationale for the data analysis techniques and measures used and the specific hypotheses tested, we clarify the purpose of this research within the context of the larger literature on writing development. In lieu of a comprehensive review of the rapidly growing body of writing research, which can be found in diverse literatures including educational psychology, cognitive psychology, developmental psychology, neuropsychology, psychoeducational assessment, teacher education, teaching of English, special education, linguistics, and medicine, we provide a brief overview of those aspects of this diverse literature that clarify how our work fits into the broader field of writing.

Almost 700 children were screened to identify 144 1st graders at risk for handwriting problems who were randomly assigned to 1 of 6 treatment conditions. Treatment was delivered to groups of 3 that met twice a week in 20-min sessions until they completed 24 lessons. Five groups received 10 min of different kinds of handwriting instruction. The contact control group received 10 min of phonological awareness training. All 6 groups composed and shared their writing for 10 min. Converging evidence across multiple measures showed that combining numbered arrows and memory retrieval was the most effective treatment for improving both handwriting and compositional fluency (composing with time limits). Thus instruction aimed at improving transcription transfers to improved text generation in beginning writers.

Longitudinal structural equation modeling was used to evaluate longitudinal relationships across adjacent grade levels 1 to 7 for levels of language in writing (Model 1, subword letter writing, word spelling, and text composing) or writing and reading (Model 2, subword letter writing and word spelling and reading; Model 3, word spelling and reading and text composing and comprehending). Significant longitudinal relationships were observed within and across levels of language: spelling to spelling and spelling to composing (Grades 1 to 7), Models 1 and 3, and composing to spelling (Grades 3 to 6, Model 1; Grades 4 to 6, Model 3); spelling to word reading and word reading to spelling (Grades 2 to 7), Models 2 and 3; spelling to word reading (Grade 1), Model 2, and word reading to spelling (Grade 1), Model 3; composition to comprehension (Grades 3 to 5), Model 3; comprehension to composition (Grades 2 to 6), Model 3; and comprehension to word reading (Grades 1 to 6), Model 3. Results are discussed in reference to the levels of language in translating ideas into written language and integrating writing and reading.Learning to write, that is, to produce legible letters and conventional spellings and create coherent written texts for many specific assignments across the curriculum, is a major task of early and middle childhood for children educated in schools. Children who do not master writing are at risk for grade retention and may not graduate because they cannot complete written assignments and take tests. Also, they may not pass high-stakes tests, which increasingly require writing (Jenkins, Johnson, & Hileman, 2004). However, there is relatively little research on the normal writing development process with which to inform educational policy, assessment, and instruction related to writing. Of the research that does exist on writing development, most of it provides (a) a snapshot of the writing process at a specific point in time or (b) a cross-sectional comparison of different children at different phases of writing development. Few longitudinal studies of writing exist, or, as Lerkkanen, Rasku-Puttonen, Aunola, and Nurmi (2004) pointed out, if the same children were tested repeatedly, the results have often been analyzed in cross-sectional fashion (e.g., . Moreover, comparatively little research has focused on writing alone. The research on writing that does exist is often focused on writing-reading relationships (Fitzgerald & Shanahan, 2000). As explained later in this section, research focused on writing alone is typically grounded in the cognitive processes of writing, such as planning, translating, and reviewing/revising, rather than on the levels of language involved in translating ideas into written products."Levels of language" is a well-established analytic tool in linguistics that is used to understand how complex, multidimensional language is structured in the mind and in the language constructions of language users (e.g., Berninger & Garvey, 1982). Linguists analyze and code different levels of langua...