Scaffolding the writing of students with disabilities through procedural facilitation: using an Internet-based technology to improve performance.

Relatively little is known about the potential of web-based programs to support and scaffold the writing performance of students with disabilities. In this study, an experimental and control group of students planned and organized their ideas in order to write expository papers about self-selected topics. The experimental group used a web-based environment that scaffolded performance by prompting attention to the topical organization and structure of ideas while students were in the situated act of composing their papers. Control students used similar writing tools, but in traditional paper-and-pencil print formats. The results of the quasi-experimental study revealed that the students in the web-based scaffolding condition produced lengthier pieces and received significantly higher ratings on the primary traits associated with writing quality. The greatest effects were evident in terms of experimental students' abilities to produce topic sentences and to generate more topically coherent pieces overall. These findings suggest the web-based software can be designed to offer universally available supports to improve the performance of struggling writers.

**********

Writing is a challenging cognitive process. According to the results of the National Assessment of Educational Progress (NAEP; 2002) only 28% of fourth-grade writers attained performance levels that were judged to be at or above proficient, whereas 72% exhibited partial mastery of the knowledge and skills that are considered fundamental for the fourth grade (Persky, Daane, & Jin, 2003). On the NAEP informational writing task, fewer than 17% of the fourth-grade students performed at the proficient level, as measured by students' production of texts that included main ideas and details and that displayed sensitivity to their targeted audience (Persky et al., 2003). Clearly, students have difficulty composing texts, and they manifest the greatest problems when they are asked to generate informational or expository texts.

Among school-age populations, students with learning disabilities (LD) face some of the most serious writing challenges, which affects their school performance (Graham, 2006; Harris, Graham, & Mason, 2003; Singer & Bashir, 2004; Troia, 2006). Students with LD perform more poorly than their typically achieving counterparts on nearly every measure of writing performance, including measures of text length, quality and organization, and mechanical and grammatical errors (Gersten & Baker, 2001; Troia, 2006). The research literature reveals several underlying problems that interfere with the attainment of proficient levels of writing.

First, a body of literature indicates that students with LD demonstrate difficulties in employing and regulating cognitive and writing strategies (Graham, 2006; Troia, 2006). Specifically, Graham and Harris (Graham & Harris, 1989, 1993; Graham, Harris, MacArthur, & Schwartz, 1991; Troia, Graham & Harris, 1999) noted that students with LD have difficulties in approaching text composition in planful ways, with concomitant deficiencies in their ability to establish purposes for writing or to prepare to write by activating background knowledge. Similarly, several studies indicate that students with LD do not anticipate or generate information consistent with prior topical information (Englert & Thomas, 1987; Williams, 2003), but instead tend to generate ideas in an associative fashion (Bereiter & Scardamalia, 1985; Graham & Harris, 2003; Singer & Bashir, 2004). The end result is an amalgamation of rambling ideas that fail to launch, present, or advance the writers' thoughts in a coherent fashion.

A second difficulty pertains to a lack of knowledge of text structure, which informs writers about the discourse-level arrangement of ideas in text. Different types of texts are organized in different ways. Meyer (Meyer, 1975, 1977; Meyer, Brandt, & Bluth, 1980), for example, identified six patterns for the organization of expository information: (a) compare-contrast; (b) problem-solution; (c) description or enumeration of traits, properties, or functions; (d) temporal or chronological sequence; (e) classification; and (f) explanation. Familiarity with these types of text structures helps writers generate the information that is required for adequate topical coverage, as well as to employ the semantic and syntactic devices that slot the required information in the appropriate textual locations (Englert, Hiebert, & Stewart, 1988; Fitzgerald & Teasley, 1986; Kamberelis, 1999).

Unfortunately, many students with LD are not conscious of the topical, semantic, and syntactic properties of these expository text structures, which negatively impacts their ability to generate organizationally sophisticated written texts (Englert & Raphael, 1988; Graham, 1990; Graham & Harris, 1989; Thomas, Englert, & Gregg, 1987; Williams, 2003; Wong, 1997). Consequently, their written texts often exclude essential information or fail to communicate the relationships among the textual ideas in a manner that satisfies the knowledge requirements of their readers (Englert et al., 1988; Thomas et al., 1987; Vaughn, Gersten, & Chard, 2000; Wong, 1997; Wong, Butler, Ficzere, & Kuperis, 1996).

Third, well-written expository texts are organized at the local level through the provision of introductory statements that orient the readers to the subtopics, text structure, and purpose of the text, and that set the stage for the information that will follow (Halliday & Martin, 1993). Each of the local subtopics is further signaled by introductory or preparatory statements, which are embellished with relevant details that offer topical breadth and depth (Halliday & Martin, 1993). Recursively, therefore, written text contains introductory and concluding statements that offer a wave-like function through the text, simultaneously setting up and wrapping up the subtopical presentations of information to achieve effective communication (Halliday & Martin, 1993).

Skilled writers use this knowledge of text organization as a basis for planning, selecting, generating, organizing, composing, monitoring, and communicating their textual ideas (Meyer et al., 1980; Spivey, 1997; Vaughn et al., 2001). In contrast, students with LD have difficulty in generating text in a manner that accomplishes a wave-like progression through their ideas (Wong, 1978). Compared to their grade-age peers, students with LD tend to generate texts that are poorly organized and that fail to launch and wrap up the meaning relationships among a particular set of ideas (Englert, Raphael, Fear, & Anderson, 1988; Englert et al., 1991; Englert & Thomas, 1987). Often the legibility of their expository texts is further threatened by writing problems associated with (a) knowledge retrieval problems, resulting in shorter pieces; (b) difficulties knowing how and what to revise, resulting in incomprehensible texts; and (c) a lack of automaticity in basic skills (grammar, punctuation, writing fluency, word retrieval) (Graham & Harris, 2003; Graham et al., 1991; Harris et al., 2003; Singer & Bashir, 2004).

Instructional Interventions

The research literature has yielded some consistent results pertaining to the instructional practices that address the writing difficulties of students with disabilities. Chief among the instructional recommendations is an emphasis on direct instruction of the expository text structures, together with the use of teaching methods and tools that highlight the organization of ideas into patterned relationships (Baker, Gersten, & Graham, 2003; Englert et al., 1991; Gersten & Baker, 2001; Vaughn et al., 2000).

The provision of organizational frameworks (semantic webs, graphic organizers), for example, helps make text structures more visible to students, enhancing their abilities to identify, organize, and generate a structured set of ideas (Berkowitz, 1986; Idol, 1987; Idol & Croll, 1987; Kim, Vaughn, Wanzek, & Wei, 2004; Meyer, 1995; Scanlon, Duran, Reyes, & Gallego, 1992; Sinatra, Beaudry, Pizzo, & Geisert, 1994; Sinatra, Berg, & Dunn, 1985; Wong, 1997; Wong et al., 1996). Organizational frameworks scaffold performance by eliciting the types of information that is expected for a given text structure, and by prompting writers to adopt a hierarchical approach to topical presentation. Relatedly, teachers who model the writing process by thinking aloud about the cognitive strategies that inform the planning, organizing, transcription, and monitoring processes have a positive effect on their students' writing performance (Baker et al., 2003; Englert et al., 1991; Englert & Mariage, 2003; Vaughn et al., 2000; Wong & Berninger, 2004). Development of students' text structure knowledge through a combination of teaching and learning techniques offers the best prospect for improving their performance by increasing students' abilities to adopt a strategic and planful approach as they gather, organize, and compose their expository ideas.

However, the provision of organizational frameworks and strategies is not always sufficient. Often students' uses of text structure and strategies must be further prompted and supported in the situated context of writing activity. A meta-analysis performed by Gersten and Baker (2001) showed the benefits of prompted writing performance through the provision of procedural facilitators or scaffolds. Procedural facilitators, which can include mnemonics, questions, prompts, think sheets, or simple outlines of important learning structures (Baker, Gersten, & Scanlon, 2002; Baker, Gersten, & Graham, 2003; Bereiter & Scardamalia, 1985; Rummel, Levin, & Woodward, 2003), aid cognition by supporting and reminding students to activate and employ the strategies during the planning, composing, and revising process (Baker et al., 2003; Singer & Bashir, 2004; Wong & Berninger, 2004; Wong et al., 1996). De La Paz (De La Paz, 1997, 1999; De La Paz & Graham, 2002), for example, used printed prompts to cue students to plan, write, monitor, evaluate, and revise their texts based on expert writing strategies and criteria. With the anchoring support of such procedural facilitators, students with LD were guided to use strategic actions that they otherwise might not have employed (Baker et al., 2002; De La Paz & Graham, 2002), resulting in an overall enhancement of their writing proficiency (Baker et al., 2003; De La Paz, 1999; De La Paz & Graham, 2002).

Technology as a Scaffolding Environment

Incorporating scaffolds into the fabric of the writing process remains an instructional challenge. By its very nature, writing is often private, whereas scaffolds imply a social or cognitive relationship between the user and the procedural facilitator (Stone, 1988). In this regard, technology may have a critical role in supporting learners in performing the writing process. Technologies can be designed to offer scaffolds that lead cognitive functions that are newly emerging, and to prompt routines and processes in a timely way just as a tutor might prompt students to employ particular writing procedures and actions. Thus, the National Commission on Writing (2003) offered its opinion that "new technologies can advance both the teaching and learning of writing" (p. 31). The report called for further investment in technology research that might aid students' writing.

Two types of organizational tools may support writing performance through the provision of technology-assisted devices: (a) graphic organizers and mapping tools, and (b) assistive technologies.

Graphic organizers and mapping tools. This type of scaffold enables writers to organize what they know into relational databases that facilitate information generation and retrieval (Jonassen & Carr, 2000). For example, Inspiration (Inspiration, 1994) is a software program that allows students to organize their ideas into a number of possible structural arrangements, such as compare-contrast, explanation, flowchart, and so on. The software extends writers' capabilities by offering mapping tools that support students as they organize their ideas. Similarly, Anderson-Inman (Anderson-Inman, Redekopp, & Adams, 1992; Anderson-Inman & Tenny, 1989; Anderson-Inman & Zeitz, 1994) developed computer-based outlining programs for improving the text comprehension and study strategies of students with LD. Studies showed that the computer-based outlining programs were effective in improving the test performance of students with LD (Anderson-Inman, Knox-Quinn, & Horney, 1996). Provided textual representations and graphic organizers of the underlying structure for a writing topic, students with LD were more successful at including the essential elements of the expository text structure (Anderson-Inman et al., 1996; Anderson-Inman et al., 1992). Similarly, mapping tools facilitated performance by advancing students' use of cognitive operations that were otherwise underemployed in reading and writing. In this manner, students could distribute or offload some of the cognitive strategies and representational work onto the technology, freeing up their memory and cognition for other generative and productive processes.

Assistive technologies. This type of organizational tool offers discourse-level prompts that help students stage their ideas and develop the macrostructure of their expository papers through the provision of cognitive frameworks. TELE-Web (Technology-Enhanced Learning Environments on the Web), for example, is an Internet-based software that offers several structural devices upon which students can frame their thoughts, words, and ideas (Englert & Zhao, 1996). At one level, TELE-Web helps students stage the text structure by reminding them to generate the introductory statements that prepare readers for the forthcoming text. At another level, TELE-Web offers reminders of the need to generate supporting evidence and details, as well as to produce concluding sentences that fulfill the discourse-level expectations of the readers. Thus, TELE-Web helps students fulfill the requirements of a well-organized text by providing a cognitive anchor on which students can generate papers that contain the global or discourse-level structural elements (introduction, body paragraphs, conclusion) of expository prose, and by prompting them to generate the topically related chunks of information that fulfill the knowledge requirements of readers. TELE-Web addresses some of the difficulties of students with disabilities in anticipating, predicting, and composing topically related chunks of information while they are in the situated circumstance of text composition.

Results of prior research have suggested the potential of the TELE-Web scaffolding to support the performance of low-achieving first-grade students (Englert, Manalo, & Zhao, 2004) and students with LD (Englert, Wu, & Zhao, 2005). Thus, the results of several studies indicate that the scaffolded and supported conditions facilitated by TELE-Web enhanced the writing performance of low-achieving writers to a greater degree than when the same students wrote about similar topics in paper-and-pencil and word processing conditions (Englert et al., 2004; Englert et al., 2005). However, these studies included a relatively small number of students, which limited the statistical power, and the interventions required that students serve as their own controls in an alternating-treatment design.

The purpose of the present research was to investigate the effects of scaffolding students' writing performance through the employment of two different conditions that were exactly similar with the exception of the online scaffolding environment available to students. In both conditions, students mapped their ideas using a graphic organizer during a planning phase, and then transformed their mapped ideas into expository texts in a writing phase. The primary difference between the two groups was their access to technology supports during the planning and writing phases: One group completed the mapping and writing tasks using a traditional paper-and-pencil format; the other group accessed the scaffolding tools and mapping technologies through the TELE-Web Internet-based software. TELE-Web scaffolds were also provided during the writing phase to highlight the textual locations where specific writing devices (e.g., writing introductions, details, conclusions) were desired. Thus, TELE-Web scaffolded writing at the point of composition, which was not as feasible in traditional paper-and-pencil formats.

The question undergirding this study was: Does the provision of text structure and organizational scaffolds in an online environment influence the writing performance (e.g., use of genre features and writing productivity) of students with disabilities who have writing difficulties? Given the focus on providing scaffolds that support students' attention to text structure, it was hypothesized that the computer-assisted scaffolding conditions would facilitate the writing performance of students with disabilities as evaluated through their incorporation of the text structure elements into their written texts. That is, students would include more of the structural properties of the expository text structures in their papers, such as topic sentences and related details, that offered depth and breadth to their coverage of their topics. However, the scaffolding was not expected to influence students' written productivity or conventional performance (e.g., capitalization, punctuation).

METHOD

Subjects

A total of 35 elementary-age students with disabilities participated in the study, 20 students in the experimental condition and 15 students in the control condition. Students were drawn from six special education classrooms across five urban schools. All special education students who received classroom accommodations and supports in writing, based on their IEPs, were first selected. Of these, only students who received writing services were included in this study. Thirteen students with LD were in the TELE-Web group, and 11 students with LD were in the comparison group. In addition, 7 students and 4 students, respectively, with disabilities other than LD were in the experimental or comparison groups (see Table 1). The mean age of students in the Tele-Web condition was 10.64, whereas the mean age of students in the comparison group was 9.64. Although TELE-Web students were slightly older, these differences were not statistically significant, t(31) = 1.80 (p > .05).

The distribution of participants across the two conditions is shown in Table 1. Reading scores were based on standardized scores obtained from the STAR Reading Test (Renaissance Learning, 2002) administered in September. The Star Reading Test is a standardized, multiple-choice computerized measure of reading performance based on the cloze procedure. Studies indicate that the STAR Reading Test yielded correlation coefficients with the Stanford Achievement Test for students in grades 1-6 ranging from .79-.89, and reliability coefficients with the Gates-MacGinitie Reading Test that ranged from .80-.89 (Renaissance Learning, 2005). To test the equivalence of the literacy skills of the students in the two conditions based on their reading scores, a t-test of students' STAR reading scores was conducted. No significant differences were found between the reading levels of the students in the two comparison groups, t(31) = .677, p > .05.

In addition, students' writing was analyzed by evaluating their pretest stories based on total number of words, t(33) = .364 (p > .05), and their papers' total primary trait score, t(31) = -.894 (p > .05). The total primary trait score was a summation of the specific text structure qualities of their papers, such as the presence of topic sentences, relevant details, depth and breadth of topic coverage, and concluding sentences. (These traits will be more fully explained in a later section.) The two groups were not distinguished on either the total number of words in the papers or the total primary trait scores, and inspections of the individual dependent variables that comprised the composite primary trait score revealed no significant differences between the groups on any dependent measure of prewriting performance.

Experimental and Control Groups

The TELE-Web group was comprised of teachers and students who had routinely used the TELE-Web software in support of their literacy curriculum. The TELE-Web teachers had participated in a series of three to four workshops in which they received information about the teacher and student interface. Specifically, the researchers demonstrated the computer software, explained its uses in literacy instruction, provided technical support, and developed procedures to ensure consistency in the research procedures. At the time of the study, the teachers and students were very familiar with the TELE-Web interface, and had used the TELE-Web interface to produce expository papers on two or more occasions.

The second intervention group consisted of exemplary special education teachers who had been instructed in a process approach to writing. These comparison teachers used effective writing strategies, including brainstorming, mapping, editing, and revising, to support their students with disabilities during the writing process. Throughout the year, they participated in university workshops that focused on effective teaching and best practices, and they were selected to be collaborating teachers who worked with teaching interns from the university. However, they differed from the first group in their school's technology readiness and their uses of technology in the classroom. That is, they were not frequent users of technology in support of their literacy curriculum, and they did not have the same access to Internet technologies at either the classroom or the building level.

Procedure

A baseline sample of writing performance was collected two weeks prior to the start of the intervention in order to gather information about students' entry-level performances in writing an informational paper about a familiar topic. To create a context for the baseline assessment, students were instructed to write an informational paper about a farm animal (e.g., pigs, chickens, goats, geese, turkeys, horses). They were told to assume that they were writing the paper for someone from another country who was not familiar with American farm animals. Students were instructed to offer their reader as much information as possible about their chosen animal. No additional structure or support was provided for this assessment, and all students produced their texts using paper and pencil. Once the papers were completed, students participated in one of the two instructional conditions.

Paper-and-pencil condition. In the paper-and-pencil condition, teachers informed students that they would be employing strategies for planning and organizing their papers. On the first day, teachers told students that they would be writing an informational paper for people who did not own any pets. Students also were told that they would be writing an informational paper that explained to their readers (a) what is the pet, (b) why it makes a good pet, (c) what it looks like, (d) what it eats, (e) what it does, and (f) how to care for the pet.

The teacher then presented a concept map where the categories were labeled with the titles of the informational categories, as shown in Figure 1. Students were told to brainstorm ideas by recording information in the relevant categories of the web. The teacher modeled how to perform this process using the example of a pet parrot, guiding students to suggest details for the category "What it looks like" (e.g., has a beak, colorful feathers, clawed toes). Finally, students were instructed to generate two or more details for each category. If students needed help, they were provided assistance by the teacher.

[FIGURE 1 OMITTED]

On the following day, students wrote their papers. The teachers instructed students to write a paragraph for each of the categories in their maps. They emphasized that the first part of the papers should provide an introduction to the topic in a way that would get readers interested in the topic. An example of an introduction for the parrot topic was presented and modeled by teachers, who also solicited several examples of introductions, which were recorded on the board.

The teachers then guided students in how to write the rest of their papers. The teachers' directions to students were as follows: "Once you have introduced your topic, write the rest of your paper. Each time you see a new category in your map, write a new paragraph. Begin each paragraph with a topic sentence that introduces the category." The teachers then guided students in generating a topic sentence and details that were relevant for one of the categories in the map. As the teachers provided these examples, they were recorded on the board. Hence, they actually modeled the desired structure and language formats.

The teachers then summarized the general organizational structure for the written paper by instructing students to generalize the composition strategy. Four writing tactics based on a study by De La Paz and Graham (2002) were explained and posted in the classroom: (a) Write an introductory paragraph (use an attention getter, pose a series of questions, introduce the topics and categories in your map); (b) add a body paragraph for each category in the map (introduce each category with a topic sentence, use keywords and details from your map); (c) write a conclusion sentence; and (d) end stories with a conclusion paragraph or sentence. Finally, the teachers reminded students to follow the conventions for writing good papers, including the inclusion of capital letters of the first letter of the first word in each sentence and a period at the end of the sentence.

TELE-Web-supported condition. The identical instructions and processes were followed by the teachers and students in the TELE-Web group. The primary difference between the two conditions was that TELE-Web teachers and students had access to the mapping tools and scaffolds available in TELE-Web, and the teacher guided and modeled the process using the TELE-Web software, including the mapping tools (concept map) and scaffolded writing tools (supported report).

1. Mapping tools. In the TELE-Web condition the teachers assigned one of the TELE-Web mapping tools: "Concept map" or "Idea organizer." An example of the concept map is shown in Figure 2. The two mapping tools were similar in their focus on providing opportunities for students to generate details that corresponded to the information categories of the animal report. However, the concept map offered a more dynamic interface that allowed students to click to add ideas (details), and then they would drag the details to fill out the animal categories. To signify the relationship among the various categories of ideas, the details generated by students changed colors as they were moved to match the color of the subordinate category. On the other hand, the idea organizer offered more static fields for text entry, which meant that students had to decide in advance in what animal category to type their various details.

[FIGURE 2 OMITTED]

As in the paper-and-pencil condition, the categories appeared in the following category boxes: (a) What is the pet, (b) Why it makes a good pet, (c) what it looks like, (d) what it eats, (e) what it does, and (f) how to care for the pet. For the category "What It Looks Like," the teachers modeled how to generate details for one category in the map, following a teaching procedure that was virtually identical to that used in the paper-and-pencil condition. The primary difference was that the teachers modeled how to generate and move details to their respective categories using the click-and-drag feature of TELE-Web to complete the organizer.

2. Writing using supported report. The second day, teachers guided students in writing their informational papers using a template that scaffolded writing performance through procedural facilitators using a writing template known as "supported report" (see Figure 3).

[FIGURE 3 OMITTED]

Again, the TELE-Web teachers followed a teaching process identical to that used in the paper-and-pencil condition. That is, they first modeled how to transform the ideas related to the category "What it looks like" into a topical paper by generating an introduction to the topic and related details. However, TELE-Web supported report offered several scaffolds to prompt students' attention to the local organization and signaling of their ideas. First, there were printed reminders of the teachers' oral instructions. The printed directions appeared at the top of the screen to reinforce the teachers' oral instructions. The directions read as follows: "Many people have a pet, but some people do not have a pet. Pick an animal. Write a paper for someone who does not know anything about that pet. Use your map. Add a paragraph for each part of your map. Introduce each paragraph with a topic sentence, and include at least two to three details. Write a concluding sentence. Remember to begin each sentence with a capital letter."

To help students stage their ideas, a topic sentence box was provided at the top of the computer screen (see Figure 3) to prompt students to generate a "topic sentence" for their texts. Immediately following the topic sentence box was a supporting details box intended to cue students to compose details that adhered to their chosen topic. Each time students chose to Add Paragraph, new topic sentence and supporting details text boxes appeared. Finally, a concluding sentence box was provided near the bottom of the computer screen, in which students could compose a summary statement to wrap up or conclude their texts.

In addition, rather than provide prompts through a poster, TELE-Web allowed students to see teacher prompts, which were brief reminders prepared by teachers about the topical content, genre-related questions, strategies, or key words of a particular text structure. These teacher prompts appeared in a pop-up window accessed by students clicking on the word Prompts (see Figure 2). The prompts disappeared when students returned to the composing area of an activity, thereby offering students a temporary rather than a permanent scaffold. In this study, the teacher provided the following genre-related prompts: "Did I tell: (a) what is the pet? (b) what the pet looks like? (c) what the pet eats? (d) what the pet does? (e) how to care for the pet? (f) why is it a good pet? (g) use topic sentences? and (h) use concluding sentences?"

Finally, in addition to the teacher scaffolds that highlighted the topic and structural features of the text structure, three additional features of the TELE-Web environment allowed students to access online support. These were mediated and controlled by students. First, TELE-Web allowed students to self-employ a spelling checker that matched words against an online dictionary. Second, students could check their text by having the computer read back their texts using a text-to-speech function. Third, once an assignment was completed, it was submitted to the teacher through the online submission feature of TELE-Web. Teachers could provide feedback and communicate when an assignment was finished, which meant that students could publish their final drafts in TELE-Web. However, during the study, the teacher did not provide written feedback on TELE-Web in order to standardize the feedback procedures across conditions.

Although students were provided with various scaffolds, they still maintained control of the supported report activity. Students determined the length and format of their text by inserting new paragraphs on an as-needed basis. When they chose to insert a new paragraph, new topic sentence and detail sentence boxes appeared. Hence, students made decisions about the nature, quantity, and quality of their texts. Once students completed their written text, the individual structural elements of their texts (topic sentences, detail sentences) were merged by the program into a single unified text. In this manner, students generated the global and local elements of the text structure in the supported report condition, but these scaffolds disappeared when students completed or submitted their texts. Students could see a full version of their final texts in a print-ready form without the prompts or instructions, and this version could be shared with other students.

In summary, the teachers and students in the TELE-Web group followed the same instructions and writing process as the control group. Students participated in the same activities across the two groups, so time was not a deciding factor. They spent approximately 30 minutes in the daily activities, spread out across three to four days. TELE-Web students were familiar with the navigation in the web-based browser, so they did not require additional time for writing their stories, and no more than four hours was spent in either of the instructional conditions.

Scoring protocol. The papers produced by students were scored based upon a primary trait scoring scheme developed by Englert (2003). This rubric incorporated the following major genre-related features marking the quality of a good personal narrative essay: (a) introduction to the paper's topic, (b) introduction to the paper's subtopics and categories, (c) adequate depth of subtopical coverage through the inclusion of relevant details, (d) breadth of content coverage through the inclusion of several subtopics that were fairly well developed, (e) conclusion, and (f) overall organization (introduction, details, and conclusion parts).

Generally, a score of "3" was assigned to students who demonstrated consistent and proficient use of a particular trait. A score of "2" was assigned to students who demonstrated a developing knowledge of the trait, but did not consistently perform it in a proficient way throughout the paper. A score of "1" was assigned to students who demonstrated a limited but emerging knowledge of the trait. Finally, a score of "0" was assigned students who did not show any demonstration of the trait. In addition, the overall holistic quality of the papers was evaluated on a scale from 0-3 points, based on a subjective rating of the paper's aesthetic and organizational quality, as well as the extent to which the individual factors worked together to make an overall impression on the reader (Elliott, Plata, & Zelhart, 1990). These traits are further explained in Table 2.

Knowledge of writing conventions was also examined using a related rubric developed by Englert (2003). The rubic scale, shown in Table 3, offered criteria for rating students' sophistication and developmental abilities in the areas of word consciousness, spelling, punctuation, capitalization, and total words. For word consciousness, the students' words were either spaced to suggest their consciousness of words, or the words were strung together with little or no consciousness of words. This variable was either scored 1 or 0 points. In rating the qualities of students' spelling, punctuation, and capitalization, a score of "3" was assigned for a particular attribute when a student demonstrated a consistent, varied, and fully formed knowledge of the trait (e.g., word consciousness, conventional spellings, punctuation, or capitalization). A score of "2" was assigned when students' performance demonstrated that their knowledge was developmentally progressing, but showed a need for further improvement (e.g., varied uses of the convention or accuracy scores ranging between 50-80%). A score of "1" was assigned when students' performance suggested that knowledge was emerging, but consistency or variety was lacking. Finally, a score of "0" was assigned when students demonstrated little or no knowledge of the trait. In evaluating students' writing productivity, two traits associated with students' total written words and topical categories were evaluated. For total words, a word count using the word counting tool available through Microsoft word was performed. For topical categories, the number of categories that were somewhat developed in the students' paper was totaled (e.g., subtopics that were supported by 2-3 relevant details that were consecutively reported).

To make certain that raters were blind to the experimental condition, the students' papers were typed in advance so that the stories could not be associated with a particular instructional condition. A scorer was trained to a criterion level of performance exceeding 85% reliability, based on agreement with other trained raters in scoring a sample of papers that did not constitute the papers used in this study. Once that criterion was attained, this scorer rated all the papers in the study.

To minimize the possibility of scoring drift or inaccuracies, a second rater randomly selected and scored 33% of the writing samples. The agreement between the two raters for cell-by-cell comparison across subjects ranged from 80%-98%, with an average overall agreement of 95%. Any differences between the raters that exceeded 2 points in the total primary trait scores were resolved through point-by-point comparison and discussion.

ANALYSIS

Students' written texts were scored on six primary traits (see Table 2), and then analyzed using an analysis of covariance (ANCOVA) or multivariate analysis of covariance (MANCOVA) to test the effect of the instructional condition while taking into account the interrelatedness among the variables, and covarying for any pretest differences. If the results of an overall multivariate analysis were significant, the individual univariate F-tests were examined to see which made the greatest contribution to the statistically significant multivariate test.

Posttest holistic and primary trait scores. First the effects of disabilities (LD, non-LD) on the holistic and total primary trait scores were tested. There were 24 students with LD and 11 students with other disabilities. Students' holistic and total primary trait scores were analyzed in a MANCOVA, covarying for pretest differences. The results showed that the effect of disability was not significant, F(2, 30) = 2.453, p = .103. Consequently, subsequent analyses were run with the total pool of subjects entered as a group. For the first analysis, we analyzed the holistic scores. Using ANCOVA with pretest scores as covariates, the results revealed a statistically significant effect for condition, F(1, 34) = 9.276, p = .005, [p.sup.2] = .225. Examination of the adjusted posttest scores showed that the papers of the students in the TELE-Web condition (M = 3.304) were rated higher in holistic quality than the papers of students in the paper-and-pencil condition (M = 2.861). The partial eta squared ([[eta].sup.2]) represents the proportion (23%) of the variance that was attributable to the condition. This effect size was considered moderately large (Cohen, 1988).

Second, we analyzed the genre-related primary trait scores in a simultaneous analysis using MANCOVA and covarying for the pretest scores. Again, the results revealed a statistically significant overall multivariate F-ratio for condition with a large effect size, F(6, 22) = 5.054, p = .002, [p.sup.2] =.580. When we examined the univariate F-ratios to determine the location of the significant effects, we found that five of the six dependent variables made a statistically significant contribution, and the sixth variable was approaching statistical significance. These results are shown in the top half of Table 4. An examination of the adjusted posttest means for the five significant dependent variables revealed that students in the TELE-Web condition wrote clearer introductions to their papers and clearer introductions to their categories, and their breadth of categorical development was more extensive than was students' in the comparison condition. Furthermore, their papers were more likely to contain a conclusion, and were more likely to be organized by including at least two of the three general parts of a paper (beginning or introduction, body paragraphs, conclusion). Finally, the papers were more likely to exhibit depth of topical development within categories. Overall, these ratings indicated that the substance of the TELE-Web students' papers was better organized to produce a coherent text that contained subtopics that were signaled and buttressed with more relevant facts and details. The effect sizes associated with these significant factors were all in the moderate ranges, with the exception of introductions to categories, which produced a large effect size.

Conventional knowledge. Second, we analyzed the posttest scores on the conventional traits associated with basic writing skills. We had not expected that differences would emerge for the conventional factors; however, the MANCOVA for condition was significant, F(6, 22) = 3.837, p = .009, [p.sup.2] = .511. The examination of the individual traits that comprised this multivariate factor indicated that two dependent variables contributed to the overall statistically significant multivariate, including students' use of punctuation and their total words. These results are shown in the bottom half of Table 4. As illustrated, the effect sizes associated with these variables were moderate. In each case, the differences favored students in the TELE-Web condition. For example, students in the TELE-Web condition averaged 170 words per story whereas students in the paper-and-pencil condition averaged 122 words.

Story Compositions

An examination of the students' compositions indicated that all students tended to improve from the baseline to posttest assessments, with large improvements attributable to the introduction of the graphic organizers. To illustrate the differences in the conditions, we selected stories that typify some of the changes that were observed. Table 5 contains the pretest and posttest stories of Leo and Marie from the paper-and-pencil and TELE-Web conditions, respectively. These students were selected because they were the students with LD who most clearly matched the average performance on the pretest and posttest of students in their respective instructional conditions. Moreover, both students performed similarly on their pretests. Using the word count feature of Microsoft Word, Leo's story was 17 words, whereas Marie's pretest story totaled 23 words. In the total primary trait scores, both received a score of 6 points. Respectively, Leo and Jamie were assigned 7 and 8 points on their total convention scores on the pretest. In comparison, the mean scores of the experimental and control groups on their total conventions were 7 and 8 points, respectively. The experimental and control groups' primary trait mean scores were 10 and 9, respectively. Thus, the two students were fairly representative of the performance of their focal group, and more important, the students were fairly similar to each other at the time of the pretest.

From pretest to posttest, Leo and Marie showed benefits from the presentation of graphic organizers combined with the writing instruction. Both students demonstrated an inadequately developed knowledge of text structures at the time of the pretest. For example, both students provided limited facts about their animal (e.g., rooster crows; pig lies in the mud), and they were unable to elaborate on their subtopical presentation with sufficient details to lend breadth and depth to their topics. Marie added only a single detail about what pigs eat (mud) and mentioned their babies. Neither student was able to introduce their subtopics with main idea sentences or to elaborate on their topics with any topical breadth or depth.

With the organizer and instruction, on the other hand, they showed growing success. On the posttest, Leo began to introduce some of his subtopics with introductory sentences ("This is a paper about dogs, You will need to care for pet dogs"). Within these subtopical areas, he adhered to the named topics and began to elaborate on the subtopics with details that partly satisfied the predicted local topical structure.

Marie, the student in the TELE-Web condition, exhibited greater improvement from pretest to posttest. In her posttest story, she used introductions, setting forth the topic at the start of her paper with an introduction that more clearly staged the main topic ("What is a dog? Dog are fun to have."). Similarly, she constructed subtopical introductions throughout her paper that prefaced the subsequent text in highly predictable ways (e.g., "Dog come in dffite (different) colores (colors)." ... "Dogs like to eat" ... "You have to tack (take) care of your pet"). She demonstrates five identifiable chunks or categories of subtopical information, which lend breadth to her paper. Furthermore, the details she provided consistently adhered to each of these subtopics, which ensured that the text progression was tight and coherent. For example, in her paragraph about dog care, she states: "You need to fead your pet. You need to love your pet. You(r) pet need some toys to play with. You need to tack your pet to the vet when it is sick. You need to walk your dog some time. Your dog will need a leash." Finally, Marie provided a concluding paragraph that showed the potential to wrap-up her paper ("Dog are good pet ... Dog are fun to have"). Although Marie interrupted this ending with facts about her own dog named Bear, she had begun to exhibit an awareness of the need to provide closure to her paper. The properties of her text aligned with the scaffolds and cognitive bootstrapping available through TELE-Web, inasmuch as she capitalized on the availability of the psychological tools to help her construct introductions, elaborate on her topic through relevant detail sentences, and write a concluding paragraph that wrapped up the topic and subtopics of her paper.

Given the scaffolding condition of TELE-Web, Marie became more successful in accomplishing a wave-like action through her text as recommended by Halliday and Martin (1993), that is, she constructed introductory statements that were embellished through relevant details, and she culminated her topical presentations with concluding sentences. Her text showed greater predictability and displayed a greater maturity from her pretest to posttest. She had begun to demonstrate her growing involvement and mastery in implementing the writing tools associated with composing well-formed expository texts.

DISCUSSION

This study examined the potential benefits of web-based scaffolding on the writing performance of students with disabilities. The findings suggested that computer-supported environments could mediate and scaffold their performance. By making writing tools and text structure strategies available during text construction, students were prompted to incorporate the various properties associated with well-organized texts. Students were more likely to include more categories related to their topics and to organize their texts globally by generating a macrostructure consisting of at least two parts that corresponded to the paper's introduction, body of related details, and/or conclusion. The organizational quality and structure of students' stories was affected, as was their overall production.

Jonassen and Carr (2000) suggest that cognitive tools are devices that support students' cognitive processes by bolstering some of the performative aspects of a task, as well as by freeing students' memories for higher-order thinking, or by reminding and prompting students to apply strategies in the context of problem solving. TELE-Web seemed to offer several cognitive tools that served as procedural facilitators. These included an explicit spatial representation or template of the text structure genre, which provided a cognitive anchor on which students could organize and stage their ideas and information. For example, when students were writing, they were prompted to generate topic introductions and relevant details. The presence of this mediating structure meant that students were not required to hold the text structure in memory since the organizational pattern was modeled and guided by the software. Consequently, information that otherwise might be omitted in paper-and-pencil tasks was present or developed with the TELE-Web scaffolds. In this respect, TELE-Web seemed to allow students to offload some of the strategic requirements of the writing process onto the machine (Sugrue, 2000). In collaboration with the mediational environments of a technology, students with disabilities were guided to perform at levels in advance of their performance in solitary and independent writing conditions.

It must be recognized, too, that the effects obtained for the TELE-Web condition surpassed those achieved by students receiving another powerful instructional intervention, namely, graphic organizers. For example, Leo's posttest story illustrates the benefits that resulted when students had access to graphic organizers that helped them to generate and organize information in ways that are consistent with well-structured expository texts. Typically, the effect sizes realized through the employment of graphic organizers are large in magnitude (d = 0.81-1.69) (Kim et al., 2005). Yet, TELE-Web offered dividends beyond those realized when graphic organizers were employed, as represented by the more substantial gains demonstrated by students in the TELE-Web condition. In support of this hypothesis, we found that TELE-Web students increased their total primary trait scores from pretest (M=6.25) to posttest (M=14.05) by 135%, whereas the students in the paper-and-pencil condition improved their pretest (M=7.33) to posttest (M=11.20) scores by only 57%. In sum, the TELE-Web students were more likely to employ the various writing tools that had been introduced as part of a cognitive process of text development. Technologies could advance what students with disabilities produced by providing scaffolds at the point of intellectual challenge (Wells, 1999).

The results suggested that TELE-Web represents a promising vehicle for delivering procedural facilitators that help students transform their organized ideas into coherent and well-signaled written texts. Like a cognitive anchor or "training wheels" of a symbolic sort, TELE-Web helped students slot and frame their introductions based on their maps, elaborate on their body paragraphs with details, and incorporate conclusions in their desired locations (see Holland, Lachiotte, Skinner, & Cain, 1998). Through the provision of these discourse-level devices, TELE-Web offered pedagogical support that enabled students to connect their newly emerging knowledge to the relevant textual sites where such knowledge might be employed (Greer et al., 2000). Hence, the intellectual work of writing well-formed texts was distributed between the student and the technology in the midst of a cognitive process that was otherwise unmediated except through the private or solitary resources of the student (Claxton, 2002; Greer et al., 2000). In this respect, TELE-Web served as a type of social actor that assisted performance by pointing writers in the direction of how they might stage the text, and ultimately, enhanced their capacity to express themselves in a manner consistent with well-organized text forms (Anderson-Inman et al., 1996). This finding suggests that a new generation of Internet-based software could be designed to support aspects of students' thinking process, and to proffer structures that helped organize their texts, resulting in more expert-like performance (Sugrue, 2000; White, Shimoda, & Frederiksen, 2000).

Further research needs to be conducted on the long-term effect of TELE-Web on writing students' performance. Ultimately, psychological tools and strategies must be internalized by writers to help them organize their writing performance in order to claim that scaffolds have accomplished their intended purpose (Kozulin, 2003). Our research on the long-term efficacy of TELE-Web (Englert et al., 2004) suggests a relationship between TELE-Web usage and learning. That is, teachers who were frequent users of the scaffolding potentials of TELE-Web produced students whose texts were rated significantly higher in quality, and that more fully incorporated the genre-related properties of well-structured texts. Further, their students showed an increasing ability to organize and introduce their subtopical presentations of information, as well as to expand on their presentations through the presence of relevant details that offered subtopical depth and breadth. Conversely, teachers who were infrequent or nonusers of the TELE-Web software produced writers who performed less proficiently on these targeted dimensions of writing. These findings suggest a connection between technology implementation and student achievement, but, again, the findings must be validated by other researchers to determine the role of technology as a mediating factor in advancing students' writing performance.

Research also needs to be conducted to answer the question of how TELE-Web functioned to scaffold writing performance. We did not unpack the specific contributions of procedural facilitators used in mapping and procedural facilitators used in drafting. Instead, we chose to incorporate several best practices in the design of a multi-component intervention based on the instructional literature in writing, which pointed to the importance of employing both organizational frameworks and writing scaffolds in order to improve writing performance. While we have conducted prior research on the efficacy of TELE-Web procedural facilitators during writing in prior research (Englert et al., 2004, 2005), we have not focused on the contribution of the mapping tools available through TELE-Web versus paper-and-pencil environments. Although we are doubtful that the differences between the two will prove significant, we recognize that such research is important before firm conclusions about the relative impact of procedural facilitators during the mapping and writing process can be drawn. It is beyond the scope of the present research to address this question.

Implications for Practice

Several aspects of this research warrant deeper consideration for teaching practice. First, although TELE-Web carried forward some cognitive resources to students with disabilities, it is important to realize that TELE-Web did not teach. We believe that TELE-Web works best when it is combined with effective instruction by teachers--in this case, by teachers' attention to the underlying text structure through their modeling, as well as through their demonstration of the language system and practices associated with the performance of the various aspects of the writing task (e.g., writing topic sentences, body paragraphs, concluding sentence, keywords). Hence, teachers cannot implement writing technologies and expect students to induce and internalize the underlying principles and text structures without guided instruction that makes explicit and visible the arrangement of ideas in texts as well. Likewise, teachers might select and employ technologies to prompt and extend their instructional influence while students are writing, but instruction must complement the technologies to advance students' metacognitive knowledge of the strategies and the procedural facilitators that might be internalized and self-employed to mediate one's writing performance.

Second, technology scaffolds must be tailored to the individual user and instructional histories of students. Students require developmentally appropriate mediational prompts and language whose function and meaning are understood (Claxton, 2002). Thus, the technology must be calibrated to the instruction, and vice versa, to connect the emerging knowledge developed in the teaching context to their realization in the written forms and structures elicited in the technology context. Unless there is an alignment between the two contexts, the application and proceduralization of developmental knowledge will fail to be fully realized (Chaiklin, 2003; Baker et al., 2002; Sugrue, 2000). The advantage of a software like TELE-Web is that it allows teachers to customize the language and prompts to match the language and strategies of the instructional curriculum to ensure that students transfer their knowledge from the teaching context to the writing context. However, teachers must evaluate each piece of software to determine its congruence to their curriculum and make conscious instructional adjustments to bridge any gaps that exist between the instructional technology, the literacy curriculum, and the user.

Third, a technology that provides procedural facilitators and prompts is likely to be influential in eliciting writing strategies that are "under development" rather than those that are fully mastered. On the one hand, many students with LD benefited from the TELE-Web scaffolds that guided them through strategy or genre-employment steps that were unfamiliar or novel. The mere availability of signs and posters that were hung throughout the classroom did not ensure that students accessed and used them as psychological tools (Kozulin, 2003). However, TELE-Web did increase the likelihood that students incorporated the targeted language devices in their writing because it was embedded in the fabric of the writing process rather than standing apart or outside of the act of writing. On the other hand, it would be a serious mistake to employ instructional scaffolds or technologies uniformly for all students, regardless of their developmental levels. For example, students who have already internalized a strong sense of text structure and authors' voice are likely to be constrained and impeded by the imposition of attention to these elements while writing. Technology cannot be viewed as being one-size-fits-all, inasmuch as scaffolds need to be added or removed in developmentally appropriate ways so that cognitive tools lie ahead of particular students' cognitive development while not outdistancing the reach of the lowest-performing participant nor impeding the automaticity of the highest-performing participant (Karpov, 2003). Falling short of that goal will threaten the potential development of students.

In terms of the application of technologies like TELE-Web to the instruction of students with LD, there is great promise. Students with LD generally have difficulties in recognizing text structure, applying memory and rehearsal strategies, and spontaneously employing cognitive strategies (Baker et al., 2002; Englert & Thomas, 1987; Graham, 2006; Graham & Harris, 1989; Wong, 1978, 1980, 1997). In addition, they have well-documented difficulties in the area of metacognition that further reduce the likelihood that they will self-instruct, monitor, or self-regulate their performance as they plan, organize, compose, and revise their ideas (Graham & Harris, 2003; Pressley, Brown, El-Dinary, & Afflerbach, 1995; Stone, 1998; Troia, 2006; Wong, 1997; Wong et al., 1996; Wong & Jones, 1982). Research also shows that the provision of prompts and scaffolds can be gainfully employed to remind students with LD to apply their knowledge and to use cognitive routines that are relevant to the task (Baker et al., 2002; De La Paz, 1997, 1999; Stone, 1998; Wong, 1978, 1980, 1997). In this respect, technologies like TELE-Web might be employed to bridge the gap between students' non-strategic approaches to writing and proficient strategic performance by helping students to recognize and organize information into well-connected patterns, by reminding them to monitor their performance, and by prompting students to apply strategies or procedures that otherwise might lie inert. Given the anchoring support of such technologies, it seems likely that students with LD can be helped to generate written papers that are better organized and more reflective of their developmental potential. The fact that TELE-Web maintains the active role of students as the task manager who makes decisions and controls the writing process is another important feature of a technology application that reflects the need to be sensitive to maintaining the metacognitive character of what it means to be a good writer.

In summary, the scaffolded conditions of TELE-Web seemed to support writing performance by prompting and guiding the cognitive work of generating, organizing, and producing ideas. Technology did serve a mediating function that helped students with disabilities perform in a qualitatively superior way (Daniels, 2001). The integration of technology scaffolds with guided instruction produced the highest levels of writing performance. Using such procedural supports, writers were better able to coordinate, elaborate, and communicate their ideas in expository texts (Baker et al., 2002). Thus, TELE-Web scaffolds accommodated gaps in learning, knowledge, and performance, and ultimately, advanced the level attained by students with disabilities when they wrote in collaboration with the technology. Further, this study highlights the potential role of online environments in supporting learners in a customizable setting. Strategically supported web-based environments can be designed to offer cognitive resources to students with learning difficulties, while remaining economical in terms of their universal access by teachers and students within and across school districts. The possibilities of web-based environments to support and mediate the performance of students with writing challenges make this a fruitful and exciting area that warrants further initiatives.

REFERENCES

Anderson-Inman, L., Knox-Quinn, C., & Horney, M. A. (1996). Computer-based study strategies for students with learning disabilities: Individual differences associated with adoption level. Journal of Learning Disabilities, 29, 461-484.

Anderson-Inman, L., Redekopp, R., & Adams, V. (1992). Electronic studying: Using computer-based outlining programs as study tools. Reading and Writing Quarterly: Overcoming Learning Difficulties, 8, 337-358.

Anderson-Inman, L., & Tenny, J. (1989). Electronic studying: Information organizers to help students study "better," not "harder"--Part 1. The Computing Teacher, 16(8), 33-36.

Anderson-Inman, L., & Zeitz, L. (1994). Beyond notecards: Synthesizing with electronic study tools. Computing Teacher, 21, 21-25.

Baker, S., Gersten, R., & Graham, S. (2003). Teaching expressive writing to students with learning disabilities: Research-based applications and examples. Journal of Learning Disabilities, 36(2), 109-123.

Baker, S., Gersten, R., & Scanlon, D. (2002). Procedural facilitators and cognitive strategies: Tool for unraveling the mysteries of comprehension and the writing process, and for providing meaningful access to the general curriculum. Learning Disabilities: Research & Practice, 17(1), 65-77.

Bereiter, C., & Scardamalia, M. (1985). Cognitive coping strategies and the problem of "inert knowledge." In S. F. Chipman, J. W. Segal, & R. Glaser (Eds.), Thinking and learning skills: Research and open questions (vol. 2, pp. 65-80). Hillsdale, NJ: Lawrence Erlbaum.

Berkowitz, S. (1986). Effects of instruction in text organization on sixth-grade students' memory for expository reading. Reading Research Quarterly, 21(2), 161-178.

Chaiklin, S. (2003). The zone of proximal development in Vygotsky's analysis of learning and instruction. In A. Kozulin, B. Gindis, V. Ageyev, & S. M. Miller (Eds.), Vygotsky's educational theory in cultural context (pp. 39-64). New York: Cambridge University Press.

Claxton, G. (2002). Education for the learning age: A sociocultural approach to learning to learn. In G. Wells & G. Claxton (Eds.), Learning for life in the 21st century: Sociocultural perspectives on the future of education (pp. 21-33). Malden, MA: Blackwell Pub.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

Daniels, H. (2001). Vygotsky & pedagogy. New York: Routledge.

De La Paz, S. (1997). Strategy instruction in planning: Teaching students with learning and writing disabilities to compose persuasive and expository essays. Learning Disability Quarterly, 20(3), 227-248.

De La Paz, S. (1999). Teaching writing strategies and self-regulation procedures to middle school students with learning disabilities. Focus on Exceptional Children, 31(5), 1-16.

De La Paz, S., & Graham, S. (2002). Explicitly teaching strategies, skills, and knowledge: Writing instruction in middle school classrooms. Journal of Educational Psychology, 94, 687-698.

Elliot, N., Plata, M., & Zelhart, P. (1990). A program development handbook for the holistic assessment of writing. Lanham, MD: UP of America.

Englert, C. S. (2003). Scoring rubrics for evaluating elementary students' written performance. Unpublished paper, Michigan State University.

Englert, C. S., Hiebert, E. H., & Stewart, S. R. (1988). Detection and correction of inconsistencies in the monitoring of expository prose. Journal of Educational Research, 81, 221-227.

Englert, C. S., Manalo, M., & Zhao, Y. (2004). "I can do it better on the computer." The effects of technology-enabled scaffolding on emergent writers' personal narrative composition. Journal of Special Education Technology, 19(1), 1-15.

Englert, C. S., & Mariage, T. (2003). The sociocultural model in special education interventions: Apprenticing students in higher-order thinking. In L. Swanson, K. R. Harris, & S. Graham (Eds.), The handbook of learning disabilities (pp. 450-467). New York: Guilford Press.

Englert, C. S., & Raphael, T. E. (1988). Constructing well-formed prose: Process, structure and metacognitive knowledge. Exceptional Children, 54, 18-25.

Englert, C. S., Raphael, T. E., Anderson, L. M., Anthony, H. M., Stevens, D. D., & Fear, K. L. (1991). Making writing strategies and self-talk visible: Cognitive strategy instruction in writing in regular and special education classrooms. American Educational Research Journal, 28, 337-372.

Englert, C. S., Raphael, T. E., Fear, K. L., & Anderson, L. M. (1988). Students' metacognitive knowledge about how to write informational texts. Learning Disability Quarterly, 11, 18-46.

Englert, C. S., & Thomas, C. C. (1987). Sensitivity to text structure in reading and writing: A comparison between learning disabled and non-learning disabled students. Learning Disability Quarterly, 10, 93-105.

Englert, C. S., Wu, X., & Zhao, Y. (2005). Cognitive tools for writing: Scaffolding the performance of students through technology. Learning Disabilities Research & Practice, 21(3), 184-198.

Englert, C. S., & Zhao, Y. (1996). Technology enhanced learning environments on the web: TELE-web. East Lansing: Michigan State University.

Fitzgerald, J., & Teasley, A. B. (1986). Enhancing children's writing through instruction in narrative structure. Journal of Educational Psychology, 78, 424-432.

Gersten, R., & Baker, S. (2001). Teaching expressive writing to students with learning disabilities: A meta-analysis. Elementary School Journal, 101(3), 251-272.

Graham, S. (1990). The role of production factors in learning disabled students' compositions. Journal of Educational Psychology, 82, 781-791.

Graham, S. (2006). Strategy instruction and the teaching of writing: A meta-analysis. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (pp. 187-207). New York: Guilford Press.

Graham, S., & Harris, K. R. (1989). Cognitive training implications for written language. In J. Hughes & R. Hall (Eds.), Cognitive-behavioral psychology in the schools: A comprehensive handbook (pp. 247-279). New York: Guilford Press.

Graham, S., & Harris, K. R. (2003). Students with learning disabilities and the process of writing: A meta-analysis of SRSD studies. In L. Swanson, K. R. Harris, & S. Graham (Eds.), Handbook of learning disabilities (pp. 323-344). New York: Guilford.

Graham, S., & Harris, K. R., MacArthur, C., & Schwartz, S. (1991). Writing and writing instruction with students with learning disabilities: A review of a program of research. Learning Disability Quarterly, 14, 89-114.

Greer, J., McCalla, G., Cooke, J., Collins, J., Kumar, V., Bishop, A., & Vassileva, V. (2000). Integrating cognitive goals for peer help: The intelligent intranet peer help-desk project. In S. P. Lajoie (Ed.), Computers as cognitive tools: No more walls (pp. 69-96). Mahwah, NJ: Lawrence Erlbaum.

Halliday, M.A.K., & Martin, J. R (1993). Writing science: Literacy and discursive power. Pittsburgh, PA: University of Pittsburgh.

Harris, K. R., Graham, S., & Mason, L. H. (2003). Self-regulated strategy development in the classroom: Part of a balanced approach to writing instruction for students with disabilities. Focus on Exceptional Children, 35(7), 1-16.

Holland, D., Lachiotte, W., Skinner, D., Cain, C. (1998). Identity and agency in cultural worlds. Cambridge, MA: Harvard University Press.

Idol, L. (1987). Group story mapping: A comprehension strategy for both skilled and unskilled readers. Journal of Learning Disabilities, 20, 196-205.

Idol, L., & Croll, V. J. (1987). Story-mapping training as a means of improving reading comprehension. Learning Disability Quarterly, 10, 214-229.

Inspiration Software, Inc. (1994). Inspiration 4.0. [Computer software]. Portland, OR: Author.

Jonassen, D. H., & Carr, C. S. (2000). Mindtools: Affording multiple knowledge representations for learning. In S. P. Lajoie (Ed.), Computers as cognitive tools: No more walls (pp. 165-196). Mahwah, NJ: Erlbaum.

Kamberelis, G. (1999). Genre development and learning: Children writing stories, science reports, and poems. Research in the Teaching of English, 33, 403-463.

Karpov, Y. V. (2003). Development through the lifespan: A neo-Vygotskian approach. In A. Kozulin, B. Gindis, V. S. Ageyev, & S. M. Miller, (Eds.), Vygotsky's educational theory in cultural context (pp. 138-155). New York: Cambridge University Press.

Kim, A., Vaughn, S., Wanzek, J., & Wei, S. (2004). Graphic organizers and their effects on the reading comprehension of students with LD: A synthesis of research. Journal of Learning Disabilities, 37, 105-118.

Kozulin, A. (2003). Pyschological tools and mediated learning. In A. Kozulin, B. Gindis, V. S. Ageyev, & S. M. Miller (Eds.), Vygotsky's educational theory in cultural context (pp. 15-38). New York: Cambridge University Press.

Meyer, B.J.F. (1975) The organization of prose and its effects on memory. Amsterdam: North Holland.

Meyer, B.J.F. (1977). The structure of prose: Effects on learning and memory and implications for educational practice. In R. C. Anderson, R. J. Spiro, & W. E. Montague (Eds.), Schooling and the acquisition of knowledge (pp. 179-200). Hillsdale, NJ: Erlbaum.

Meyer, B.J.F., Brandt, D. M., & Bluth, G.J. (1980). Use of top-level structure in text: Key for reading comprehension of ninth-grade students. Reading Research Quarterly, 16, 72-103.

National Commission on Writing in America's Schools and Colleges. (2003). The neglected "R": The need for a writing revolution. New York: College Entrance Examination Board.

Persky, H. R., Daane, M. C., & Jin, Y. (2003). The nation's report card: Writing 2002. Washington, DC: National Center for Education Statistics.

Pressley, M., Brown, R., El-Dinary, P. B., & Afferbach, P. (1995). The comprehension instruction that students need: Instruction fostering constructively responsive reading. Learning Disabilities Research & Practice, 10, 215-225.

Renaissance Learning. (2002). STAR Reading. Wisconsin Rapids, WI: www.renlearn.com.

Renaissance Learning. (2005). STAR Reading: Computer-adaptive reading test and database. Technical manual. Wisconsin Rapids, WI: Author.

Rummel, N., Levin, J. R., & Woodward, M. M. (2003). Do pictorial mnemonic text-learning aids give students something worth writing about? Journal of Educational Psychology, 95(2), 327-334.

Scanlon, D. J., Duran, G. Z., Reyes, E. I., & Gallego, M. A. (1992). Interactive semantic mapping: An interactive approach to enhancing LD students' content area comprehension. Learning Disabilities Research and Practice, 7(3), 142-146.

Sinatra, R., Beaudry, J., Pizzo, J., & Geisert, G. (1994). Using a computer-based semantic mapping, reading, and writing approach with at-risk fourth graders. Journal of Computing in Childhood Education, 5(1), 93-112.

Sinatra, R. S., Berg, D., & Dunn, R. (1985). Semantic mapping improves reading comprehension of learning disabled students. Teaching Exceptional Children, 17, 310-314.

Singer, B. D., & Bashir, A. S. (2004). Developmental variations in writing composition skills. In C. A. Stone, E. R. Silliman, B. J. Ehren, & K. Apel (Eds.), Handbook of language and literacy: Development and disorders (pp. 559-582). New York: Guilford.

Spivey, N. N. (1997). The constructivist metaphor. Reading, writing, and the making of meaning. New York: Academic Press.

Stone, C. A. (1998). The metaphor of scaffolding: Its utility for the field of learning disabilities. Journal of Learning Disabilities, 31(4), 344-364.

Sugrue, B. (2000). Cognitive approaches to web-based instruction. In S. P. Lajoie (Ed.), Computers as cognitive tools: No more walls (pp. 133-162). Mahwah, NJ: Erlbaum.

Thomas, C. C., Englert, C. S., & Gregg, S. (1987). An analysis of errors and strategies in the expository writing of learning disabled students. Remedial and Special Education, 8, 21-30.

Troia, G. (2006). Writing instruction for students with learning disabilities. In C. A. MacArthur, S. Graham, & J. Fitzgerald (Eds.), Handbook of writing research (pp. 324-336). New York: Guilford Press.

Troia, G. A., Graham, S., & Harris, K. R. (1999). Teaching students to plan mindfully: Effects on the writing performance of students with learning disabilities. Exceptional Children, 65(2), 235-252.

Vaughn, S., Gersten, R., & Chard, D. J. (2000). The underlying message in LD intervention research: Findings from research syntheses. Exceptional Children, 67(1), 99-114.

Wells, G. (Ed.). (1999). Dialogic inquiry: Toward a sociocultural practice and theory of education. New York: University of Cambridge Press.

White, B. Y., Shimoda, T. A., & Frederiksen, J. R. (2000). Facilitating students' inquiry learning and metacognitive development through modifiable software advisers. In S. P. Lajoie (Ed.), Computers as cognitive tools: No more walls (pp. 97-132). Mahwah. NJ: Lawrence Erlbaum Assoc.

Williams, J. P. (2003). Teaching text structure to improve reading comprehension. In H. L. Swanson, K. R. Harris, & S. Graham (Eds.), Handbook of learning disabilities (pp. 293-305). New York: Guilford Press.

Wong, B.Y.L. (1978). The effects of directive cues on the organization of memory and recall in good and poor readers. Journal of Educational Research, 72, 32-38.

Wong, B.Y.L. (1980). Activating the inactive learner: Use of questions/prompts to enhance comprehension and retention of implied information in learning disabled children. Learning Disability Quarterly, 3, 29-37.

Wong, B.Y.L. (1997). Research on genre-specific strategies for enhancing writing in adolescents with learning disabilities. Learning Disability Quarterly, 20, 140-159.

Wong, B.Y.L., & Berninger, V. W. (2004). Cognitive processes of teachers in implementing composition research in elementary, middle, and high school classrooms. In C. A. Stone, E. R. Silliman, B. J. Ehren, & K. Apel (Eds.), Handbook of language and literacy: development and disorders (pp. 600-624). New York: Guilford.

Wong, B.Y.L., Butler, D. L., Ficzere, S. A., & Kuperis, S. (1996). Teaching low achievers and students with learning disabilities to plan, write, and revise opinion essays. Journal of Learning Disabilities, 20, 197-212.

AUTHORS' NOTES

This research was part of a larger project, Technology Enhanced Learning Environments on the Web (TELE-Web), funded by a grant from the office of Special Education Programs (No. H0327A010012) in the U.S. Department of Education. This research was also partly supported by the Center for the Improvement of Early Reading Achievement, funded by the Office of Educational Research and Improvement. The opinions expressed in this article do not necessarily reflect the position, policy, or endorsement of the U.S. Department of Education.

Please direct correspondence to: Carol Sue Englert, School of Education, Michigan State University, Lansing, MI 48901; [e-mail]

CAROL SUE ENGLERT, Ph.D., is professor, Michigan State University.

YONG ZHAO, Ph.D., is professor, Michigan State University.

KAILONNIE DUNSMORE, Ph.D., is assistant professor, Michigan State University.

NATALIA YEVGENYEVNA COLLINGS, Ph.D., is assistant professor, Michigan State University.

KIMBERLY WOLBERS, is a doctoral student, Michigan State University.

Table 1

Summary of Demographic Information for Participants

                                   TELE-Web Group   Comparison Group
Ethnicity
  Caucasian                              13                6
  African-American                        7                6
  Hispanic                                0                3

Disability
  Learning Disabilities                  11                11
  Cognitive Impairment                    2                 2
  Emotional-Impairment                    2                 2
  Other Disabilities (1)                  5                 0

Age in Years                           10.64              9.64

Reading Grade Level                     1.71              1.49

Writing Performance
  Mean Primary Trait Total Score        6.25              7.33
  Mean Holistic Score                   2.00              2.13
  Mean Total Number of Words           46.00             42.53

(1) Other Disabilities include students with autism spectrum
disorders and students with physical and other health impairments.

Table 2

Rubric for Primary Trait Scoring

                        3-Proficient                2-Developing

Introduction to     * Clear introduction       * Fairly good
the paper's topic     to topic, purpose          introduction to
                      and text structure         topic, but may be a
                      in a well-defined          little vague about
                      sentence that              the purpose, or
                      launches the topic         offers a good detail
                      in a general way,          sentence rather than
                      and that stages the        a good topic-staging
                      details that follow        introduction

Introduction to     * Clear & consistent       * Some Introduction
the category          introduction to all        at Category Level
                      categories                 (main idea-type
                    * Details adhere to the      sentence) but not
                      specific topic sen-        entirely consistent,
                      tence or main              or provides general
                      idea                       rather than specific
                                                 topic introduction

Depth of            * Details are connected    * Some categories are
categories            & consecutive and          well developed and
                      link to categories in      supported by 2-3
                      clear terms                contiguous details
                    * Main ideas are solidly   * May leave some
                      backed with sufficient     question about cate-
                      evidence                   gory, but answers
                                                 most questions with
                                                 details

Breadth of          * Covers nearly all of     * Some categories are
categories            primary categories to      present with 2-3
                      provide coherent,          contiguous details
                      complete                 * Some subtopical
                      information                categories are
                    * Clear & detailed           omitted
                      subtopical coverage
                    * No obvious gaps or
                      extraneous categories

Conclusion          * Clearly stated           * Implicit statement
                      conclusion that wraps      that implies closure,
                      up the subtopical          but does not precisely
                      presentation and goes      sum up the exact
                      back to the original       purpose of the paper
                      topic.                     or the set of prior
                                                 details; author may
                                                 close paper without
                                                 specific connection
                                                 to the topic

Overall             * Includes all three       * Has two of the three
organization          parts of paper, in-        parts (e.g., intro-
                      cluding an introduc-       duction, details,
                      tion to topic, body        and/or conclusion)
                      of paper with 2-3
                      related details, and
                      conclusion

                         1-Emerging                0-Undeveloped

Introduction to     * Use of title only to    * Can't tell the paper's
the paper's topic     communicate the           topic, or the entire
                      event or purpose          paper is associative

Introduction to     * Presents 2-3 related    * No clear introduction
the category          facts that make it        to category, or can't
                      possible for reader to    tell the specific topic
                      infer the category      * Categories and sub-
                    * Consecutive facts         topics are vague or
                      may have some             hard to figure out
                      extraneous details      * List of associative
                                                details

Depth of            * Infrequent              * Associative, random,
categories            development of            disorganized ideas
                      details in categories   * No development
                      (e.g., without 2-3        within categories
                      details)                * Leaves many
                    * Inconsistent effort       unanswered
                      with sufficient details   questions

Breadth of          * Subtopics are vague     * Narrative
categories            or missing              * Associative (hard for
                    * Inadequate subtopical     reader to follow &
                      coverage                  hard to identify
                    * Sketchy                   any category)
                    * Vague categories        * No development
                    * Confusing                 across categories
                                              * All categories are
                                                1-sentence

Conclusion          * A suggestion from       * No concluding
                      the author of the         sentence; uses detail
                      need for closure, but     as final sentence
                      no specific well-
                      defined closing
                      sentence; author may
                      use a primitive device
                      to signal closure

Overall             * Has one of the three    * Not organized or
organization          parts                     associative

Table 3

Rubric for Scoring Writing Conventions

                   3-Proficient              2-Developing

Word
consciousness

Spelling           * Accurate spelling for   * Somewhat accurate
                     at least 80% or more      spelling (50-80% of
                     of the text, which is     the text)
                     easily readable         * Use of invented
                                               spellings preserve
                                               sound-symbol
                                               relationships
                                             * May overgeneralize
                                               spelling rules
                                             * Text is fairly readable

Punctuation        * Accurate use of         * Some uses of
                     punctuation marks         punctuation marks
                   * Varied uses of            but not consistent
                     punctuation (at least     in all circumstances
                     two different types       (less than 90% of
                     of punctuation used).     the time)
                                             * Limited but fairly
                                               accurate use of a
                                               single type of punctu-
                                               ation (e.g., periods)

Capitalization     * Accurate use of         * Consistent uses of
                     capital letters           capitals at the
                   * Varied uses of            beginning of a
                     capitalization that       majority of the
                     imply knowledge of        sentences (e.g., 80%)
                     different rules         * Somewhat
                     (capitals at beginning    developed knowledge
                     of sentences, proper      of capitalization
                     names, etc.)              present, but not
                                               entirely applied to all
                                               situations

                      1-Emerging                0-Undeveloped

Word               * Spacing between         * No spacing
consciousness        all words                 between words

Spelling           * Initial consonants or   * Random letter strings
                     single letters            or pictures, but no
                     represent words,          knowledge of letter-
                     sounds and syllables,     sound correspond-
                     showing some              ences
                     emergent knowledge
                     of letter-sound
                     correspondences

Punctuation        * Infrequent or           * No punctuation
                     inconsistent use of
                     any type of
                     punctuation

Capitalization     * Minimal or              * Random or no
                     inconsistent attempt      capitalization; may
                     to capitalize             only capitalize "I"
                     beginning words of
                     sentences

Table 4

MANCOVA on Writing Measure

                         Genre-Related Primary Traits (d f = 1, 34)

                     Tele-Web
                     Adjusted (1)     Comparison
                     Posttest         Adjusted
Variables              Mean         Posttest Mean   F-value   p-ratio

Intro to Paper         2.156           1.525         6.435    .017 *
Intro to Category      2.755           1.660        31.876    .000 ***
Depth                  2.587           2.084         3.898    .059
Breadth                2.900           2.467         7.224    .012 *
Conclusion             1.523            .636         4.679    .039 *
Organization           2.596           2.207         4.611    .041 *

                      Genre-Related Primary
                       Traits (d f = 1, 34)

                      Partial
Variables              Eta (2)     Power

Intro to Paper        .192         .686
Intro to Category     .541        1.000
Depth                 .126         .478
Breadth               .211         .736
Conclusion            .148         .551
Organization          .146         .544

                             Conventions (d f = 1, 34)

                    Tele-Web
                  Adjusted (1)   Comparison
                    Posttest       Adjusted
Variables             Mean       Posttest Mean   F-value   p-ratio

Word
Consciousness           .937          .950         .030      .864
Spelling               3.636         3.552         .061      .807
Punctuation            2.540         1.946        6.671      .016 *
Capitalization         1.740         1.680         .040      .843
Categories             4.678         4.296        1.956      .173
Total Words          170.728       122.030        5.749      .024 *

                        Conventions
                       (df = 1, 34)

                     Partial
Variables            Eta (2)      Power

Word
Consciousness         .001        .05
Spelling              .002        .076
Punctuation           .198        .395
Capitalization        .001        .079
Categories            .068        .221
Total Words           .176        .428

Adjusted for pretest scores.

* Significant at the .05 level.

*** Significant at the .001 level.

Table 5

Leo and Marie's Stories

                   Paper-and-Pencil Condition

     Leo's Pretest Story              Leo's Posttest Story

                                          Pet Dogs

Farm animal is a rooster.          This is a paper about
A rooster sat on a house.          dogs. Dogs are pets.
And in the morning                 That many people have.
the rooster crows.                 A pet is a friendly
                                   animal that lives
                                   with a family. Some dogs are
                                   medium size.
                                   Some brown and black
                                   You will need to care
                                   for pet dogs. You will
                                   need to Feed it, And
                                   give it a bath. It Some
                                   times they do crazy things.

                 TELE-Web Condition

   Marie's Pretest Story            Marie's Posttest Story

                                            Dogs

Pigs like to roll in               What is a dog? Dog
drat in they are could (called)    are fun to have.
babeys. Mom pig have               Dog come if dffite
babeys by the day. Pig like        (different) coloers Bare is
to eat mudd.                       black and brown. bare
                                   have teeth like other
                                   dog.

                                   Dog like to eat Bear
                                   like to eat bones like
                                   other dog.

                                   You have to tack care
                                   of your pet you need to
                                   fead your pet. you need
                                   to love your pet. Your
                                   pet need some toys to
                                   play with. you need to
                                   tack your pet to the
                                   vet when it is sick. You
                                   need to walk your dog
                                   some time. Your dog
                                   will need a leash.

                                   Dog like to have fun
                                   too Bear like to watch tv.
                                   Bear like to play outsiad
                                   and he like to chasec
                                   us in we fetch the ball
                                   to him. Bear like to run
                                   out said. Bear sitdown
                                   when we tell him to.
                                   Bear like to eat are shoes.

                                   Dog are good pet. Bear
                                   is a good pet. You sould
                                   allwas love your pet.
                                   Dog are fun to have.
                                   Bear is nies sometime.