Archive article #00–09

Variations in Interactive Writing Instruction: A Study in Four Bilingual Special Education Settings

Anne Graves, San Diego State University,
Gene Valles, San Diego State University,
& Robert Rueda, University of Southern California


The body of research on writing instruction has been accumulating for the past ten years (Christenson, Thurlow, Ysseldyke, & McVicar, 1989; De La Paz & Graham, 1997; Englert, Garmon, Manage, Rozendal, Tarrant, & Urba, 1995; Graham & Harris, 1989; Graves, Semmel, & Gerber, 1994). Several studies indicate that writing-as-a-process instruction is effective in improving both the quality and amount written in compositions of students with learning disabilities (Bos, 1991; Englert et. al, 1995; Englert, Raphael, & Manage, 1994; Graham & Harris, 1994; Rapp, 1997; Zaragoza & Vaughn, 1993). The results of additional studies indicate that a combination of writing-as-a-process with strategy instruction is best for improving the quality of compositions (Graham & Harris, 1989; Graves, Montague, & Wong, 1990; MacArthur, Graham, Schwartz, & Schafer, 1995; Montague & Graves, 1993; Wong, Butler, Ficzere, & Kuperis, 1996). In addition, a few studies of writing instruction for special education students who speak a first language other than English indicate similar results (Ruiz, 1995a, 1995b; Ruiz & Figueroa, 1995; Graves, Valles, & Prodor, in preparation).

Ruiz and her colleagues have investigated the effects of the Optimal Learning Environment (OLE) program on literacy development in several studies of individual students with learning disabilities who are English language learners. OLE is a writing-as-a-process writing instruction program based on socio-cultural learning theory (Bos & Fletcher, 1997; Gallimore, Tharp, & Rueda, 1989). The program is structured to provide contextualized, meaningful, and purposeful activities such as students writing their own books and reading those books to other classes. OLE creates an opportunity for language development, reading and writing to enhance learning.

The OLE program includes an integration of oral language, listening, speaking, reading, and writing. One of the goals in OLE is for students to develop communicative competence (Ruiz, 1995a). The learning environment is designed in a way which facilitates this development by including the following elements: 1) interactive journals in which the teacher responds to students' daily entries in writing each day to provide modeling of written dialog, 2) Writer's Workshop based on writing-as-a-process in which students go through planning, drafting, editing, revising, final drafting, and publishing each time they produce a written product, 3) mini-lessons including strategic writing, 4) patterned writing and reading in which students read and copy key phases such as: "Brown bear, brown bear what do you see? I see a red bird looking at me. Red bird, red bird what do you see? I see a yellow duck looking at me" (Carle, 1970), 5) creating text for wordless books, 6) shared reading with predictable text, 7) literature conversation with read-alouds, 8) literature study with response journals, 9) student-made alphabet wall charts, and 10) drop everything and read time (D.E.A.R.). Results indicated that the use of OLE improved the writing performance of these individuals and provided a mechanism for non-writers and readers in English and Spanish to develop improved English and Spanish performances (Ruiz, 1 995a, 1 995b).

The OLE program is much like the Early Literacy Project (ELP) developed by Englert and her colleagues (Englert et al., 1995) for students with learning disabilities, but the latter was not designed for bilingual students. Like OLE, the ELP is also based on socio-cultural theory and includes some of the following elements: thematic units, unsupervised silent reading, partner reading and writing, sharing chair which provides students with opportunities to ask questions and act as informants to peers and teachers, story discussions, author's center for writing-as-a-process projects, and morning news which provides students with a language experience activity in which the teacher writes about students' personal experiences as the student dictates. The ELP includes an emphasis on the following four principles: 1) meaningful and purposeful activities, 2) teaching self-regulated learning, 3) responsive instruction in which students who need assistance may receive direct instruction, strategy instruction, or less explicit instruction, and 4) building communities of learners through collaborative learning and problem-solving among students. Both the OLE program and the ELP program are focused on student experiences and background knowledge each of which provide authentic purposes for reading and writing and the content of literacy activities.

The OLE program is very similar to the ELP in that it is based on creating a contextualized environment for learning. The rationale for such a learning environment may be even more compelling for students who are Spanish speakers with learning disabilities (Gallimore et al., 1989; Graves, 1998). First, Spanish is a language with 99% phonological consistency. Therefore, students are likely not to have the difficulty with decoding that students experience in English. In addition, because English immersion is the predominant model for Spanish speakers in the schools, students who are labeled in special education are often students who have not had Spanish language learning opportunities. If students are to be able to salvage their potential for literacy based on native language work, it makes sense that a meaningful and purposeful use of the language would resurrect their motivation and their appreciation for language and literacy (Ruiz & Figueroa, 1995).

The Ruiz studies (Ruiz, 1995a, 1995b; Ruiz & Figueroa, 1995) have contributed to the knowledge base on effective instruction in writing for English language learners with learning disabilities. Ruiz's work supports bilingual education theory and contributes to the knowledge base in bilingual special education. In many teacher preparation programs around the country, those in training to be bilingual special education teachers are taught the techniques used in OLE and emphasis is placed on the socio-cultural context for learning (Baca & Cervantes, 1998; Gallimore et. al., 1989; Ruiz & Figueroa, 1995; Ruiz, Figueroa, Rueda, & Beaumont, 1992).

The Ruiz work and the experiences of many other practitioners in the field (Baca & Cervantes, 1998) have inspired a set of recommendations for practice. Those recommendations typically include the aspects described above which comprise the OLE and ELP curricula. Interns in a teacher training program at a Southern California University were prepared to teach writing using the methodology described above. Each intern had previously earned an elementary teaching credential and experience ranged from 0 to 9 previous years of elementary teaching (case one teacher—4 years; case two teacher—0 years; case three teacher —3 years; and case four teacher—9 years). Each intern was in the second of a two-year internship program in special education when this study began and had no previous experience in special education. This paper provides four case-studies of writing instruction in the classrooms of four interns in bilingual special education settings to determine which methods they employed for teaching writing and how the writing of the bilingual students with learning disabilities changed over time.

The purpose of this study is to learn how the interns applied what they were taught in teacher preparation and to report the data on student progress in each of the four classroom situations. Spanish speaking and bilingual students with learning disabilities are learning to write in all four settings. All four teachers are certifiably bilingual in Spanish and English, however, only one teacher (case one) was a native Spanish speaker the others learned Spanish as a second language. The research includes self-reports of teachers' instructional choices, observations of the teachers and their students, and an analysis of the compositions of the students in each setting at the beginning of a ten-week period, at the end of a ten week period, and in a one year follow-up.

The following research questions were addressed: 1) how are different approaches to interactive writing with bilingual students with learning disabilities implemented? 2) when compared to the beginning point, what are student outcomes on number of words written and quality of compositions after ten weeks? 3) What are student outcomes on number of words written and quality after one year?

Participants and Procedures
From a pool of ten interns in the Bilingual Personnel Preparation Program in Special Education at San Diego State University, four volunteered to participate in this study. Each participating teacher-intern worked at a school in which the majority of the student population was low socioeconomic status demonstrated by 80% or greater of each school population qualifying for "free lunches". In each setting, the teachers included all of their students for the purposes of the study. All students in this study across the four settings were English language learners with learning disabilities according to specific school district labeling processes. All students were tested in both Spanish and English and labeled due to a significant lag in both languages. All students in this study had specific Individualized Education Plan (IEP) goals in written expression. Students ranged from early production to intermediate fluency in English language learning with much greater proficiency in Spanish (Echevarria & Graves, 1998). In three of the classroom cases, all the students spoke Spanish much more fluently than English and students ranged from second to fifth grade. In the sixth grade classroom case, students who spoke Spanish, Tagalog, Vietnamese, Cambodian, or Lao were also more fluent in their native languages than in English.

As a way of triangulating the data from the descriptions of the instruction, teachers agreed to continue teaching as they would normally for a period of ten weeks. Each teacher agreed to submit weekly logs accounting for the number of hours they spent on writing instruction each week, the attendance of students, the teaching techniques used, the assignments in which students participated, and the time students spent involved in writing activities per week. The four teachers agreed to random observations during their writing instruction on four occasions during the ten weeks. Samples of student work were copied by teachers and submitted on two occasions during the ten week period. Observations of each of the teachers were conducted to verify teaching logs, to obtain descriptions of writing instruction in the four settings, and to document examples of questions, corrections, and styles of interaction. Each teacher administered an initial prompt before the ten weeks began and a ten-week prompt at the end of the ten weeks of instruction, which consisted of a picture of cave people throwing spears at giant mastodons taken from the Test of Written Language II (Hammill & Larsen, 1993). Students were asked to write a story about the picture. In all prompts students were given approximately 15 minutes. One year later, teachers were interviewed and they re-administered the same picture prompt which was used in both the initial and ten-week prompts the year before.

Four teacher-interns with four very different bilingual special education settings were studied: (1) a self-contained setting for bilingual students in grades three, four, and five (Case One); (2) a resource program for primarily bilingual students in grades three, four, five, and six (Case Two); (3) a resource program for both Spanish and English speakers in grades two and three (Case Three); (4) a resource program for English-as-a-second-language learners in grade six (Case Four).

Measures for Student Data
Writing samples based on picture prompts were taken on three occasions: January, 1995, March, 1995, and February, 1996. For all the data, words were counted by two researchers and the number of words was recorded separately. Counts were compared and it was determined that 45 out of the 50 stories had different word counts. The researchers were asked to go back and recount the stories for which differences were reported and in all but one story those differences were resolved. The number of words reported differed by one word in one of the stories. Therefore, across the two counters, the reliability was computed by simple percentage at 98% agreement.

The quality of the compositions was measured by the Story Quality Scale (Graves & Montague. 1991) which consists of scores from one to five for coherence, organization, and episodic structure for a possible total score of 15. Based on the scoring criteria, a score of 0 to 4 is considered a weak composition; a score of 4.1 to 8 is considered moderately good; a score of 8.1 to 12 is good; and a score of 12.1 to 15 is excellent. The range is derived from five previous studies including students with and without learning disabilities (Graves et al., 1990; Graves et al., 1994; Graves, Semmel, & Gerber, in preparation; Lewis, Graves, Ashton, & Kieley, 1997; Montague, Graves, and Leavell. 1991). In this study, reliability across two scorers was .86, and reliability has been previously established on all five aforementioned studies ranging from .81 to .88.

The four bilingual special education settings that were studied are:
(1) a self-contained setting for bilingual students in grades three, four, and five (Case One);
(2) a resource program for primarily bilingual students in grades three, four, five, and six (Case Two);
(3) a resource program for both Spanish and English speakers in grades two and three (Case Three);
(4) a resource program for second language learners in grade six (Case Four).

In the first three cases, the primary language of the students is Spanish and students are encouraged to speak in Spanish in the each of the three special education settings. In case four, students are encouraged to use English. For each case description, we report information from observations of teachers and teacher self-reports (see Table 1).

TABLE 1: Summary of Information from the Four Settings
Settings Sex of Teacher Job Status # Students
(pre / post / final)
Grades Age Range Primary Language
(Case 1)
M Intern
1 yr Exp.
11 / 9 / 6 3, 4, 5 8.1-11.4 Spanish
Resource/Gen. Ed.
(Case 2)
F Intern
1 yr Exp.
7 / 7 / 4 3 ,4 ,5 7.5-12.3 Spanish
(Case 3)
F Intern
1 yr Exp.
12 / 12 / 5 2, 3 7.4-9.3 Spanish
(Case 4)
F Intern
1 yr Exp.
12 / 10 / 8 6 11.5-13.2 Spanish, Lao, Vietnamese,

Following the descriptions of each of the four settings, we present analyses of compositions written by students in each of the four settings including data on amount of words written and story quality.

Case One, Mr. Ramirez: Interactive Journals
Mr. Ramirez, a bilingual Spanish/English speaker, taught Spanish/English bilingual students with learning disabilities in a self-contained special education classroom for approximately three hours per day. The students were primarily Spanish speaking from grades three, four, and five with an age range of 8.1 to 11.4 years. The class included 11 students who took the initial prompt, seven boys and four girls. Of the 11, two left school before the ten week prompt leaving seven boys and two girls. In the one-year follow-up, only six boys were left from the original sample. For those six, data were analyzed at each stage of the study to gain information about performance.

The students participated in most school activities and were mainstreamed for various subjects individually. Mr. Ramirez taught a fully integrated curriculum focusing on reading, writing, listening, speaking, functional living skills, and math, often within the same thematic unit. For example, students experienced a helicopter landing at the school during the time they were studying transportation and vehicles. They talked about it, wrote about it in their journals, learned spelling words from their journal entries, created math problems about it, drew pictures of it, and shared their work. Students wrote in their journals daily and the teacher provided written responses to each student daily. Individually, he conferred with each student and required the student to read his or her own work as well as the teacher's response. Students wrote about life experiences, school events, lessons, or stories they read.

Mr. Ramirez reported that his written responses to students' journal entries generally included words that needed correction so as to provide a correct model. He did not correct spelling and grammatical errors overtly, but often focused on an error in formulating a journal response to model correct form. For example, a student wrote: "tengo dos ermano". He wrote back: "Tienes dos hermanos. Esta bueno." The teacher's response modeled the correct spelling and form of the plural for the word "brothers" in Spanish and the use of a capital letter at the beginning of the sentence and a period at the end. Mr. Ramirez reported that students typically began to demonstrate improved spelling as a result of this procedure.

Mr. Ramirez reported interactive journal writing and sharing for approximately 25 minutes a day for ten weeks. Observations of the classroom confirmed this schedule. Teacher-student interactions were observed in which the teacher listened intently to the students' journal entry usually commenting about the content. The teacher would then model for students by writing back immediately and reading with expression the written comments to the students. The students often appeared very motivated to write by quickly approaching the task and smiling or laughing during the process. Students were encouraged to write in Spanish initially in this setting; between the end of the ten-weeks and the follow-up prompt a year later a few students began to write in English.

Case Two. Mrs. Roberts: Optimal Learning Environment (OLE)
Mrs. Roberts, a bilingual Spanish/English speaker, taught Spanish/English bilingual students with learning disabilities in a resource setting. For mornings only, students were in the general education setting for three weeks at a time and then rotated into the resource room for three weeks at a time and so on throughout the year. This pattern was repeated throughout the school year and the composition of each group changed as the students' needs changed. The students were primarily Spanish speaking from grades three, four, five, and six with an average age range of 7.5 to 12.3 years. One girl and six boys took the initial prompt and the ten-week prompt. One year later, however, only four boys were left from the original sample. Data were analyzed at each stage of the study to gain comparative information about these four students.

Mrs. Roberts taught using the Optimal Learning Environment (OLE) program that was described by the teacher as a comprehensive developmental curriculum. The language of instruction in the resource setting was in Spanish. For the three-week periods in the resource room, she included some version of the nine OLE elements described earlier in this paper. Typically, during the first few minutes of each morning the students wrote in their journals for 10 to 15 minutes. She responded to their writing each day with her own written comments.

Students used Writer's Workshop to plan, write, edit, revise, and produce perfect/publishable final copy. Each student's final copy was illustrated and became one page of a class book, which was laminated and kept in the classroom for sharing and reading. One book was about the families of the students in the class. The writer's workshop elements included a focus on preliminary activities such as prewriting, planning, and talking in groups with peers about proposed writing (see D. Graves, 1983 for a full description). Students chose their own topics and wrote books about those topics to share with the class. Students focused on conveying meaning to their peers, sharing their work at each phase. Teachers conferred with students at times to assist them with editing and revising. The final draft for the class book was to be a perfect copy and usually included illustrations.

Daily, students read from children's literature, predictable texts, and anything they wanted during the drop everything and read (D.E.A.R.) time that lasted about 15 minutes. The teacher often read books in Spanish and then had the students read the books in small groups. Popular folktales in Spanish such as Juana, Princesa de la Sal (1935) and Pituso (1954) were read across two or three days. Discussions and written responses ensued. Predictable text from pocket charts, in books, and from the work of other students was also read on a daily basis. For example, " Aranas, aranas por todas partes'. Hay diez aranas en la telarana. Aranas, aranas por todas partes! Hay nueve aranas en la telarana...." (Williams, 1996).

Observations of Mrs. Roberts indicated confirmation of the three weeks in resource followed by three week in general education schedule. The students participated in at least an hour of reading and writing activities in the resource room with sharing, illustrating, and organizing work taking the rest of the time. Observations of the general classroom revealed far less student participation in reading and writing activities. Two observations of general education classrooms indicated approximately 20 minutes of reading and ten minutes of writing one day and 17 minutes of reading and 12 minutes of writing the other. In the general education classrooms students were often required to write in English in the resource room students were encouraged to write in Spanish.

Case Three. Ms. Arquet: Expository Writing-as-a-Process
Ms. Arquet, a bilingual Spanish/English speaker, taught Spanish/English bilingual students with learning disabilities in a pullout resource program in which she taught writing for approximately 90 minutes per week. This group of students spoke Spanish with an emphasis on emerging English. Students were from second and third grade with ages ranging from 7.4 to 9.3 years. Ms. Arquet had a group of five girls and seven boys take the initial prompt and the ten-week prompt. The next year only five boys were left in the sample for the follow-up test. The data from these five students were analyzed at each stage of the study to gain information about performance in which the teacher, Ms. Arquet, taught using an expository Writer's Workshop approach.

The teacher created her own writer's workshop with a focus on expository report writing. As in the other classrooms, this teacher tried to create a fully integrated, comprehensive language arts setting by incorporating reading, listening, talking, and writing into the writing-as-a-process, writer's workshop activities. Students engaged in topic selection within a broad context. For example, if students were studying Africa in their general education classroom, the teacher would allow students to select any African animal and write a report. In the planning phase, the students, with teacher support, gathered information on the animal and engaged in note taking on index cards. Students worked together around a table with the teacher, talking and sharing as they worked. Students wrote first drafts in a notebook from their prepared note cards and the teacher edited their work as they went. As in OLE, students wrote drafts until they had a "perfect copy." Students illustrated their work and the teacher posted the reports in the hall or in the classroom.

Ms. Arquet reported the systematic use of Writer's Workshop by spending the first two weeks on journal writing and sharing, the next three weeks on journal writing and planning, the next three weeks on drafting, editing, and revising and the last week on publishing and disseminating the class book. Observations of her setting confirmed her claims about use of Writer's Workshop and the approximate time students had to practice writing. Students were animated and very verbal during the four observations of this setting. They talked about their animals in Spanish as they gathered information. Students were encouraged to write in Spanish initially in this setting; between the initial prompt and the follow-up students had switched to writing in English.

Case Four. Mrs. Jones: Combination of Approaches
Mrs. Jones, a bilingual Spanish/English speaker, taught bilingual students with learning disabilities in a sixth grade middle school resource program in which students had several periods a day with the resource teacher. The majority of the instruction was in English with an English-as-a-second-language focus because students' native languages included: Tagalog, Vietnamese, Cambodian, Lao, and Spanish. The teacher reported bilingual interactions between the Spanish speaking youngsters and herself only occasionally for clarification. She was unable to provide bilingual mediation to the majority of the class. The students ranged in age from 11.5 to 13.2 years.

Only four of her 12 students spoke Spanish (three girls and one boy); two boys spoke Lao, three boys spoke Cambodian, two boys spoke Vietnamese, and one boy spoke Tagalog. These 12 students took the initial prompt but the Tagalog-speaking boy and one of the Cambodian-speaking boys moved before the ten-week prompt. One year later, the sample had shrunk to eight, three Latinas, two boys who spoke Lao, two boys who spoke Cambodian, and one boy who spoke Vietnamese were left in the sample. Data were analyzed and reported from this sample at each stage of the study to gain information about performance.

Mrs. Jones provided instruction for the students for approximately 250 minutes a week and generally integrated reading, writing, listening, speaking, and math activities. According to her records, students wrote for approximately 60 minutes each week. The teacher typically involved students in planning activities that allowed them to create a visual representation of their thinking before they wrote. For example, she would model webbing or mapping a story and ask students to do the same before they wrote. This teacher gave weekly spelling tests on easily decoded words, but focused on memorization of common words in English which are not spelled according to rules (i.e., gone, done, could, does). Students wrote in journals daily but the teacher was unable to respond personally to every entry. At times her procedure was similar to interactive journals (like in Case One) in that the teacher would respond in writing, but she did not make a regular habit of doing so. She reportedly did eventually read all the entries, however, and used their writing as a way of gathering information about students' writing skills. The students did not know when she would respond to their work, but they knew she would eventually read all of the entries.

During the ten weeks, she reported journal writing for three to five days per week approximating ten minutes per session. For seven of the ten weeks she reported either a brainstorming or a mapping activity of approximately 20 minutes in duration. She reported two editing/revising activities of 20 minutes each and two publishing/sharing times for approximately 20 minutes each across the ten weeks. She required spelling tests once a week on frequently used words and allowed students to practice with partners before the tests. The spelling activity generally lasted 25 minutes per week. She reported a range of 30 to 80 minutes per week for writing.

Quality and Quantity Data from Written Compositions
Student compositions were collected at the beginning of the study, after ten weeks, and at the end of one year of study. Word count means by case (see Table 2) and quality score mean scores by case (see Table 3) were derived from the compositions of students who participated in all three of the written prompting sessions.

TABLE 2: Number of Words Written
Settings PRE POST 1 Year Follow-up
(Case 1)
35.2 41.6 39.0
(Case 2)
22.8 30.0 52.2
(Case 3)
23.0 37.8 42.0
(Case 4)
65.6 44.2 46.5

TABLE 3: Quality Measures
Settings PRE POST 1 Year Follow-up
(Case 1)
3.44 4.33 3.60
(Case 2)
3.36 4.00 7.00
(Case 3)
3.40 3.73 3.91
(Case 4)
4.08 4.54 4.88

Case One. Mr. Ramirez: Interactive Journals.
Minimal differences existed between the average number of words written by six students in this setting either from the initial prompt (m = 35.2) to the ten-week prompt (m = 41.6) and from the ten-week prompt to the one year follow-up (m = 39.0). However, on the quality measure students improved from the "weak" category in the initial prompt (m = 3.44) to the moderately good category in the ten-week prompt (m = 4.33) but the improvements were lost in the one year follow-up prompt (m = 3.60). For example, Luis wrote more words and higher quality work from the initial prompt to the ten-week prompt and while his progress is maintained his work doesn't appear markedly different from the ten-week prompt to the prompt a year later (see Figures 1 and 2).

FIGURE 1: Initial Prompt for Luis in Case One Setting

FIGURE 2: Ten-Week and Follow-up Prompts for Luis in Case One Setting

In the interview after the one year follow-up, Mr. Ramirez reported a change in the service delivery model at his school. During the first ten week phase of the study, students were in a self-contained classroom with him for approximately three hours per morning. The next year, the school switched to a resource model in which students had approximately an hour a day in the special education setting, and he complained about his lack of control over the writing instruction students received. He felt very strongly that the previous model had allowed students to progress more rapidly and was concerned that students had "lost ground." After one year, his writing instruction was limited to two to three times a week with interactive journal opportunities once a week at best. He had been forced to switch most of the children over to English because they were writing in English in the general education settings and Spanish work was not acceptable to most of the teachers. He thought that switching them to English before they had substantial Spanish skills might affect their abilities in written expression, particularly on the follow-up probe.

Case Two. Mrs. Roberts: Optimal Learning Environment (OLE).
The number of words written by the four students in this setting did not appear to differ substantially from the initial prompt (m = 22.8) to ten-week prompt (m = 30.0), but they did write more words from the ten-week prompt to one year follow-up (m = 52.2). On the quality measure students improved from the initial prompt (m = 3.36) to the ten-week prompt (m = 4.00), and improved markedly in quality from the ten-week prompt to the one year follow-up (m = 7.00). The follow-up prompt scores are over 100% improved from the initial prompt scores and students are solidly in the "moderately good" range indicating moderately good organization, coherence, and episodic structure in the stories. For example, Alejandro wrote this quite short, weak piece initially and one year later he wrote a longer substantially improved composition (see Figure 3).

FIGURE 3: Initial and Follow-up Prompts for Ricardo in Case Two Setting

In the one-year follow-up interview, Mrs. Roberts reported the satisfaction she felt with the OLE program. She reported glowing approval from students, teachers, and parents. She was able to exercise the same system during the following year that she had in the previous year, and was certain that students had progressed substantially in their writing. She noted the comprehensive nature of OLE and the rigorous daily requirements for reading and writing in various different approaches.

Case Three. Ms. Arquet: Expository Writing-as-a-Process.
The five students in this setting wrote more words from the initial prompt (m = 23.0) to the ten-week prompt (m = 37.8), and maintained progress but did not write substantially more words from ten-week prompt to the one year follow-up (m = 42.0). On the quality measure, students wrote higher quality stories from the initial prompt (m = 3.40) to the ten-week prompt (m = 3.73) but the scores did not reach the "moderately good" range. Stories in the follow-up prompt were still improving but still did not reach the "moderately good" range (m = 3.91). For example, Arturo wrote a short and fairly weak few sentences in the beginning and wrote an improved, though still weak, piece ten weeks later (see Figure 4).

FIGURE 4: Initial and Ten-week Prompts for Arturo in Case Three Setting

His work one year later was very similar to the ten week sample. He appeared to be breaking through some of his writing difficulties from the beginning to the ten week point but was unable to continue those strides throughout the next year. Arturo's work is indicative of three students in the sample, but the work of the other two is worth noting.

Carlos (see Figure 5) is a student with severe writing difficulties. He consistently produced work through all three prompts that looked like gibberish. When Carlos was asked what his story said he began to provide words which are written above the random letters depicted in Figure 5.

FIGURE 5: Follow-up Prompt for Carlos in Case Three Setting

At the other extreme, Edgardo's work improved dramatically from 30 words to 78 words (see Figure 6). The quality of his work also progressed from the "weak" to the "moderately good" range.

FIGURE 6: Follow-up Prompt for Fernando in Case Three Setting

During the interview after the one year follow-up, Mrs. Arquet indicated that students had worked hard during the year and that she had seen improvements. She reported that she had not been able to continue some of the writing projects just as they had been conducted before due to some changes in scheduling at the school. She wondered if switching to English had adversely affected the progress some of the children had made previously when their writing had been assessed in Spanish. She believed that transitioning to English would yield lower scores at least temporarily until students could strengthen their written language skills in the new language. Naturally she was quite pleased with Edgardo's work and very troubled by Carlos' work, indicating that Carlos probably needed some other types of interventions.

Case Four. Mrs. Jones: Combination of Approaches.
The eight students in this setting wrote more words on the initial prompt (m = 65.6) than on the ten-week prompt (m = 44.2), and did not write significantly more words from the ten-week prompt to the one year follow-up (m = 46.5). Students all wrote in English on all three writing prompts. On the quality measure, students wrote higher quality stories from the initial prompt (m = 4.08) to the ten-week prompt (m = 4.54). All eight students wrote compositions higher in quality for the one-year follow-up but the progress was not substantive given the one year time span (m = 4.89). Pon's work is illustrative of the sixth graders' compositions (see Figures 7 & 8).

FIGURE 7: Initial Prompt for Pon in Case Four Setting

FIGURE 8: Ten-week Prompt for Pon in Case Four Setting

Though the students in this group maintained scores in the low "moderately good" range, the fact that they are sixth graders should be taken into consideration. In previous studies, students with learning disabilities in sixth grade were often able to produce stories in the six and seven range while their general education counterparts almost always produced stories rated between ten and 15 (Graves et al., 1994).

At the end of the one year follow-up period, Mrs. Jones said that many of her students were unmotivated and hated to write. She thought that most of the students had improved from the previous year, however, but wished that they were required to write in more of their classes. She did note that one possible explanation for the decrease in word count from the initial to the ten-week prompt was that as quality improved and students became more precise writers their word volume may have decreased.

After ten weeks, students in each of the four settings demonstrated some improvements in writing quality. Interactive journal writing, Writer's Workshop, OLE, and a combination of journal writing, brainstorming/planning, and spelling practice all yielded improved written compositions after ten weeks. The commonalties among classrooms were that teachers spent at least one hour per week teaching writing, were teaching students to write for real audiences, provided instruction on mechanics, and tried to create a community of learners in the classroom.

After one year, the OLE program (Case Two) results were the only ones which exceeded the progress noted after the ten week period one year earlier. OLE is a formal program with specific requirements for activities and time spent with students each week. Apparently, over the course of the year, the quality of the instruction which students received yielded great strides in their writing. In addition, other literature has indicated the strong effects of the OLE program and this study seems to confirm those earlier studies (Graves et al., in preparation; Ruiz, 1995a, 1995b; Ruiz & Figueroa, 1995).

In the follow-up interviews, the teachers in Case Two (OLE) and Case Four (Sixth Grade: Mixed Approaches) both expressed concerns regarding the Spanish/English issues and whether students were ready to transition to English. Pressure for students in general education to use English in school can force teachers to make instructional decisions that they might not make otherwise. OLE provided a system by which instruction and practice in the native language could continue until the teacher thought that students were ready for the transition to English. However, the students in Case Four did not have any opportunity to express themselves in their native languages and this may have impeded progress in literacy development. This issue may be even more critical for students with learning disabilities than for students in general education, because some of these students may not learn to write and read at all unless their Spanish language skills can be used in the initial literacy process.

Limitations of the Study
This research was designed from the beginning as descriptive though we provided as much quantitative information as possible in our attempt to paint a clearer picture of the settings. This is not a quantitative comparison across cases because it was not designed to yield such information. The study would have been a quantitative one if we could have trained teachers and required them to uphold specific teaching approaches, included more students, observed and collected data from the ten-week prompt to the follow-up prompt, and randomly assigned students to each of the conditions. Perhaps, we could offer more information on the lack of progress in at least three of the cases had we been able to conduct a quantitative study. A descriptive study was conducted because we believe that initial explorations in the area of bilingual writing instruction were appropriate before the more structured quantitative studies could be conducted effectively.

The attrition among participants in the study is also an important limitation. The one year follow-up data was severely limited due to the loss of important data which may have yielded different outcomes. However, we only included those youngsters who participated in every phase of the study.

Educational Implications and Future Research
This type of research is intended to provide information about the state of the art of bilingual special education instruction in writing. Replications with larger sample sizes would be critical in reaching any definitive conclusions. However, a paradigm shift discussed by Rueda and others (Gallimore et al., 1989) has occurred in bilingual special education and many teachers are selecting instructional methods that emphasize interaction, constructivism, and meaningful reading and writing activities. Much additional research is needed to study components of various practices to determine the their strengths and weaknesses.

Issues regarding the continued use of Spanish by students who are labeled with learning disabilities when English is not well developed continue to be a central focus of discussions in the schools (Baca & Cervantes, 1998). This research seems to indicate that Spanish writing practice is helpful to students and future research questions need to be answered regarding when and under what conditions transition to English is appropriate.

Baca, L., & Cervantes, H. (1998). The bilingual special education interface (3rd ed.). Upper Saddle River, NJ: Prentice Hall.

Bos, C. 5. (1991). Reading -writing connections: Using literature as a zone of proximal development for learning. Learning Disabilities Research and Practice, 6, 251-256.

Bos, C. S., & Fletcher, T. V. (1997). Sociocultural considerations in learning disabilities inclusion research: Knowledge gaps and future directions. Learning Disabilities Research and Practice 12 92-99.

Carle, E. (1970). Brown bear. brown bear what do you see? New York, NY: Henry Holt.

Christenson, S., Thurlow, M., Ysseldyke, J., & McVicar, R. (1989). Written language instruction for students with mild handicaps: Is there enough quantity to ensure quality? Learning Disability Quarterly, 12, 219-229.

De La Paz, S., & Graham, 5. (1997). Strategy instruction in planning: Effects on the writing performance and behavior of student with learning difficulties. Exceptional Children, 63, 167-82.

Echevarria, J., & Graves, A. (1998). Sheltered content instruction: Teaching English-language learners with diverse abilities. Boston, MA: Allyn and Bacon.

Englert, C. S., Garmon, A., Manage, T., Rozendal, M., Tarrant, K., & Urba, J. (1995). The early literacy project: Connecting across the literacy curriculum. Learning Disability Quarterly, li8, 253-277.

Englert, C. S., Raphael, T. E., & Manage, T. V. (1994). Developing a school-based discourse for literacy learning: A principled search for understanding. Learning Disability Quarterly, L7, 2-32.

Gallimore, R., Tharp, R., & Rueda, R. (1989). The social context of cognitive functioning in the lives of mildly handicapped persons. London: Falmer Press.

Graham, S., & Harris, K. R. (1989). Component analysis of cognitive strategy instruction: Effects of learning disabled students' compositions and self-efficacy. Journal of Educational Psychology, 81, 353-361.

Graham, S., & Harris, K. R. (1994). Implications of constructivism for teaching writing to students with special needs. Journal of Special Education, 28, 275-289.

Graves, D. (1983). Writing: Teachers and children at work. Portsmouth, NH: Heinemann.

Graves, A. (1998). Instructional strategies and techniques for middle school students who are learning English. In R. Gersten and R. Jimenez (Eds.). Promoting learning for culturally and linguistically diverse students. Belmont, CA: Wadsworth.

Graves, A., & Montague, M. (1991). The use of cues in narrative story writing. Learning Disabilities Research and Practice 7 273-280.

Graves, A., Montague, M., & Wong, Y. (1990). The effects of procedural facilitation on the story composition of students with learning disabilities. Learning Disabilities Research 5, 88-93.

Graves, A., Semmel, M., & Gerber, M. (1994). The effects of story prompts on students with and without learning disabilities. Learning Disability Quarterly, 17, 154-164.

Graves, A., Semmel, M., & Gerber, M. (in preparation). The effects of long and short story prompts on students with and without learning disabilities: A follow-up study.

Graves, A., Valles, E., & Prodor, C. (in preparation). The effects of optimal learning environment (OLE) vs. traditional instruction on compositions of bilingual students with learning disabilities.

Hammill, D., & Larsen, 5. (1993). Test of Written Language II. Austin, TX: Pro-Ed.

Lewis, R., Graves, A., Ashton, T., & Kieley, C. (1997). Enhancing writing skills of students with learning disabilities. Learning Disabilities Research and Practice 11 212-220.

MacArthur, C. A., Graham, S., Schwartz, S. S., & Schafer, W. D. (1995). Evaluation of a writing instruction model that integrated a process approach, strategy instruction, and word processing. Learning Disability Quarterly, 18, 278-292.

Montague, M., & Graves, A. (1993). Teaching narrative composition to students with learning disabilities. Teaching Exceptional Children, 56 45-48.

Montague, M., Graves, A., & Leavell, A. (1991). Planning, procedural facilitation, and narrative compositions of junior high school students with learning disabilities. Learning Disabilities Research and Practice, 6, 219-224.

Rapp, W. H. (1997). Success with a student with limited English proficiency: One teacher's experience. Multiple Voices for Ethnically Diverse Exceptional Learners 2 21-37.

Ruiz, N. T. (1995a). The social construction of ability and disability I: Profile types of Latino children identified as language learning disabled. Journal of Learning Disabilities, 28, 476-490.

Ruiz, N. T. (1995b). The social construction of ability and disability II: Optimal and at-risk lessons in a bilingual special education classroom. Journal of Learning Disabilities 28, 491-502.

Ruiz, N. T., & Figueroa, R. A. (1995). Learning-handicapped classrooms with Latino students: The optimal learning environment (OLE) Project. Education and Urban Society, 27, 463-483.

Ruiz, N. T., Figueroa, R. A., Rueda, R., & Beaumont, C. (1992). History and status of bilingual students in special education. In R. Padilla & A. Benavides (Eds.), Critical perspectives in bilingual education (pp.349-380). Tempe, AZ: Bilingual Press.

Williams, R.L. (1996). Aranas, aranas por todas partes! Cypress, CA: Creative Teaching Press.

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 29 197-212.

Zaragoza, N., & Vaughn, S. (1993). The effects of process writing instruction on three 2nd-grade students with different achievement profiles. Learning Disabilities Research and Practice, 7, 184-193.

This research was conducted as part of CIERA, the Center for the Improvement of Early Reading Achievement, and supported under the Educational Research and Development Centers Program, PR/Award Number R305R70004, as administered by the Office of Educational Research and Improvement, U.S. Department of Education. However, the contents of the described report do not necessarily represent the positions or policies of the National Institute on Student Achievement, Curriculum, and Assessment or the National Institute on Early Childhood development, or the U.S. Department of Education, and you should not assume endorsement by the federal government.

© 2001 CIERA. All rights reserved.