Article

Learning to Write in Braille: An Analysis of Writing Samples from Participants in the Alphabetic Braille and Contracted (ABC) Braille Study
Jane N. Erin and Tessa S. Wright
Abstract: This article reports the results of data from 114 writing samples of 39 children who read braille and who were included in the Alphabetic Braille and Contracted Braille (ABC) Study between 2002 and 2005. Writing characteristics, miscues, and composition characteristics are analyzed, and two case studies are included.
Altbough there is a growing body of research on the development of literacy in young readers of braille, limited attention has been paid to the mechanical and conceptual process of how young children who are blind learn to write. Writing is generally taught to braille-reading students using a braillewriter, although occasionally the skill is introduced through the use of a slate and stylus or an electronic braille device. Each tool requires different patterns of coordination than writing in print, which raises questions about how the differences in physical production influence the quantity and quality of writing. Children who are blind also have distinctive experiences with conceptual learning, suggesting that variations in the content of written products may exist. However, there is little published evidence of how children who are blind learn to write, particularly during the early elementary years when children flrst learn the form and meaning of writing. Evaluating writing samples of young children is complex because writing involves both physical transcription and the creation of meaning through words. Some researchers have emphasized measures that assess the process of transcription, such as copying words and sentences; others have analyzed production through such measures as correct punctuation, sentences, words written, and words correctly spelled (Coker & Ritchey, 2010). The limited vocabulary and varied writing proficiency of young writers make it challenging to conduct reliable analyses. In a review of the literature on curriculumbased writing measures, McMaster and Espin (2007) found weak relationships between the quality of holistic writing

I2LJ EARN C E U S ONLINE

by answering questions on this article. For more information, visit: <http://jvib.org/CFUs>.

2011 AFB, All Rights Reserved

Journal of Visual Impairment & Blindness, July 2011

389

Article
samples and specific measures; they recommended that qualitative as well as quantitative features of writing be considered, especially in the lower grades. Reliable writing assessments of young children pose an even greater concern when the children are visually impaired (that is, are blind or have low vision), given the distinctive factors related to writing and the absence of research. Research has primarily investigated students who were at the mid-elementary school level or older. Kreuzer (2007) analyzed the writing productions of fourth- and fifth-grade students who had low vision or were blind. She gathered two writing samples each from 15 students who were blind, 15 with low vision, and 15 who were sighted; one sample was elicited from a prompt about an imaginary scenario, and the second was elicited from a prompt about the students' own experiences. From these samples, Kreuzer analyzed data related to the number of words and the percentage of abstract nouns and adjectives. Holistic scores were assigned for each composition, and the use of abstract and concrete nouns, pronouns, adjectives, and adverbs was assessed. Both the students who were blind and those with low vision used more concrete and fewer abstract nouns in their imaginative samples than did the sighted students, and the students who were blind used fewer adjectives. The holistic writing scores were significantly lower for the students who were blind or had low vision in the writing samples that were based on the students' experiences. It is interesting to note that the students with low vision wrote significantly shorter samples than did the sighted students for both imaginative and experiential writing. Although many students in Kreuzer's study chose to use electronic notetakers to produce their samples, no student with low vision chose to write using a computer or any adaptive device. Several studies have addressed the writing production of high school students. Koenig (1987) analyzed writing samples produced by 84 students who were blind and who were aged 9, 13, or 17. He compared the samples to those of sighted students according to protocols developed for the National Assessment of Educational Progress (1981; cited in Koenig, 1987). The analyses included rhetorical effect, cohesiveness, syntax, spelling, and types of errors. Students in the 9- and 13-year-old age groups performed similarly to their sighted peers, and in some areas, such as spelling and rhetorical effect, they outperformed the sighted students. The 17-year-old students performed similarly to their sighted peers, but their samples were weaker in rhetorical effect, cohesiveness, and spelling. Ryles's (1997) study investigated the literacy skills of students with visual impairments who leamed braille early with those of students who learned braille later or not at all. When the writing samples of these three groups were rated by English teachers along with the writing samples of sighted students, the samples of the early readers of braille were rated as high as or higher than those of the sighted students, unlike the samples of the later readers of braille and the students who were visually impaired who did not read braille. The literature related to the writing skills of students who are blind suggests wide variation among school-aged students who read braille, with possible areas of difficulty related to the use of vocabulary, the structure of the composition, and cohesiveness
2011 AFB, All Rights Reserved

390

Journal of Visual Impairment & Blindness, July 2011

CEU Article
for some students. However, the scant Literature, the differences in research designs, and the absence of investigations of writing at the early elementary levels yield an incomplete body of information about how braille writing skills develop in young children. The writing samples produced for the ABC Braille study provide an initial longitudinal investigation of a process that can illuminate features that affect the quality of writing of children who read braille. This article reports the findings from the data on writing that addressed the following questions: 1. Is there a relationship between words per passage (WPP) or words per sentence (WPS) and students' knowledge of contractions? 2. Is students' knowledge of contractions related to the quality of compositions? 3. What types and frequency of miscues occurred in the writing samples? 4. What mechanical characteristics of braille writing were demonstrated by the students? 5. Are there differences in the scores on composition and the frequency of miscues of students who were the highest and lowest achievers? 6. What patterns of writing development were evident in the individual students? learned braille contractions early and those who learned them later (Wall Emerson, Holbrook, & D'Andrea, 2009). The study yielded extensive data about the participants' general literacy development, including writing skills. The data were gathered from writing samples produced annually by young readers over a two- to five-year period between preschool and the third grade. The Institutional Review Board at Peabody College, Vanderbilt University, approved the project according to the guidelines of the World Medical Association's Declaration of Helsinki on Ethical Principles for Medical Research Involving Human Subjects. In the spring of each year of the ABC Braille study, each participating student produced a writing sample to provide comparative information on writing proficiency and spelling in an unstructured task. The participants included students from the United States and Canada who were recruited through announcements in professional journals and at meetings; they were blind, were identified as readers of braille by their educational teams, and had no disabilities other than blindness. A researcher gave an oral prompt to initiate the writing task, and the students were asked to write independently about the topic. The prompts were about personal experiences, such as a favorite thing to do at home or a favorite thing to do for a holiday. No time limit was imposed, so the length of the samples depended on the students' motivation and interest as well as writing stamina. The researchers were encouraged to say as little as possible while the students were writing, although it was sometimes necessary to repeat the prompt to encourage writing. They videotaped each child's hands on the braillewriter

Method
This article reports data from writing samples that were produced by students who read braille and whose literary skills were documented in the ABC Braille study. The research, funded by the American Printing House for the Blind with substantial support from the Canadian Braille Literacy Foundation, sought to identify distinctions between young readers who

2011 AFB, All Rights Reserved

Journal of Visual Impairment & Blindness, July 2011

391

CEU Article
to document writing techniques and to record rereading. The writing data were analyzed from 113 samples produced between 2002 and 2007 by 39 students. The students who produced the samples were permitted to write with any preferred device. Two samples were produced using electronic braille devices, and the others were produced using Perkins Braillewriters. Seven prewriting samples were also reviewed but not analyzed because they did not include conventional braille symbols.
CHARACTERISTICS OF THE COMPOSITIONS THAT WERE ANALYZED

study (Schoch, 2008). Using this form, the researchers assigned from 1 to 3 points for each sample in each of the three categories of content; structure; and voice, style, and tone, for a total of up to 9 points. A Pearson two-tailed correlation was conducted to determine whether there was a relationship between the quality of a composition and the level of contractivity, as well as the level of reading achievement. Miscue analysis Miscues for all the samples were counted and classified according to whether they were phonetic, braille related, or unknown. A writing miscue was a word that was not conventionally written using either contracted or uncontracted braille or any combination of correct contractions. Braille-related errors were categorized as a missed dot, an added dot, horizontal transposition, vertical transposition, the duplication of letters, or the use of whole words as part words.

Contractivity and length of words and passages The relationship between knowledge of contractions (contractivity) and WPP and WPS was examined using a two-tailed Pearson bivariate correlation analysis. For both the WPP and WPS, the Pearson correlation analysis was used to examine both the relationship between the overall mean WPP and WPS of each student's writing samples and mean contractivity, as well as the relationship between scores at each grade level and the number of contractions known at that grade level. Contractivity was calculated on the basis of the total number of contractions that each student mastered during each year of the study, as reported annually by the students' teachers of students with visual impairments on the checklist provided in the Assessment of Braille Literacy Skills (Koenig & Farrenkopf, 1995). Characteristics of the compositions The quality of the content of the compositions was analyzed according to a threecategory ranking system developed for the
392 Journal of Visual Impairment & Blindness, July 2011

Writing mechanics
During the second year of the study, the researchers decided also to document the students' writing mechanics. They were asked to videotape from an angle where hand and finger movements could be observed. In many cases, it was not possible to do so or the observers overlooked the taping, so a visual analysis of the students' writing mechanics was possible for only about half the samples. Each videotape was rated for appropriate fingering, use of the line spacer, firm pressure on keys, checking work during writing, and changing a line on hearing the bell. Items were rated as evident consistently, sometimes evident, or not observed.
2011 AFB, All Rights Reserved

Article
H I G H AND LOW ACHIEVERS

To explore the relationship between writing skills and reading achievement, the researchers compared the writing compositions of the six highest-achieving readers in the study with those of the seven lowestachieving readers. Low-achieving students were those whose reading assessments were at or below grade level throughout the study, while high-achieving students were those whose reading assessments were at or above grade level throughout the study (Wall Emerson, Sitar, Erin, Wormsley, & Herlich, 2009). A /-test was used to compare the mean number of WPS, WPP, and composition scores of the two groups.

Results
CONTRACTIVITY AND LENGTH OF WORDS AND PASSAGES

passages or the absence of a composition format), and most of the samples that could be scored consisted of simple sentences, such as "I like to play at home." Twenty-nine students had samples that could be scored from Grades 1 and 2; only 20 students had samples from Grade 3, and 11 had samples from Grade 4. Therefore, only samples from the Grade 1-Grade 2 combination were scored to ensure that the results were not due to maturity or missing samples. The average score for each student from Grade 1 and Grade 2 was calculated, and the relationship of this score with the level of contractivity was calculated. The Pearson two-tailed correlation for contractivity and Grade 1-Grade 2 content scores was .391 (p = .0435), which was not statistically significant (Schoch, 2008).
WRITING MISCUES

The two-tailed Pearson bivariate correlation analysis of WPP and WPS with contractivity was calculated using mean scores across grade levels as well as individual raw scores for each grade level. The only statistically significant correlation was a positive correlation between overall WPP and contractivity; as the students mastered more contractions, the overall mean number of WPP increased. The Pearson two-tailed correlation for contractivity and WPP scores was .388 (p = .016).
COMPOSITION AND CONTRACTIVITY

No significant relationship between contractivity and composition scores was found. Composition scores were calculated only for students who had writing samples for both Grades 1 and 2 because few students had complete samples in kindergarten (because of the shortness of

Writing miscues were analyzed for all the writing samples. In total, 321 miscues were identified in the 113 writing samples. These miscues were classified as phonetic, which were attempts to spell according to the sounds of the word; braille, which were associated with braille rules or configurations; and unknown, for which the reason for the error could not be determined. Of all the writing miscues, 81% (260) were phonetic, reflecting the students' attempts to write an unfamiliar word on the basis of their knowledge of the relationships of sounds and symbols. Examples of these miscues included pritty for pretty and sandgo for San Diego. Only 13% (40) of the miscues were braille related; these miscues were symbols that were probably an error in braille writing, as judged by the similarity of the symbols. These

2011 AFB, All Rights Reserved

Joumal of Visual Impairment & Blindness, July 2011

393

CEU Article
Boxl Types of braille miscues.
Missed dots (9) Uf for of ing for ttie 2356c for because p for and ing for the 23456ent for went mame for make vast)46n for vacation (2) sifdes for slides. Thin for then (2) win for went slidegh for slides trice for twice emportant for important wun for run Con for in (2) h for to ben25 for bench Theen for fften D/ng for doing (2) G/ng for going stiakers for stickers tg for fo en for on drings for irt/ngs /'p for I'm holied for holiday griving for driving dtis for friends dun for iun fer for her sicher for s/sfer su/ng for singing ritble for writing 256at for whaf dump for y'ump ghe for fhe ouside for outside (2)

Added dot (8)

Horizontal transposition (12)

Vertical transposition (7)

Duplicate letters (1) Whole words used as part words (5)

miscues were classified by type (see Box 1). Horizontal (left-right) transpositions were the most common miscue, with 12 occurrences. The only braille rule-based miscues were 5 uses of whole-word contractions as part-words, as in writing ding for doing. The remaining 6% (21) of miscues were configurations with no identifiable origin, such as ceebord used for an unknown word. The number of miscues per sample varied, and was mainly associated with the length of a composition. The largest number of miscues was from a student who was among the lowest reading achievers; this student's miscues increased by grade level in keeping with the increased length of his writing passages. The second largest number of writing miscues was generated by a student who was the highest achiever in the study; the majority of her miscues were phonetic spellings in kindergarten and

Grade 1, when she wrote much longer passages than other students at the same grade level and included complex words, such as decided (spelled disideed). Occasionally miscues reflected limitations in students' exposure to literacy, as when one capable student described her family's vacation stay at "Embassee Sweets." A sighted student would probably have learned the correct spelling of the name of the hotel from exposure to signs, napkins, pens, and other items that are available in hotel rooms.
WRITING MECHANICS

Writing mechanics were videotaped beginning in the second year of the study. Videotapes could be analyzed for only 75 of the 114 writing samples because of missing videotapes or difficulty with the camera angle. The skills that were rated included appropriate fingering, the use of the line spacer, firm

394

Journal of Visual Impairment & Blindness, July 2011

2011 AFB, All Rights Reserved

CEU Article
pressure on keys, checking work during writing, and changing a line on hearing the bell. Correct fingering was inconsistent across grade levels. The proportions of students who used correct fingering were 40% in kindergarten, 64% in Grade 1, 54% in Grade 2, 37% in Grade 3, and 50% in Grade 4. It seems surprising that the consistency of fingering did not improve with greater experience. The frequency with which the students checked their own work during the writing process was also documented. Two of three kindergarten students did not check their work. In Grade 1-Grade 4, 62%70% of the students consistently checked their writing samples while writing. Between 16% and 28% of the students never checked their work. More than 90% of all the students in the videotaped samples consistently used the line spacer when appropriate, and 95% pressed firmly on the keys. The students failed to stop writing at the bell sometimes or always in about 25% of the rated samples, usually in kindergarten or Grade 1. Although the students had a choice about how to produce their writing samples, only 2 of the 113 samples were produced on a device other than a Perkins Braillewriter. One student used the Mountbatten Brailler because her finger strength was limited because of a physical condition, and the other used a BrailleNote. In both cases, the electronically produced sample was much longer than previous samples and was used in the last year of the study. Grade 4. One student produced a 76-word sample, in contrast to the previous 24-word sample. The other student wrote a 284-word sample, the longest sample produced by any student in the study and contrasted with her second longest sample of 33 words. These two samples did not differ from previous samples with regard to miscues or punctuation errors, but the first student wrote sentences in her electronic sample that were twice the length of the sentences in her previous samples. No conclusions can be drawn about the effects of using an electronic device for spontaneous writing, but this is an important issue for future research.
H I G H AND LOW ACHIEVERS

The eight highest-achieving students were compared with the seven lowestachieving students with regard to WPS, WPP, and composition. There were no significant differences in the groups on any of the measures, as assessed with a twotailed i-test. Miscues were also examined for differences. The only notable difference was that the number of miscues in the samples of the high-achieving students tended to remain consistent or to decrease over time, whereas the number of miscues in the samples of the low-achieving students remained stable or increased.

Case studies
One of the values of longitudinal research is the opportunity to observe change over time as students' productions are compared with their own writing from the previous year. Although each child attempted to produce writing samples during each year of the study, only two children produced five writing samples for all five years of the study because of their time of entry into the study. The students whose samples are described here are examples of individual change over time. Their samples are not presented to respond to the primary

2011 AFB, All Rights Reserved

Journal of Visual Impairment & Blindness, July 2011

395

CEU Article
Table 1 Length of compositions and sentences and percentages of miscues for Patricia.
Kindergarten Words per passage Mean words per sentence Percentage of miscues compared to total words 68 9.9 (Two run-on sentences.) 19 Grade 1 149 6.8 14 Grade 2 207 12.2 2 Grade 3 130 11.8 .05 Grade 4 95 12 2

questions of the study but, rather, to demonstrate the variability of individual change in young students.
PATRICIA

Patricia, one of the highest-achieving students in the study, produced her first writing sample in kindergarten and wrote a sample each year through Grade 4. She was introduced to contractions beginning in kindergarten, when she used some whole-word alphabet contractions as well as with and the. She had been taught all 189 contractions by Grade 2. By her last year in the study. Grade 4, she was reading at the 8th-grade level. Her spelling score in Grade 4 was at the 12th-grade level on the Brigance (1999). Length of composition and miscues Table 1 presents data from Patricia's five samples regarding the number of WPP and WPS, as well as the percentage of miscues among the total words in each sample. Patricia's kindergarten passage was longer than that produced by any other student in the study at the kindergarten level, and the length of her passages increased from kindergarten to Grade 1 through Grade 2, when she produced a 207-word passage. The first two passages had fewer WPS, but for the next three years, the number of WPS was similar, ranging from 11.8 to
396 Journal of Visual Impairment & Blindness, July 2011

12.2. By Grade 4, Patricia was approaching the standard adult range of 15-20 WPS ("Ask Oxford," 2010), although the length of her sentences stabilized in the last three samples. Miscues compared with total words decreased over the first three samples, but remained stable at l%-2% after Grade 2. Patricia's kindergarten sample contained more writing miscues than did any other student's sample at that level, in part because the sample was longer, and Patricia used a more varied vocabulary than most students. Patricia's kindergarten sample included 12 writing miscues, most of which were invented spellings, such as prck for park. On three occasions, Patricia wrote gerund forms of "do" and "go" using an alphabet contractions such as ding for doing. In Grade 1, 21 of 149 words contained writing miscues. As in kindergarten, most miscues were phonetic spellings of unfamiliar words: groseree for grocery, latter for ladder, and cusins for cousins. In Grade 2, the number of miscues decreased markedly; only 5 miscues were present among 205 words. Patricia was then familiar with conventional spelling, and her only miscues were vestables for {vegetables), friuts (for fruits), seezin (for season), sucksed (for succeed), and eqitment (for equipment). She used contractions appropriately and single spaced

2011 AFB, All Rights Reserved

CEU Article
written material. The Grade 3 sample included only one miscue, the use of lose for loose, and the Grade 4 sample also included only one true miscue (actionally for occasionally). Composition and content Patricia's average score for the content of compositions was 7.6 out of a possible 9 points, with the highest scores of 9 assigned for the Grade 2 and Grade 3 essays. These essays were long and rich in content. From the beginning, Patricia's writing showed an understanding of writing as a form of meaning. In kindergarten, Patricia already showed an awareness of the reader when she asked, "Do you like to play with Mom and Dad?" She was the only child who used voice in this way from the beginning, indicating an awareness of the reader's perspective. When asked to read the story, Patricia read it fiuently and with meaning, regardless of writing errors. I like doing with my Mom is riting and read my favorite with my Mom is eating. Mu fravx FAVORITE thing ding with my Dad is going to with car and griving DRIVING to the prck PARK. I love my Mom SHOM and Dad to do you like your Mom or Dad. To and I like to play with Mom for Dad to. Do you like to play with Mom and dad The Grade 1 sample about Halloween experiences still contained no clear introduction or conclusion; however, Patricia described events in sequence and used standard words with many invented spellings. She also included complex sentences with subordinate clauses as in the
2011 AFB, All Rights Reserved

following: "When my cusins and I got to the groseree store, we got stiakers there." Her Grade 2 composition about baseball included evidence of an introduction as well as two clear topical sections: eating healthy foods and working together with a team. For the first time, a conclusion was present: "Now you know how easy it is to work together in a game or sport. So tell that to your teacher or team mates that read all about baseball and have fun." Patricia's writing now reflected a more complex understanding of voice; she was writing to an audience and organized her thoughts to promote understanding. The Grade 3 and Grade 4 samples were shorter than the Grade 2 sample. In the conclusion to her Grade 3 sample, she used a simile that was especially interesting because it referred to color: "The weekend is like a crayon box. You don't know what color you are going to get." The Grade 4 sample was even briefer and lacked the enthusiasm of the previous essays. There was no development or originality of content in this essay, perhaps as a result of an uninteresting prompt or of Patricia's diminishing enthusiasm with writing tasks with age. Although her essays were long and well constructed, Patricia rarely used adjectives or descriptions except in the case of identifier nouns, such as grocery store or crayon box; the general purpose of her writing was to relate information.
MARY

Mary also generated five samples during her five years in the study, and she was introduced to contractions more gradually than was Patricia, although she still learned contractions at a faster rate than 24 of the 43 students in the study. By
Journal of Visual Impairment & Blindness, July 2011 397

CEU Article
Table 2 Length of compositions and sentences and percentage of miscues for Mary. Kindergarten Words per passage Mean words per sentence Percentage of miscues compared to total words 24 6 17 Grade 1 12 6 0 Grade 2 49 6.1 10 Grade 3 26 13 0 Grade 4 22 11 9

Spring of Grade 2, she had learned 73 contractions; in Grade 3, she had mastered 148; and by Grade 4, she knew all 189 contractions. In spite of the teacher's documentation that she had learned most contractions, Mary did not spontaneously contract any words until her third writing sample, produced in Grade 2. In the last three samples, she consistently contracted and, the, and like but did not use contractions on 10 other opportunities. Mary's reading progress closely matched the other students in her grade. In Grade 3 and Grade 4, she was on grade level in reading and vocabulary, and her spelling scores were a year beyond grade level during both years. Length of compositions and miscues Although her reading was on grade level, Mary wrote brief passages that ranged from 12 to 49 words, with the longest produced in Grade 2. For the first three grades, her sentences were short, with about 6 words per sentence; in Grades 3 and 4, the length increased, and she used greater variety in syntax. There were few miscues (see Table 2). In kindergarten, Mary wrote sandbox as sad box, pink as pipk, and hole as hols. No miscues occurred in Grades 1 and 3, but in Grade 2, she wrote slides as slfdes, shoot as shot, and merry-goround as mary ground. In her final sample, she wrote holiday as holday and
398 Joumal of Visual Impairment & Blindness, July 2011

because as dots 2356c. Of the nine miscues, three appear to be related to braille writing errors {pipk for pink, slfdes for slides, and 2356c for because). Composition and content Mary's average score for content was 4.6 out of a possible 9 points, with a low score of 3 in Grade 1 and a high score of 6 in Grade 2. Her writing samples were characterized by repetitive simple sentences beginning with / or we. Her first sample, in kindergarten, consisted of four simple sentences: "I like to ride my bike." "I have a pink bike." "I play in my sand box." "I dig a hole in the sand box." The Grade 1 sample was similar, with just two simple sentences. By Grade 2, Mary again wrote simple sentences beginning "I like . . ." but she ended with a sentence that suggested a conclusion: "It is fun playing at recess." The Grade 3 essay included two compound sentences with a subject of I and multiple verbs: "Also I like to plant and ride my bike and sleep." The final Grade 4 composition had greater variability in syntax and an expression of preference: "I like it because it is very fun and you get a lot of presents." Mary's compositions were short, and the content was limited to telling the reader what she liked to do. In response to the prompts, which all began with "What is your favorite . . .?" Mary provided information but did not develop a fully

2011 AFB, All Rights Reserved

CEU Article
formed response. Of the 133 words generated in her five essays, only one was an adjective or adverb. Although Mary became more competent in writing varied sentences with more complex structures, the content of her responses was devoid of detail or complex thought. Eor her, writing appeared to be an academic exercise of creating correct sentence structures that addressed the prompt question. The essays generated by Patricia and Mary demonstrate their growth in both the mechanics of writing and their ability to convey meaning and personal thoughts through writing. One student used detailed language to embellish her thoughts, while the other generated sequences of simple, factual sentences that expressed her preferences in response to the prompts. Both students wrote their longest essays in Grade 2, and their final Grade 4 essays were shorter than those in the two previous years. This change in length raises questions about possible changes in motivation, either in general writing or in the use of the research writing prompts as students mature. Both girls became more competent over the five, years they participated in the study, particularly with regard to a decreased number of miscues and clearer structural development. associated with such knowledge. The relationship with WPP and knowledge of contractions suggests the possibility that using contractions may reduce the effort needed to produce more material, although the significance was modest. Only 13% of the miscues were related to characteristics of the braille code, even among the youngest students; 81% were attempts to use phonetic rules in spelling. This finding contrasts with the common belief that the early use of contractions promotes poor spelling, especially when considered with the overall finding of the ABC Braille study that the students in the group spelled better than did their peers (Wall Emerson, Holbrook, & D'Andrea, 2009). It also supports the conclusions of the ABC Braille study that the literacy difficulties of students who were blind were not associated mainly with the braille code but were the same problems experienced by all young readers who are working to resolve the rules and irregularities of a complex written language. The two case studies presented here suggest the value of using regular writing samples as a means of documenting students' writing over time. For future research, writing prompts that stimulate more imaginative content may be considered to avoid the repetitive and formulaic responses that some children produced. This option was supported by Kreuzer's research (2007), which found superior holistic scores by students with visual impairments who were writing imaginative samples. Predictably, the types and frequency of miscues changed with maturity for most children, but competent writers like Patricia showed a more rapid decrease in writing miscues after Grade 1.

Discussion
The analysis of 114 writing samples from young readers of braille showed wide variability in writing features, such as the length of passages, the length of sentences, and the frequency of miscues. While the overall number of WPP was related to the students' knowledge of contractions, the number of WPS and the holistic content of their compositions were not

2011 AFB, All Rights Reserved

Journal of Visual Impairment & Blindness, July 2011

399

CEU Article
Although the data on the mechanics of writing were confounded by the limited number of usable videotapes, some findings of interest emerged. One was the fact that fingering on the braille keys became less consistent with age. Perhaps the students used more spontaneous fingering patterns to increase their speed as they became more familiar with writing. Teachers of students with visual impairments may also reduce their emphasis on the mechanics of writing as the children grow older, especially if the children are producing accurate braille. The younger children often used the index finger, rather than the thumb, to operate the space bar, which may be related to differences in the size of their hands and their motor development. Although these questions are of interest, they should be interpreted with caution since there were was a small number of rated samples for young children. In addition, interobserver agreement was not assessed for the videotaped observations, which is a limitation of the study. For teachers of young readers of braille, the results suggest the value of writing samples as a means of documenting skills over time. Aspects of writing, such as planning and producing a full essay, writing words accurately, and developing an effective writing style, can be assessed only through analyses of written text. Teachers of students with visual impairments who leave spontaneous writing instruction to classroom teachers will miss opportunities to observe the applied use of contractions, as well as the students' ability to resolve literacy problems. The limitations of this analysis and the scant literature on the development of writing skills among students who read
400 Joumal of Visual Impairment & Blindness, July 2011

braille support the need for research on how these children learn to write. As technology changes, speed and the demands of the task of writing increase for the general population. The effects of using technological writing tools are unknown, but in this study the two students who used technology wrote longer essays, which suggests the need for further investigation. The students in this study developed their writing skills in widely varied ways: some students wrote long, expressive essays that were filled with miscues, while others wrote short unimaginative passages that were free of errors. In parallel with the general findings of the ABC Braille study, these results support the importance of leaming to read and write in a meaningful context in which individual assessment, experience, and intervention can ensure leaming over time.

References
Ask Oxford (2010). Retrieved from http://www. askoxford.com/betterwriting/plainenglish/ sentencelength Brigance, A. H. (1999). Brigance Comprehensive Inventory of Basic SkillsRevised (CIBS-R). North Billerica, MA: Curriculum Associates. Coker, D., & Ritchey, K. (2010). Curriculumbased measurement of writing in kindergarten and first grade: An investigation of production and qualitative scores. Exceptional Children, 76, 175-193. Koenig, A. (1987). A study of expressive writing skills of blind students including partial replication of the National Assessment of Educational Progress Third Writing Evaluation. (Doctoral dissertation, Vanderbilt University). Dissertation Abstracts International, 48, 1734A. Koenig, A., & Farrenkopf, C. (1995). Assessment of braille literacy skills. Houston, TX: Region IV Education Service Center. Kreuzer, D. (2007). An analysis of writing

2011 AFB, All Rights Reserved

CEU Article
practices in 4th- and 5th grade students with visual impairments (Doctoral dissertation. University of Califomia, Berkeley). UMINo. 3331676. McMaster, K., & Espin, C. (2007). Technical features of curriculum based measurement in writing: A literature review. Journal of Special Education, 41, 68-84. National Assessment of Educational Progress. (1981). The third assessment of writing: 1978-79 released exercise set, Princeton, NJ: Author. Ryles, R. (1997). The relationship of reading medium to the literacy skills of high school students who are visually impaired. (Doctoral dissertation. University of Washington). UMI No. 9819296. Schoch, C. (2008). Summary of writing analysis of ABC Braille study. Unpublished manuscript. Department of Disability and Psychoeducational Studies, University of Arizona, Tucson. Wall Emerson, R., Holbrook, M. C, & D'Andrea, F. M. (2009). Acquisition of literacy skills by young children who are blind: Results from the ABC Braille study. Journal of Visual Impairment & Blindness, 103, 610-624. Wall Emerson, R., Sitar, D., Erin, J., Wormsley, D., & Hedich, S. (2009). The effect of consistent structured reading instruction on high and low literacy achievement in young children who are blind. Journal of Visual Impairment & Blindness, 103, 595609.

How to Contact JI/78

SUBMIT

To submit an article. Research Report, or Practice Report for peer review, e-mail it to Dr. Duane R. Geruschat, editor in chief, JVIB: <jvib@ jhmi.edu>; or mail it to Lions Vision Center, 550 North Broadway, 6th Floor, Baltimore, MD 21205. Inquiries should be sent to: <jvibeditor@afb.net>.
CONTRIBUTE

To offer information on a program, conference, product, or promotion for possible publication in From the Field, News, or Calendar, contact: Rebecca Burrichter, senior editor, AFB Press, 2 Penn Plaza, Suite 1102, New York, NY 10121; fax: 917-210-3979; e-mail: <rebeccab@ afb.net>.
ADVERTISE

To advertise in JVIB or to receive information on advertisement rates, contact: Anne Durham, sales and marketing manager, American Foundation for the Blind, Huntington, West Virginia; e-mail: <adurham@ afb.net>.
SUBSCRIBE

Jane N. Erin, Ph.D., associate editor for practice, JVIB, and professor. Department of Special Education, Rehabiiitation, and School Psychology, College of Education, University of Arizona, P.O. Box 210069, Tucson, AZ 85721; e-mail: <jerin@ u.arizona.edu>. Tessa S. Wright, Ph.D., assistant professor of practice and coordinator. Program in Visual Impairment, University of NebraskaLincoln, 202G Barkley Memorial Center, Lincoln, NE 68583: e-maii: <twright5@unl.edu>.

To subscribe to JVIB, contact: AFB Press, P.O. Box 1020, Sewickley, PA 15143; phone: 800-232-3044 or 412-741-1398; fax 412-741-0609; email: <afbsub@abdintl.com>; web site: <www.afb.org/store>.
SEARCH

To find JVIB, on the visit: <www.afb.org/jvib>.

web,

2011 AFB, All Rights Reserved

Journal of Visual Impairment & Blindness, July 2011

401

Copyright of Journal of Visual Impairment & Blindness is the property of American Foundation for the Blind and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.