Using Technology to Enhance theWriting Processes of Students withLearning DisabilitiesCharles A. MacArthur

Abstract

This article reviews the ways that computers can support writing by students with learning disabilities, with an emphasis onapplications that go beyond word processing. Following an overview of research on word processing is a discussion of software thatassists with the basic processes of transcription and sentence generation, including spelling checkers, speech synthesis, wordprediction, and grammar and style checkers. Next, applications that support the cognitive processes of planning are reviewed,including prompting programs, outlining and semantic mapping software, and multimedia applications. Finally, the use of computernetworks to support collaboration and communication with diverse audiences is addressed.

M ark, a sixth-grade studentwith a severe reading dis-ability, enjoys talking about

his interests in class. However, on writ-ing tasks, his spelHng is often so un-usual that even he cannot read whathe has written. Consequently, Markis extremely reluctant to write. Inschool, he uses a word processor thatprovides speech synthesis and a per-sonalized word bank. He discusses histopic individually with the teacher,who adds key words from this dis-cussion to his word bank. As he writes,Mark can select many of the words heneeds from the word bank, rather thanhaving to spell them. The speech syn-thesizer provides a check on the accu-racy of his writing and supports hisreading as he prepares to share hiswriting orally with his peers.

Each month a flurry of activitybreaks out in Mrs. Adams's class asthe deadline for the class magazineapproaches. Students rush to completeand edit their best writing. The edi-tors for the month collaborate on lay-out and production. Desktop pub-lishing software enables them toproduce a professional-looking prod-

uct, with headlines, graphics, and neatcolumns of text. This regular publish-ing project has had a dramatic impacton the amount and quality of writingproduced by this class. Parents andpeers who read the magazine cannottell from the product that the writersare all students with learning disabili-ties (LD).

As these examples illustrate, com-puters are flexible writing tools thatcan enhance writing processes in manyways. They can support the basic skillof being able to produce legible textwith correct mechanics, as well as themore complex cognitive processes ofplanning, writing, and revising textand the social processes of collabora-tion and communication with an au-dience.

The support provided by comput-ers may be especially beneficial forstudents with LD, who often find writ-ing frustrating. Students with LDperform less well than their peers ona variety of written language tasks(Englert, Raphael, Anderson, Gregg,& Anthony, 1989; Graham, Harris,MacArthur, & Schwartz, 1991). Theyoften have difficulty with the physi-

cal demands and conventions of writ-ing and with fluent production of sen-tences. Many students with LD havedifficulty coordinating the complexcognitive processes of setting goals,generating content, organizing theirwriting, and evaluating and revisingtheir text. However, recent researchdemonstrates that instructional pro-grams that provide (a) a supportivesocial context for writing in the class-room, (b) meaningful writing tasks,and (c) instruction in writing processescan improve the writing achievementof students with LD (Englert, Raphael,Anderson, Anthony, & Stevens, 1991;MacArthur, Graham, & Schwartz,1993).

The purpose of this article is to re-view specific ways in which comput-ers can support the writing processesof, and enhance writing instruction for,students with LD. Because word pro-cessing is now quite common inschools (Becker, 1993) and publishedresearch reviews on word processingare available (Bangert-Drowns, 1993;Cochran-Smith, 1991; MacArthur,1988), the present review starts with abrief overview of that research. Themajor focus of the article is on com-

JOURNAL OF LEARNING DISABimiESVOI-UME 29, NUMBER 4, JULY 1996

PAGES 344-354

VOLUME 29, NUMBER 4, JULY 19% 345

puter applications that go beyondword processing. Following the over-view of word processing is a discus-sion of computer tools that can assistwith the basic processes of transcrip-tion and sentence generation. Next,applications that support the cogni-tive processes of planning are re-viewed. Finally, the use of computernetworks to support collaborativewriting and communication with di-verse audiences is addressed. It isimportant to note that many of theseapplications, particularly the morerecent advances, have little or no re-search support. Because developmentgenerally precedes research in thisfield, this article is not limited toresearch-supported techniques; how-ever, the discussions clarify the extentto which the techniques are supportedby research.

Word Processing

Word processors have severalcapabilities that may influence thewriting process. First, the editing fea-tures of the word processor allowwriters to make frequent revisionswithout tedious recopying. Conse-quently, writers may make more revi-sions, and it is possible that this easeof revision will encourage students toconcentrate on content while writinga first draft and edit for mechanicslater (Daiute, 1986a). The potentialimpact of word processing on revis-ing is significant, as revision is animportant aspect of the composingprocess that distinguishes expert writ-ers from younger and less skilled writ-ers (Fitzgerald, 1987). Students withLD, in particular, have a limited con-ception of revising as being an oppor-tunity to correct errors, and theirrevisions are restricted primarily tominor changes that do not affect theoverall meaning or quality of writing(MacArthur, Graham, & Schwartz,1991).

Simply having access to word pro-cessing has little impact on the revis-ing behaviors of students with LD; for

example, MacArthur and Graham(1987) found no differences in thenumber or type of revisions such stu-dents made using paper and pencilcompared to using word processing.Furthermore, the final drafts of pa-pers written on a word processor didnot differ from those written by handon any of the measures used inthe study, including overall quality;length; story structure; vocabulary;syntactic complexity; or errors in spell-ing, capitalization, and punctuation.Only two minor differences werefound between handwriting and wordprocessing: More deletions were madewith the former; and word processingresulted in more revisions during writ-ing of the first draft, whereas nearlyall revisions with handwriting weremade while writing the second draft.

However, instruction in revision incombination with word processing cansignificantly increase the amount andquality of revision by students withLD (Graham & MacArthur, 1988; Mac-Arthur, Schwartz, & Graham, 1991;Stoddard & MacArthur, 1993). Grahamand MacArthur taught elementary-school students with LD a strategy forrevising opinion essays. The strategyfocused on substantive revisions, suchas stating the thesis clearly, giving andsupporting reasons, increasing thecoherence of text, and closing with asummary statement. MacArthur andhis colleagues (MacArthur, Schwartz,& Graham, 1991; Stoddard & Mac-Arthur, 1993) provided instruction ina peer revising strategy in which pairsof students with LD learned to helpeach other revise papers written on aword processor. In all three studies,strategy instruction in combinationwith word processing resulted in in-creases in substantive revisions andimprovement in overall quality of com-positions.

Second, word processors give stu-dents the power to produce neat,printed work and to correct errorswithout messy erasures. The literatureon process approaches to writing andwhole language places considerableemphasis on the value of publishing

students' writing (Calkins, 1991). Themotivation provided by printed pub-lished work may be especially impor-tant for students who struggle withhandwriting and mechanics. Comput-ers make it possible to publish in awide range of professional-lookingformats. Desktop publishing programsmake it easy to produce newsletters,illustrated books, big books, businessletters, signs and posters, and manyother forms of work.

A third feature of word processorsthat is mentioned less often is thevisibility of the text on the screen(MacArthur, 1988). This visibility, to-gether with the use of typing ratherthan handwriting, can facilitate col-laborative writing among peers andscaffolded interactions betweenteacher and student. Peers can worktogether, sharing responsibility forgenerating ideas, typing, and revisingin flexible ways, as both partners cansee and read the text easily and typ-ing does not identify separate contri-butions. Daiute (1986b) studied pairsof low-achieving elementary-schoolstudents working on a series of col-laborative writing tasks. Through care-ful analysis of student talk and theresulting written products, she docu-mented ways in which the studentslearned writing techniques from eachother.

In addition to facilitating peer col-laboration, word processing can en-hance scaffolded interactions betweenteachers and students. The visibilityof the text on the screen enables teach-ers to more easily observe students'writing processes and intervene whenappropriate (Morocco & Neuman,1986). Teachers can scaffold students'writing by sharing writing tasks andproviding appropriate coaching(Cochran-Smith, 1991). The visibilityof text on the screen can also supportteacher-directed group lessons. Usinga large monitor or projection panel,teachers can model writing processesand discuss strategies for planning andrevising. As with the other features ofword processing, the impact of greatercollaboration depends on the instruc-

346 JOURNAL OF LEARNING DISABILITIES

tional program and the skill of theteacher.

Finally, typing is substantially dif-ferent from handwriting. Typing isprobably inherently easier than hand-vi'riting, especially for students withhandv '̂riting problems. On the otherhand, typing can also be a barrier, asit is not a standard part of curricula.Students need some typing instruc-tion if they are to use word proces-sors regularly {MacArthur et al., 1993).

A relatively large number of stud-ies have investigated the use of wordprocessing with nondisabled popu-lations. A recent meta-analysis byBangert-Drowns (1993) found that useof word processing in writing instruc-tion programs had a positive, thoughrelatively modest, impact on students'writing. Cochran-Smith (1991) con-cluded that students have positive at-titudes toward word processing butthat the impact of computers on thequality of students' writing and writ-ing processes depends on teachers'strategies for using word processing,and on the social organization of theclassroom. MacArthur (1988) dis-cussed the potential benefits and prob-lems associated with using wordprocessing with students with LD andreviewed the limited and inconclusiveresearch with that population. Morerecent studies (MacArthur et al., 1993;Morocco, Dalton, & Tivnan, 1990) haveindicated that word processing com-bined with effective writing instruc-tion can enhance the writing of stu-dents with LD.

A growing number of writing sup-port tools are available that go beyondword processing. The next section fo-cuses on computer tools that supportthe basic writing processes involvedin transcription and sentence genera-tion.

Sentence Generationand Transcription

For competent adult writers, thebasic processes of formulating gram-matically correct sentences and tran-

scribing them into written languageare relatively automatic. In fact, evi-dence from studies comparing dicta-tion and handwriting suggests that fornormally achieving students, the me-chanics of writing cease to be a limit-ing factor by the end of elementaryschool. The dictated stories of primary-grade children are superior to theirwritten stories (King & Rentel, 1981),but by fifth or sixth grade, dictatedcompositions, although longer, are notqualitatively better than handwrittenones (Hidi & Hildyard, 1983; Scar-damalia, Bereiter, & Goelman, 1982).In contrast, the dictated compositionsof students with LD have been re-ported to be substantially longer andqualitatively superior to their com-positions written via handwriting orword processing (Graham, 1990;MacArthur & Graham, 1987). The diffi-culties of students with LD with tran-scription processes—spelling, capitali-zation, punctuation, and usage—arewell documented (Graham et al., 1991).The dictation studies suggest that thesedifficulties interfere with the overallcomposing process. This interferencemay take a number of forms (e.g., stu-dents may avoid using words theycannot spell). The effort devoted tomechanical issues may reduce thecognitive capacity available for plan-ning and revising processes. Studentsmay also write less because of theeffort involved or because of low self-confidence.

These sentence-generation and tran-scription processes are importantthroughout the stages of writing. Con-sequently, this section discusses com-puter tools designed for use duringboth initial drafting and revising.

Spelling Checkers

The most widespread and generallyuseful tools to support transcriptionare spelling checkers. Nearly allword processors designed for adults,and most recent versions targeted atschools, include an integrated spell-ing checker that can be accessed with-out leaving the word processor.

Spelling checkers perform two func-tions: They identify misspelled words,and they suggest correct spellings.

Although clearly useful, spellingcheckers do have limitations, espe-cially for students with serious spell-ing problems (Dalton, Winbury, &Morocco, 1990). Two of those limita-tions pertain to the identification ofmisspellings. First, spelling checkersflag proper nouns and special termsas errors. Second, and more impor-tant, they fail to flag niisspelled wordsthat are other words correctly spelled,including homonyms and "other cor-rect words" (e.g., back for bake or whetfor went). In two recent studies,MacArthur, Graham, and De La Paz(in press) found that about 26% to 38%of spelling errors made by fourth-through eighth-grade students withLD fit into this second category. Dalton(1988) reported that approximately40% of spelling errors made by fourth-grade students with LD were not iden-tified by spelling checkers. Similarresults have been reported for non-disabled students (Mitton, 1987). Thisproblem might be ameliorated bysmaller or adjustable-size dictionarieshaving fewer uncommon words.

Two potential limitations pertain tosuggesting correct spellings. First,spelling checkers fail to suggest thecorrect spelling for many words, espe-cially severe misspellings. Differentcheckers vary in their ability to sug-gest the correct word; for example,MacArthur, Graham, and De La Paz(1994) reported that eight spellingcheckers found the correct spelling for46% to 66% of the words the checkersflagged as misspelled. Second, evenwhen the spelling checker suggests thecorrect word, students with LD maynot be able to identify the correct wordfrom the list. Spelling checkers con-vert the writer's task from producingthe correct spelling to recognizing itfrom among a list of similar words.This recognition task can be difficultfor poor spellers, especially if the listis long. Of course, a trade-off existsbetween the length of the list and inclu-sion of the correct word. Two soft-

VOLUME 29. NUMBER 4, JULY 19% 347

ware design strategies are availableto help with this issue: synthesizedspeech to pronounce the words in thelist, and definitions of words in thelist.

A recent study of middle school stu-dents with LD who had moderate tosevere spelling problems provideddata on the overall usefulness of spell-ing checkers (MacArthur, Haynes, &Graham, 1994). Twenty-six studentswrote stories and revised their spell-ing using a spelling checker. Theymisspelled 4% to 35% of their words.The spelling checker flagged 63% oftheir errors, missing 37% that werehomonyms or other correct words. Thecorrect spelling was suggested for 58%of the flagged words, or 36% of allerrors. Students were able to correct82% of the errors with correct sugges-tions and 23% of the errors when thecorrect suggestion was not offered.Overall, students corrected 36% oftheir errors using the spelling checker.

A final issue about spelling check-ers is whether they can be used topromote greater spelling skill as wellas to directly compensate for poorspelling. One instructional techniqueis to have students attempt to locatemisspelled words and circle them ona printout prior to using the spellingchecker. Handheld spelling checkers,which require students to identifypotentially misspelled words and typethem on the checker, may encouragethis strategy. Another approach is toteach students to use their knowledgeto try alternate spellings prior to rely-ing on the computer-generated list ofsuggestions. These strategies have thepotential to transfer to editing spell-ing errors without a computer.

Speech Synthesis

Speech synthesis software (or hard-ware) translates text into speech. It isnot as natural-sounding as digitizedspeech, which is recorded, but its ad-vantage is that it can be used to speakany text. Word processors with speechsynthesis enable students to hear whatthey have written and to read what

others have written; this capabilitymay support writing by allowing stu-dents with writing problems to usetheir general language sense to moni-tor the adequacy of their writing(Rosegrant, 1986). For example, stu-dents may notice incomplete or awk-ward sentences, misspelled words, orerrors of meaning. In an instructionalcontext that focuses on meaningfulcommunication, talking word proces-sors may help bridge the gap betweenwhat children want to express andwhat they have the skills to read andwrite. Speech synthesis can scaffoldboth reading and writing, for example,by helping students read language-experience stories and the writings oftheir peers.

Research investigating the potentialof speech synthesis to improve writ-ing among elementary-school childrenis limited. Borgh and Dickson (1992)compared word processing with andwithout speech synthesis with non-disabled second- and fifth-grade stu-dents. Both versions of the wordprocessor incorporated a specialprompting feature: Each time a periodwas typed, signaling the end of a sen-tence, a prompt appeared on the screenreminding the student to reread thesentence and consider revising it. Stu-dents did more revising after eachsentence with the speech synthesis andless revising at the end of the story.No differences were found in lengthor quality of writing.

Research on and adoption of speechsynthesis have been slowed by prob-lems with the quality of the speech.High-quality speech synthesis, usingexpensive hardware (e.g., DECtalk, nodate), is nearly as comprehensible asrecorded speech, but less expensivehardware has been found to have se-rious comprehensibility problems(Mirenda & Beukelman, 1987). Thequality of less expensive software-onlyspeech synthesizers has been improv-ing, though they still do not rival thecomprehensibility of digitized (re-corded) speech.

Currently, a number of word pro-cessors and related writing tools that

have speech synthesis are available.The Talking Textwriter (no date) wordprocessing software has been avail-able for some time for both Apple IIand IBM; however, it uses the rela-tively poor Echo speech synthesizer.My Words (1993) provides a straight-forward word processor with a vari-ety of text-to-speech options; it willread letters, words, sentences, or thefull text. Write:Outloud (1993) pro-vides speech synthesis with similaroptions for reading letters, words, sen-tences, or the full text. In addition, itprovides spell checking.

Word Prediction andWord Banks '

Word prediction was originally de-veloped for individuals with physicaldisabilities to reduce the number ofkeystrokes required to type words andsentences. However, it may have po-tential for students with serious prob-lems with spelling, punctuation, andsyntax, as well. A brief description ofone word prediction program willprovide information about the basicfunctions (see Figure 1). Co:Writer(1992) is a commercially available pro-gram for Macintosh computers thatsupports word prediction for any wordprocessor. The user types in a win-dow on top of the word processor. Aseach letter is typed, the software pre-dicts the intended word (offering alist of adjustable length). If the in-tended word is in the list, the user cantype the number of the word or pointand click with the mouse to insert theword in the sentence. If the correctword is not present, the user contin-ues to type the next letter and so on.Predictions are based on spelling, syn-tax, and the words previously usedby particular users. Speech synthesisis available to read the words in thelist of suggestions, as well as the com-pleted sentence. When the sentence iscomplete, it is transferred into theword processor.

A related type of software is wordprocessors that include word banks.

348 JOUKNAL OF LEARNING DISABILITIES

w File Edit Format Options Font Speech

The Frog PrinceRuniiiuuiiiuiuuuii ijjjj,^ •iipiiaii

Once upon a time in a far off land, therelived a beautiful princess. Though she waslovely indeed, she was not well liked becauseshe thought only of herself.

CoiLUriter Demo Ulriter

One day she was playing in the g1:good2: game3:garden

FIGURE 1. CoiWriter word prediction software with a word processor in the background. (The user has just typed "g"and the software has suggested three possible words.)

My Words (1993), for example, is aword processor with speech synthesisand a word bank (see Figure 2). In thedefault operation, all words typed inthe word processor are collected in analphabetized list of words at the sideof the screen. This list can also beedited and locked separately from thewriting. As a user types, the list auto-matically scrolls to find words be-ginning with the letters typed. Forexample, as the student types a t, thelist scrolls to the first word startingwith I. The student can continue totype or select a word from the list byclicking with the mouse. Such pro-grams provide a limited version ofword prediction in a format that maybe less intrusive than word predic-

tion programs designed for individ-uals with physical disabilities.

Another approach to providingvocabulary support for young childrenis the use of word and picture banks,as in Kid Works 2 (1992). This pro-gram includes "boxes" of nouns, verbs,and adjectives. Each word is accom-panied by a picture and can be pro-nounced by speech synthesis. Studentscan incorporate these words into sto-ries as pictures to create rebus stories,or can translate them into words.

For students with LD, word predic-tion software and word processorswith word banks assist transcriptionduring the process of writing, ratherthan during revision. They have thepotential to support spelling, capitali-

zation, and sentence formation. Thispotential is untested; I am not awareof any empirical research on wordprediction with students with LD. Ourresearch group has conducted prelimi-nary studies with word prediction andword bank software, but no resultsare available at this time.

Grammar and Style Checkers

Grammar and style checking soft-ware goes beyond spelling to checksyntax, sentence structure, punctua-tion and capitalization, and writingstyle. Several sophisticated style check-ers are on the market for college andadult writers, but reviewers seem todisagree on whether the advice they

VOLUME 29, NUMBER 4, JULY 1996 349

provide is helpful. In any case, theyappear to be of limited value forelementary-school students or poorwriters. They successfully identify rela-tively minor grammatical and stylis-tic problems, but often do not interpretserious grammatical and mechanicalerrors correctly. One program. WriteThis Way (Emerson & Stern Assoc,1992), was designed specifically tomeet the needs of students with LD.It consists of a basic word processorwith speech synthesis, spell checking,and grammar checking. Unfortunately,the grammar checker does not appearto be successful at identifying errorsin the writing of students with or with-out LD. In an informal review (Mac-Arthur, 1994), the checker was usedto proof 10 writing samples written

by elementary-school students withLD and those same papers with spell-ing, punctuation, and grammaticalerrors corrected. It did not flag themajority of grammatical errors in theuncorrected writing versions, and itincorrectly flagged many errors in thecorrected versions. When it did locatean error, its diagnoses were often dif-ficult to comprehend. In summary, atthis time I am not aware of any usefulgrammar-checking software for stu-dents with LD.

Planning Processes

Experienced writers typically devotea substantial portion of writing timeto planning activities. They set goals

in terms of the intended audience,generate content through memorysearch and information gathering, andorganize their material carefully(Flower & Hayes, 1981). Students withLD may have difficulty with all ofthese component processes (Englertet al., 1989; Graham et al., 1991). Typi-cally, they begin writing after devot-ing minimal time to planning. Theyoften have problems generating suffi-cient appropriate content and thusproduce short compositions with lim-ited information. They typically lackawareness of common text structuresthat could help them in organizingmaterial and in generating more con-tent. In addition, limited backgroundknowledge may interfere with boththeir reading comprehension and their

w File Edit Story Font Size Style Color

UioidsthatthethemThenTheythingthoughttotootreostreesupwaswasn'twhenwokeyellyelling

Sounds «car crashparakeetlaughtertarzan yell

O

O

iThe Dream

The funniest that happened to me in the 8thgrade was when 1 fell asleep in class,ik the classwas social studies and Mrs. harrison was theteacher. They laughed.«l She was talking aboutthe history of slaves in Africa. Tarzan wassaving the African people. I started to dreamabout Tarzan. He looked like a movie star andwa5 saving all the African people by swingingdown from the trees and yelling his yell that wasscaring them.^k Then I thought I heard Tarzancalling me so I came down from the treas and amonkey came too. The class I enjoy the most ismusic. When I woke up , I was yelling out aTarzan's call and everybody heard it. Theylaughed. The teacher was yelling at me for

FIGURE 2. My Words, a word processor with speech synthesis and a word bank.

350 JOUEiNAL OF LEARNING DISABILITIES

writing (Garner, Alexander, & Hare,1991).

This section considers several typesof software that have the potential tosupport these students' planning pro-cesses and facilitate their access tobackground knowledge. Revising pro-cesses, which also present problemsfor poor writers, were discussed ear-lier, in the sections on word process-ing and transcription support.

Prompting

The interactive capabilities of com-puters can be used to develop pro-grams that prompt writers to engagein planning processes, by asking thema series of questions or presentingreminders. The most common prompt-ing software presents a series of ques-tions designed to help writers generateideas prior to writing. These questionscan be designed around particular textstructures. For example, for a newsarticle, a program might prompt stu-dents with "who, where, why, andwhen" questions; for a story, it mightprompt them with questions aboutcharacters, problem, action, and reso-lution. Writer's Helper (1990) containsa large collection of interactive plan-ning programs, including programsthat support brainstorming, freewrit-ing, and categorizing, as well as struc-tured questions. At a somewhat iowerlevel of interactivity, several wordprocessors permit teachers to enterseries of questions that can be locked.Students write their answers betweenthe questions.

I am not aware of any research onthe impact of using such planningprograms. However, some researchindicates that simple text-structureprompts may enhance the writing ofstudents with LD (Montague, Graves,& Leavell, 1991). Prompting programsmight also be used in conjunction withstrategy instruction based on textstructures (Graham et al., 1991).

Research has been conducted onsoftware programs that provideprompts during the composing pro-cess to remind students to engage in

planning or revising strategies.Salomon (1992) developed a specialwriting program that provided guid-ance (in the form of questions) before,during, and after writing. For example,questions before writing concernedaudience, purpose, and content; ques-tions during writing addressed elabo-ration, organization, explicitness; andquestions after writing posed evalua-tion issues. Salomon reported that thequality of writing produced by stu-dents using this prompting programimproved, and that those gains gen-eralized to writing with paper andpencil. Daiute (1986a) found that aprogram that prompted students torevise during composing was effec-tive in increasing the amount of revi-sion.

Outlining and SemanticWebbing

Outlining and semantic webbing arecommon practices for organizing ideasprior to writing, both in school andamong experienced writers. Many so-phisticated adult word processorsinclude outlining capabilities. Farlyprograms for semantic mapping wererestricted by screen size and limitedgraphics capabilities. A sophisticatedprogram for semantic webbing. Inspi-ration (1994), is available for Macintoshcomputers (see Figure 3). The programpermits the creation of semantic webson the screen with elements that canbe easily rearranged for experimenta-tion with different arrangements ofideas. Hidden notes can be attachedto fhe main ideas in the web. The en-tire web, including notes, can be au-tomatically converted into an outlineprior to writing. Most of the semanticwebbing activities that teachers useon paper can be carried out on thecomputer using this program. Forexample, teachers can create blankwebs that model particular text struc-tures (e.g., compare/contrast, descrip-tion). Whether the advantages offlexible rearrangement of ideas andneatness outweigh the simplicity of

webbing on paper is an open ques-tion and probably depends on a num-ber of student, task, and instructionalfactors. I am not aware of any researchon the use of computers to supportsemantic webbing.

Multimedia

The potential of multimedia soft-ware to enhance writing processes isjust beginning to be explored as newsoftware tools are developed. Al-though multimedia can also serve asa new means of publication and helpto compensate for weak basic skills, ifis discussed in this section on plan-ning processes because it has the po-tential to promote the generation ofideas and provide background knowl-edge for planning. For the purposesof this review, multimedia includesprograms that integrate drawing toolswith writing as well as programs thatinclude video and sound.

For young children, "writing" oftenconsists primarily of drawing pictures.Children learn how to tell stories, de-scribe their experiences, and explorewhat they know by drawing and talk-ing about their drawings. Some of themost popular programs for homemarkets are drawing and writing pro-grams for children, such as Kid Pix(1992) and Kid Works 2 (1992). KidPix has a wide variety of fanciful draw-ing tools but limited text capabilities;Kid Works 2 permits children to cre-ate pictures using a palette of draw-ing tools and to write stories to ac-company their pictures; it will thenplay back the story, displaying thepictures and reading the story withsynthesized speech. With other soft-ware, children can create visual envi-ronments that serve as backgroundsfor story writing. For example. Story-book Weaver (1992) provides back-ground scenes, objects, and animatedfigures for children to use in creatingpictures. As they write the stories togo with the pictures, they can rely ona large picture dictionary for wordsthey do not know how to spell. Sounds

VOLUME 29, NUMBER 4, JULY 1W6 351

COQ .

JZ

Q .COQ .

to

coo

COQ .

EoCO

tm.

_oQ.

E

Q.

EmX0}

ECO

o

c

3

ocCO

D)CQ.

aE

iD

352 JOURNAL OF LEARNING DISABILITIKS

and music can be added to theirstories. The Explore-a-Science series(1993) includes science topics (e.g.,whales) for expository writing.

Older students are generally ex-pected to convey their ideas via writ-ing without the support of drawing.However, students at any age benefitfrom the use of visuals and othermedia in preparation for writing. Thisbenefit may be especially importantfor students with more limited literacyskills or prior knowledge in a particu-lar area, due to an educational dis-ability or a different cultural andlinguistic background (Daiute, 1992).Daiute reported on a multimediaproject with students with poor writ-ing skills from diverse cultural back-grounds. Students collected visualsand sounds that had personal mean-ing to them (e.g., snapshots, music),wrote about these images and sounds,and used simple multimedia softwareto combine them with their text. Thisdescriptive study provided anecdotalevidence that multimedia can moti-vate students to write more and helpthem find new ways to express them-selves.

Although multimedia has the poten-tial to enhance writing, there are draw-backs as well. The addition of graphicsand sounds to compositions mayresult in reduced focus on text—thetext becomes relatively less importantin carrying the meaning. Creatinggraphics and sounds also requires timeand attention, which may detract fromattention to the written text. Bahr,Nelson, and Van Meter (1994) investi-gated the effect of using graphics-based writing software on the writingof fourth- through eighth-grade stu-dents with LD. The software allowedstudents to create scenes and then typestories about those scenes. The authorsconceptualized the graphics featuresas an aide to planning what to write,and compared this software to text-based planning software that pre-sented questions based on storygrammar. Students typed answers tothese questions and used their re-sponses as plans for their stories. The

stories produced using the two typesof software were compared on narra-tive maturity (a measure of overallquality) and several quantitative indi-cators related to length (e.g., numberof words and T-units). No significantdifferences were found between thetwo conditions; however, these resultsshould be interpreted with caution, asonly 9 students participated. Clearly,further research is needed on the ad-dition of graphics features to writingsoftware.

Multimedia may have the potentialto enhance reading and writing incontent-area tasks as well. Multime-dia can be used to extend backgroundknowledge and to encourage studentsto explore their ideas prior to writing.Research on anchored instruction,which uses video to provide a mean-ingful, shared context for learning, hasdemonstrated the potential of multi-media materials to make informationmore usable for problem solvingby connecting new learning to back-ground knowledge (Cognition andTechnology Group, 1993). In oneproject, researchers at Vanderbilt de-veloped a multimedia composing toolthat uses video to provide backgroundinformation on a particular topic as abasis for writing activities for individu-als with reading and writing disabili-ties (Hasselbring & Goin, 1991). Afterviewing a video segment, studentsreceive graduated supports in read-ing and writing about the topic pre-sented in the video; they can returnto the video at any time to explorethe content further. Another projectat Vanderbilt (Kinzer, Hasselbring,Schmidt, & Meltzer, 1990) explored theuse of video news reports to teach textstructures and how to use the struc-tures in reading and writing. Studentslearned about the typical structure ofnews reports (i.e., "who, what, when,where, why, and how" questions) andhow to use that structure in note tak-ing, brainstorming, and revising. Stu-dents' comprehension of television andoral news reports improved, and theyincluded more structural elements intheir written news reports.

Collaborative Writing andPublishing with Networks

As mentioned earlier in the sectionon word processing, computers canchange the social context for writingby supporting publishing and collabo-rative writing in the classroom. For-mats for publishing can be furtherextended with multimedia software.In addition, networks (both local areanetworks within a school and telecom-munications networks) can offer ex-panded opportunities for collaborativewriting and communication with di-verse audiences.

In one early educational applicationof telecommunications, Riel (1985)developed an electronic newswire thatinvolved students from geographicallydiverse cultures in collaborative pro-duction of a newspaper. Because thestudents from California and Alaskadid not share the same cultural knowl-edge, they had to struggle to commu-nicate clearly. Using a similar networkwith seventh-grade students fromIsrael, Cohen and Riel (1989) con-ducted a study to explore the effectsof writing for authentic audiences ofpeers from different cultural back-grounds. They reported that essayswritten by students for distant peerswere superior to essays written to begraded by their teachers, because theformer were more explicit and de-tailed. Telecommunication networksfor use by schools are supported by anumber of state and national organi-zations.

Local area networks can also be usedto support communication in writing.Peyton and Batson (1986) used a net-work within a classroom to teach writ-ing to students with hearing impair-ment whereby all discussion andinteraction were conducted in writ-ing. Students viewed the ongoing con-versation on their computer screensand participated by typing their re-marks. For students with hearing im-pairment, the network provided animmersion approach to masteringFnglish. Students with LD may alsobenefit from writing on such a net-

VOLUMb 29, NUMBER 4, JULY 353

work, The network interaction canchange the social context for writingby providing a connection betweenconversation and formal writing.

Concluding Comments

As the capabilities of computershave increased in the past decade, avariety of exciting new tools have beendeveloped that have the potential toenhance the writing of students withLD. Basic tools, such as spelling check-ers, have become common even onsimple word processors. Printing fea-tures and desktop publishing havebecome more powerful and easier touse. The quality of speech synthesishas improved, and the variety of pro-grams providing speech has expanded.Word prediction software has becomeavailable to support access to writingand reading. Telecommunications net-works are accessible to schools will-ing to invest in modems and a phoneline. Multimedia programs that inte-grate drawing and writing are widelyavailable, and programs that integratephotographs, video, and sound withwriting will become increasingly avail-able in the next few years.

The challenge for special educatorsis twofold: First, existing research onword processing makes it clear thatsimply providing technology to teach-ers and students will not result inimprovements in students' writing.Effective instructional methods mustbe developed that make use of thepower provided by these tools toenhance the writing of students withLD. Second, as this review reveals, re-search on computers and writing hasbeen limited primarily to studies ofthe effects of basic word processing.Researchers need to go beyond wordprocessing to investigate the effectsof instruction using a range of tech-nological tools to support writing. Thisreview has attempted to provide aframework for both the developmentof instructional methods using tech-nology and research on their effective-ness. Teachers, administrators, teacher

educators, and researchers need tocollaborate in this effort to transformthe potential of technology into reality.

ABOUT THE AUTHOR

Charles A. MacArthur, PhD. is an associateprofessor in the Department of EducationalStudies at the University of Delaware. Hisresearch interests are in literacy, strategy in-struction, and technology applications withstudents with learning disabilities. Address:Charles A. MacArthur, Department of Educa-tional Studies, University of Delaware, New-ark. DE 197U.

AUTHOR'S NOTE

Preparation of this article was partially sup-ported by a grant from the U. S. Departmentof Education (No. m80G30033). However,the opinions expressed are those of the aulhorand no official endorsement is implied.

REFERENCES

Bahr, C. M., Nelson, N. W., & Van Meter,A. (1994). Planning to write: Comparisonof a text-based versus a graphics-based tool.Manuscript submitted for publication.

Bangert-Drowns, R. L. (1993). The wordprocessor as an instructional tool: Ameta-analysis of word processing inwriting instruction. Review of EducationalResearch. 65. 69-93.

Becker, H. J. (1993). Teaching with andabout computers in secondary schools.Communications of the ACM, 36(5), 69-72.

Borgh, K., & Dickson, W. P. (1992). Theeffects on children's writing of addingspeech synthesis to a word processor.lournal of Research on Computing in Edu-cation. 24. 533-544.

Calkins, L. M. (1991). Living hetween thelines. Portsmouth, NH: Heinemann.

Cochran-Smith, M. (1991). Word process-ing and writing in elementary class-rooms: A critical review of related litera-ture. Review of Educational Research. 61,107-155.

Cognition and Technology Group atVanderbilt Learning Technology Center.(1993). Integrated media: Toward a theo-retical framework for utilizing theirpotential, journal of Special Education Tech-nology, 32(2), 71-85.

Cohen, M., & Riel, M. M. (1989). The effectof distant audiences on students' writ-

ing. American Educational Research jour-nat. 26. 143-159.

Co:Writer ]Computer program]. (1992).Wauconda, IL: Don Johnston Develop-mental Equipment.

Daiute, C. (1986a). Physical and cognitivefactors in revising: Insights from studieswith computers. Research in the Teachingof English, 20, Ul-\59.

Daiute, C. (1986b). Do 1 and 1 make 2?Patterns of influence by collaborativeauthors. Written Communication, 3. 382-408.

Daiute, C. (1992). Multimedia composing:Extending the resources of kindergartento writers across the grades. LanguageArts. 69. 250-260.

Dalton, B. M. (1988). A comparative study offive spell checkers' analyses of learning dis-abled and normally achieving fourth gradestudents' written compositions. Unpub-lished manuscript. Harvard GraduateSchool of Education, Cambridge, MA.

Dalton, B., Winbury, N., & Morocco, C. C.(1990). "If you could just push a but-ton": Two fourth-grade boys with learn-ing disabilities learn to use a computerspelling checker. Journal of Special Edu-cation Technology, 10. 177-191.

DECtalk IComputer hardware], (no date).Maynard, MA: Digital Equipment Corp.

Emerson & Stern Associates. (1992). WriteThis Way [Computer programl. Petaluma,CA: Interactive Learning Materials.

Englert, C. S., Raphael, T. E., Anderson,L. M., Anthony, H. M., & Stevens, D. D.(1991). Making writing strategies andself-talk visible: Cognitive strategy in-struction in writing in regular and spe-cial education classrooms. AmericanEducational Research Journal. 28, 337-372.

Englert, C. S., Raphael, T. E., Anderson,L. M., Gregg, S. L., & Anthony, H. M.(1989). Exposition: Reading, writing, andthe metacognitive knowledge of learn-ing disabled students. Learning Disabili-ties Research, 5, 5-24.

Explore-a-Science [Computer program].

(1993). Acton, MA: Bradford.Fitzgerald, J. (1987). Research on revision

in writing. Review of Educational Research.57. 481-506.

Flower, L., & Hayes, ]. R. (1981). A cogni-tive process theory of writing. CollegeComposition aitd Communication, 32, 365-387.

Garner, R., Alexander, P. A., & Hare, V. C.(1991). Reading comprehension failurein children. In B. Y. L. Wong (Ed.), Learn-ing about learning disabilities (pp. 284-307).San Diego: Academic Press.

354 JOURNAL OF LEARNING DISABILITIES

Graham, S. (1990). The role of productionfactors in learning disabled students'compositions. Journal of Educational Psy-chology, 82,78-1-791.

Graham, S., Harris, K., MacArthur, C. A.,& Schwartz, S. S. (1991). Writing andwriting instruction with students withlearning disabilities: A review of a pro-gram of research. Learning DisabilityQuarterly, 14,89-114.

Graham, S., & MacArthur, C. (1988). Im-proving learning disabled students' skillsat revising essays produced on a wordprocessor: Self-instructional strategytraining. The Journal of Special Education,11, 133-152.

Hasselbring, T., & Goin, L. (1991). Enhanc-ing writing through integrated media.The Writing Notebook. 9{1), 27-29.

Hidi, S., & Hildyard, A. (1983). The com-parison of oral and written productionsof two discourse types. Discourse Pro-cesses, 6,91-105.

Inspiration (Version 4.0) [Computer pro-gram]. (1994). Inspiration Software.

Kid Pix [Gomputer program). (1992).Novato, CA: Broderbund.

Kid Works 2 [Computer program]. (1992).Torrance, CA: Davidson & Associates.

King, M. L., & Rentel, V. M. (1981). Re-search update; Conveying meaning inwritten texts. Language Arts, 58. 721-728.

Kinzer, C. K., Hasselbring, T. S., Schmidt,C. A., & Meltzer, L. (1990, April). Effectsof multimedia to enhance writing ability.Paper presented at the annual confer-ence of the American Educational Re-search Association, Boston.

MacArthur, C. A. (1988). The impact ofcomputers on the writing process. Excep-tional Children, 54. 536-542.

MacArthur, G. A. (1994). [Review of agrammar checker designed for studentswith learning disabilities]. Unpublishedraw data.

MacArthur, C. A., & Graham, S. (1987).Learning disabled students' composingunder three methods of text production:

Handwriting, word processing, and dic-tation. The Journal of Special Education,21(3), 22-42.

MacArthur, G. A., Graham, S., Haynes,J. A., & De La Paz, S. (in press). Spellingcheckers and students with learning dis-abilities: Performance comparisons andimpact on spelling. The Journal of SpecialEducation.

MacArthur, G. A., Graham, S., & Schwartz,S. (1991). Knowledge of revision andrevising behavior among learning dis-abled students. Learning Disability Quar-terly, 14. 61-73.

MacArthur, G. A., Graham, S., & Schwartz,S. S. (1993). Integrating word processingand strategy instruction into a processapproach to writing. School PsychologyReview, 11,671-681.

MacArthur, G. A., Schwartz, S. S., & Gra-ham, S. (1991). Effects of a reciprocal peerrevision strategy in special educationclassrooms. Learning Disabilities Research& Practice, 6, lQ-\-l\O.

Mirenda, P., & Beukelman, D. R. (1987).Gomparison of speech synthesis intelli-gibility with listeners from three agegroups. Augmentative and Alternative Com-munication, 3, 120-128.

Mitton, R. (1987). Spelling checkers, spell-ing correctors and the misspellings ofpoor spellers. Information Processing andManagement, 23, 495-505.

Montague, M., Graves, A., & Leavell, A.(1991). Planning, procedural facilitation,and narrative composition of junior highstudents with leaming disabilities. Learn-ing Disabilities Research & Practice, 6,219-224.

Morocco, C, Dalton, B., & Tivnan, T. (1990,April). The impact of computer-supportedwriting instruction on the writing qualityof 4th grade students with learning disabili-ties. Paper presented at the annual meet-ing of the American Educational Re-search Association, Boston.

Morocco, G. C , & Neuman, S. B. (1986).Word processors and the acquisition of

writing strategies, journal of LearningDisabilities, 19, 243-247.

My Words [Computer program]. (1993).Dimondale, MI: Hartley.

Peyton, J. K., & Batson, T. (1986). Com-puter networking: Making connectionsbetween speech and writing. ERIC/CLLNews Buiietin, 10(1), 1, 5-7.

Riel, M. M. (1985). The computer chroniclesnewswire: A functional learning envi-ronment for acquiring literacy skills, jour-nal of Educational Computing Research, 1,317-337.

Rosegrant, T. J. (1986, April). It doesn't soundright: The role of speech output as a primaryform of feedback for beginning text revision.Paper presented at the annual meetingof the American Educational ResearchAssociation, San Francisco.

Salomon, G. (1992, February). Metacognitivefacilitation and cultivation during essaywriting: The case of the " Writing Partner."Paper presented at the Third Interna-tional Conference on Cognitive Educa-tion, Riverside, CA.

Scardamalia, M., Bereiter, C, & Goelman,H. (1982). The role of production factorsin writing ability. In M. Nystrand (Ed.),What writers know: The language, process,and structure of written discourse (pp. 173-210). New York: Academic Press.

Stoddard, B., & MacArthur, C. A. (1993).A peer editor strategy: Guiding learningdisabled students in response and revi-sion. Research in the Teaching of English,17, 76-103.

Storybook Weaver [Computer program].

(1992). Minneapolis, MN: MECG.

Talking Textwriter [Gomputer program].

(no date). New York: Scholastic.

Write:Outloud [Gomputer program].(1993). Wauconda, IL: Don JohnstonDevelopmental Equipment.

Writer's Helper [Computer program).(1990). Iowa City, IO: CONDUIT, Uni-versity of Iowa.