Numerous studies have shown that students transfer pen-and-paper writing procedures into their computer-assisted writing processes. As Gail Hawisher (1988) states in her review of 11 case studies of computer-assisted composition:

A general theme emerging from these studies is similar to a conclusion drawn from the experimental research [the author reviewed 18 experimental studies]. That is, a writer's or student's particular habits and strategies for composing apparently take precedence over the machinery. (p. 12)

The quest to successfully integrate the computer into a process-centered pedagogy has led to significant debate (Curtis, 1988). One of the important controversies is whether word -processing concepts need to be emphasized along with writing in a computer-assisted composition class. Of course, we have given our students at least some instruction on using word-processing packages. But agreeing upon the kind, amount, and sequencing of such instruction has been problematic. The policy of devoting a significant amount of instructional time solely to teaching students to use computer hardware and software (e.g., word-processing packages) has come under substantial attack (Euchner, 1983; Frisch, 1987; Hollifield, 1983). In 1985, Elizabeth Sommers and James Collins concluded that "the heart of the matter is to teach writing, not word-processing" (p.27). Yet, as writing teachers who use computers, we know that mastery of word-processing programs facilitates strategies that are difficult with pen and paper, if not impossible.

Sometimes the problem of not mastering a word-processing program is acute. Robert Mack, Clayton Lewis, and John Carroll (1983) found that 10 experienced office temporaries were unable to type in, revise, format, and print out a one-page letter after two full days of working with instructional manuals. After considering these findings, Carroll and Mack (1984) concluded that although the manuals were designed for passive learning, human beings learn actively--by doing, thinking, and knowing. Richard Bullock (1985) found that basic writers more successfully integrated the word processor into their composing processes after he provided them with word-processing instruction. He concluded that keyboard fluency is an important issue and that instructors of computer-assisted composition may, among other things, need to provide typing tutorials (pp. 51, 54). In another study, Colette Daiute (1986) found that junior high school students who had received both typing and word-processing instruction performed better on the final drafts of essays than other students. It would seem, then, that some instruction in word processing is important, but we do not yet know how much to provide for students who are also struggling with writing problems.

It may very well be necessary to accommodate the physical dimension of computer-assisted composing because, if writing is a mode of knowing, whatever interferes with inscription also interrupts thinking. In fact, one might argue that the following description of computer use in a basic writing classroom calls into question whether or not we can indeed afford to leave students using word processors largely "to their own devices."

The study described below was conducted over a 10-week period. The purpose of the study was to assess how three basic writers adapted word processing to their writing.

Rather than concerning itself chiefly with outcomes and products this qualitative study focused upon the composing processes of students in one section of computer-assisted basic writing at Ohio State University (OSU). Three volunteer subjects, Fred, Karnak, and Ann, provided the data for the study (all subjects have been given pseudonyms). In order to gather data, I observed all classes, nearly all of the students' interactions with word processors, and several of their composing sessions in their dormitories. The students worked with word-processing software only in an OSU computer lab. My role in the class was what Stephen Doheny-Farina and Lee Odell (1985) identify as "participant as observer," a researcher who interacts with subjects only enough to establish himself or herself as an acceptable presence and to clarify any information received.

My data included all drafts of each assigned essay, notes and observations of students' writing, and taped and written interviews of the subjects and their instructors (see Appendix). By including several kinds of data and multiple perspectives on the subjects, the study was able to produce a triangulated view of the activities observed. Data was analyzed by factoring, making metaphors, splitting variables, and other techniques suggested by Matthew Miles and Michael Huberman (1984).

At the time of the study, Ohio State was an open-admissions university. I observed a section of the second of a two-course series for students who scored below 16 on their ACTs and who wrote placement essays that exhibited significant incoherence and/or other serious problems. The "process-centered" curriculum required students to submit several drafts of each essay, and class sessions in the section I observed were most often workshops featuring both instructor-student and peer editing. Sample essays as well as clause and punctuation rules were discussed in class.

The course was the third computer-assisted section that Carol, the instructor, had taught. A doctoral student in English at Ohio State, she had previously taught a class using WYLBUR and also a microcomputer-assisted class. The examined class met three times a week but only once during the quarter in the computer lab. During the third classroom meeting, Carol explained the functions of the software and assigned BANKSTREET WRITER'S instructional handout as homework. The next class session was held in the computer lab. During this session, Carol gave hands-on instruction with the Apple 2e s and 1982 BANKSTREET WRITER, directing students step-by-step how to start up the machines, write, scroll, erase, manipulate the cursor, cut, paste, transfer modes, and save a text. Carol then told students to practice revising on screen, moving from the WRITE mode to the EDIT mode. The 1982 version of the software required one to enter the EDIT mode before doing any on-screen revising.

For the rest of the 10-week quarter, the 15 students worked with the computers during the 29 1/2 hours of open lab time. Students were required to use the word processor at least as a transcription tool to print each submitted draft, but their word-processing skills were not observed or evaluated. After the initial two instructional sessions, students were given no formal guidance and little informal guidance in integrating the computer into their composing processes. Thus, their largely "natural" approaches to the machine were observed.

Fred, who grew up on a farm, was a dairy science major. The snowmobiling accident he described in his essay resulted in his missing half of a year in high school. His score on the ACT was 9, and his grades placed him in the 18th percentile of his graduating class. While Fred had taken typing in high school, he was not proficient in this skill by the time that he reached college.

Karnak's ACT score was 14; a weak placement essay caused him to be put into developmental English. Born in India, Karnak had come to the United States before his first birthday but had returned to India for two years during junior high school, two years in which he had little exposure to English. A capable student fluent in four languages, Karnak was majoring in chemical engineering at Ohio State. Although he had not had typing instruction, he had taken a semester-long programming class at his affluent suburban American high school, where he graduated in the 88th percentile.

Ann's written response to the study's questionnaire indicates that she had not been inspired by her high school teachers: "I thought my instructors were boring, therefore, I did not try." Nonetheless, she had placed in the 63rd percentile of her high school graduating class. Although she had scored 10 on the ACT, she wrote a fairly strong placement essay. A major reason for Ann's placement in developmental English was her nonstandard use of verb forms that created sentences including this one taken from a paper that she submitted in her first developmental English class at OSU: "By listening to her tips, I had achievement in my talents."

Summary of Activities at the Word Processor
The chief activity of Fred, Karnak, and Ann at the computer was transcribing previous drafts and notes. All subjects also did some editing at the keyboard, although Karnak and Ann preferred to edit hard copy. Furthermore, Ann did some on-screen global revisions. All subjects also revised and edited extensively on printouts, then used the computer for transcribing. In only two of the sessions observed did the subjects print out a draft, make immediate changes, and then print out another draft. Yet research suggests that a use of hardcopy may be fruitful with the word processor (Haas and Hayes, 1986). In the only instance in which this mode was employed, the changes were superficial--Fred changed from double- to triple-spacing, and Karnak deleted extra spaces between sentences. Clearly, the writers described failed to realize the full capabilities of word processing on computers.

It is perhaps more important, then, to note how the computers were not used. They were used neither for pre-writing nor for writing. Although Fred had no chance to pre-write or write in the lab during his first essay, because he saw the machine as a transcribing tool, it is doubtful that he would have used it for pre-writing or writing had he had the chance. In addition, all subjects generated little on the screen during revision.

Perceived Benefits of Word Processing
All three subjects liked using the word-processing software and believed that it helped them, chiefly in three ways. First, the students could transcribe more easily because they could correct typing mistakes easily. Second, the students believed that they could revise more because they did not have to retype the complete essay. Lastly, the students believed that they could edit more, again because they did not have to retype or use correction fluid or eraser tape. Exhibiting a common attitude of basic writers, Karnak wrote in his journal that his improved capacity to fix surface errors would have a major impact upon his writing: "The word processor unable [sic] me to correct spelling and punctuation mistakes easily without writing the whole paper over again. As a result of the computer, l will have fewer mistakes and my overall writing will greatly improve."

The subjects perceived that the chief function of the word-processing package was to help them avoid retyping rather than to help them generate or combine or reorganize ideas on the screen.

Problems Using the Word Processor
Transferring. Students experienced difficulties because they transferred their writing habits into their computer-assisted composing processes. Recopying was a significant problem for Fred, who inscribed the same draft three different times, the first and second time with pen and then the third time with the word-processing program.

Fred also transferred to his composing habits his over-reliance upon others to perform his writing tasks. During his previous developmental writing class, after he had generated ideas for an assignment, Fred had John, his instructor, identify which ideas to use in his essay: "I had to point them out," John said. Carol said that Fred had to be "pushed" to develop his ideas, that he needed too much encouragement. This over-reliance upon others was frequently evident in Fred's approach to word processing. In one instance after the two training sessions, he had a lab assistant load his disk, retrieve his file, and type in his file name. Fred also asked the assistant how to enter a word into the text. Later during the session, Fred had another writing instructor re-explain how to enter words onto the text and how to scroll. While scrolling, Fred lost his place in the text; consequently, the instructor reexplained cursor movement and then showed Fred how to set the print format and print. When asked, the instructor then gave Fred his fifth explanation of how to erase and his third explanation of how to save the text. Rather than refer to his software instructional handout or to the flowchart of the program drawn on the blackboard, Fred relied upon others to provide this information and often to operate the computer for him.

All writers for the most part used computers as a typewriter to transcribe their drafts. In an interview conducted after he completed his essay, Fred told me that he saw the computer as a typewriter during this period.

During his rewriting, Karnak paid very little attention to his previous draft, and this idiosyncrasy caused him sometimes to revert to previous material, including, in one instance, a thesis that Carol had already edited out of his text. He transferred this habit into his computer-assisted composing, leading Carol to say at the end of the term, "Once he gets something down, he thinks it's carved in stone." Had Karnak written his first draft using the word-processing software then revised that file, his previous draft would have been the focus of his revising efforts.

Ann transferred her habit of reverting to a previous schema to her computer-assisted writing process. Although told by Carol to avoid narration, Ann wrote with pen and paper a narrative description of her participation in a horse show for her first draft and spent the rest of the first week revising her paper in longhand, trying to make the essay more expository. In her first word-processed draft, much of her description is in the present tense; Ann lapsed into the past tense in the middle of her essay. In so doing, she shifted from supporting her informative thesis with a generalizable example to describing a specific incident, seemingly for the purpose of telling a story. In one instance, after introducing her competitors in a typical event, Ann added, "We are all good, but you never know which one of us the judge is going to like the best. My apprehension grew stronger, I felt my [stomach knot grow tighter]." Ann made 15 such shifts into the past tense while using the word-processing program. Thus, Ann continued to make the same kind of errors she made with pen and paper.

Avoiding Writing. Of the four subjects, Fred used the computer to avoid writing by printing out his short text triple-spaced in an attempt to make the draft look longer. In this way, his revision was merely a change in formatting.

Ignoring and Forgetting and Instructions. Because he did not learn computer functions, Fred spent much lab time soliciting others to manipulate the computer for him. In the instance described above, 45 minutes of such activity yielded one word of text. Karnak was more familiar with the word-processing commands, but he still forgot to use the cut-and-paste functions, instead he erased and retyped slowly. Fred's problems may have been caused in part by his using the computer only once every five or six days, but he also ignored the word-processing handout and the hands-on instructional session, entering his first draft instead of listening to Carol.

Typing Interference. Because both Fred and Karnak had to look at the keyboard to type, they had problems generating text on the screen. In fact, Karnak said that his poor typing was the chief reason that he did no pre-writing or writing on the screen. He entered his words by the hunt-and-peck method, typing mostly with his right hand. Fred generated only part of a sentence on the screen.

Writing Environment Problems. Conditions in the students' writing environment were also responsible for their performance with the computers. Ann and Karnak complained that they did not have enough access to the computers and that this lack of access was a major reason why they did no pre-writing or writing with the machine. All three students were carrying heavy course load s for basic writers, and the lab was a significant distance from their dormitories and their classrooms. Therefore, a lack of time to travel to the lab decreased their access, as did the fact that the lab was only open to them for 29 1 /2 hours per week. Karnak and Ann also complained in interviews about noise pollution in the lab--rasping printers and talking--al though they never mentioned this noise during the numerous writing sessions I observed in the lab. Ann also said that glare from the sun made her monitor difficult to read . Although, these lab conditions were contributing factors, they did not cause most of the problems identified in this analysis. Until close to the end of the course, Fred perceived the word-processing program as having the same transcribing function as the typewriter. This misperception, his reliance upon others to manipulate his machine for him and his "hunting and pecking" typing made it unlikely that he would have generated or revised more material on the screen. Ann's chief problem, vacillating between schema, occurred both on paper and on the screen. Finally, Karnak told me in an interview that the chief reason that he did no writing on the screen was the result of his poor typing ability. As he told me and as my observations suggested, the first activity of his writing process was to write a draft of his essay, "go[ing] with the flow" of his thoughts. But because he was not a touch typist, he was prevented from generating material on the screen.

This study indicates that basic writers who were given little guidance in integrating the computer into their composing processes generated little material on the screen and in other ways made far less than full use of the word processor. Clearly, the limitations of the 1982 BANKSTREET WRITER software caused some of the basic writers' problems in generating material on the screen. Because students had to switch modes by calling up a menu and then had to give other commands in order to erase, search and replace, cut and paste, and in other ways revise, the software caused the physical dimension of writing to interrupt rather than complement the cognitive dimension. But, beyond the limitations of the word-processing program, another cause of physical interference was that students were not given intensive and extensive training with the software. Therefore, they did not know how to use some of the functions.

One may argue that software today is sufficiently user-friendly that students need little instruction in or evaluation of their word-processing procedures. But the great demand for instruction for today's programs is evidenced by the numerous word processing software courses offered by colleges, technical schools, and computer stores, all of which might suggest that programs may not be so user-friendly. Linda Fleit (1987), president of EDUTECH International, a computer consulting company for higher education states that:

My idea of a user-friendly system is one that requires a manual of fewer than five pages written in English with a list of at least 25 things to try before you call for service. In fact, user-friendly computing does not exist, not by a longshot. (p.96)

I would agree with Fleit. The study reported here suggests that students might realize more of the potential benefits of the computer if we taught and evaluated word-processing skills as well as writing throughout the term. Piano scales are not sonatas, but practicing them facilitates the performance of sonatas.

By sitting next to Fred, Karnak, and Ann while they composed, I identified crucial word-processing and writing-process problems--that Fred did not know how to enter the revising mode, that Karnak ignored previous drafts when revising, that Ann shifted points of view, and so on. This study suggests that rather than just teaching composing and word-processing skills to the class en masse and examining their written products, we would also benefit from observing writing processes in action in the computer lab. Specifically, the instructor might sit, as I did, next to the student and observe a writing session, helping the writer change rather than transfer writing habits. The instructor could also make suggestions to help the student use the word processor to its greatest potential for that student's composing style, ensuring that such critical functions as cut-and-paste procedures, functions that Fred and Karnak overlooked, were not ignored.

Would students' reactions to our monitoring offset any benefits? As Andrea Herrmann (1985) found, writing becomes a more overtly social act when students compose in a lab. People seem to be drawn to others' screens to see what is "on." Thus, the milieu of the lab is not incompatible with such a pedagogy.

Therefore, we need to continue to consider how much word-processing instruction should be combined with writing instruction. It may be that students using word-processing software in composition classes have produced mixed results not because of too much direction, but rather because of not enough. We need to re-examine our assumptions about the teaching of technology within the context of writing classes.

Geoffrey Cross teaches in the English Department of the University of Louisville.

The following data collection methods were used in the study to produce a triangulated view of the activities observed.

Observation

• Participant observation during entire period of every class meeting (28 classes + final exam).

• Participant observation of eight of the ten writing sessions in which students used computers (recorded students' written words, computer functions used, verbal comments).

• Participant observation of several other writing sessions.

Written Questionnaires

• Open-ended written questionnaire given to students during observation period.

Interviews

• Follow-up interviews conducted in response to written questionnaire answers.

• Frequent discussions with students before and after class and after composing.

• Several conversations and one one-hour interview with Carol, the students' CAC instructor.

• A 45-minute interview with each of the students' previous basic writing instructors.

Document Collection

• High school class ranks, ACT scores, OSU placement essays, papers written for previous basic writing class, developmental English diagnostic and final exams, journals required for the class, all drafts and related materials of the examined composition.

Textual Analysis

• All changes made in all preliminary writings and drafts of each participant.
  

Bullock, R. M. (1985). The lure of the cursor, the fear of the byte: Affective responses to word processors. Computers and Composition, special issue, Selected papers from the Conference on Computers in Writing: New Directions in Teaching and Research (pp. 45-55). University of Minnesota.

Carroll, J. M., & Mack, R. L. (1984). Learning to use a word processor by doing, by thinking, and by knowing. In J. C. Thomas & M. L. Schneider (Eds.), Human factors in computer systems (pp. 13-53). Norwood, N J: Ablex.

Curtis, M. S. (1988). Windows on composing: Teaching revision on word processors. College Composition and Communication, 39, 337-343.

Daiute, C. (1986). Physical and cognitive factors in revising: Insights from studies with computers. Research in the Teaching of English, 20, 142-147.

Doheny-Farina, S., & Odell, L. (1985). Ethnographic research on writing: Assumptions and methodology. In L. Odell & D. Goswami (Eds.), Writing in nonacademic settings (pp. 503-535). New York: Guilford.

Euchner, C. (1983). Schooling in 'basics' must precede computer literacy. In K. Klein & D. Strother (Eds.), Planning for microcomputers in the curriculum (p. 209). Bloomington, IN: Phi Delta Kappa.

Fleit, L. H. (1987). Overselling technology: Suppose you gave a computer revolution and nobody came? Chronicle of Higher Education, 33(32), 96.

Frisch, A. (1987, March). Using language and using computers: Toward a 'useful' view of literacy. Paper presented at the Conference on College Composition and Communication, Atlanta, GA.

Haas, C., & Hayes, J. R. (1986). What did I just say? Reading problems in writing with the machine. Research in the Teaching of English, 20, 22-35.

Hawisher, G. (1988). Research in computers and writing: Findings and implications. Paper presented at the annual meeting of the American Educational Research Association, New Orleans, LA.

Herrmann, A. W. (1985). Writing on the computer: Marginal, selective, and dynamic leaders. Paper presented at the UCLA Conference on Computers and Writing: New Directions in Teaching and Research, Los Angeles, CA. (ERIC Document Service Reproduction No. ED 259 368)

Hollifield, J. H. (1983). In reality, high tech means low skills, poor pay. In K. Klein & D. Strother (Eds.), Planning for microcomputers in the curriculum (pp. 202-204). Bloomington, IN: Phi Delta Kappa.

Mack, R., Lewis, C. & Carroll, J. (1983) Learning to use office systems: Problems and prospects. ACM Transactions on Office Information Systems, 1, 10-30.

Miles, M. B., & Huberman, A. M. (1984). Qualitative data analysis: A sourcebook of new methods. Beverly Hills: Sage.

Sommers, E. A., & Collins, J. L. (1985). Microcomputers and writing. Computers and Composition, 2(4), 27-35.