Robert Hymer

Conventional approaches to writing instruction are characterized both by a reliance on paper and pencil and by what might be called "workbook" methodology: students are presented with formal principles concerning language use and then given assignments whose chief function is to get them to employ the principles in their writing. In the past several years there have been two major developments in the teaching of writing: word-processing programs on microcomputers (Dudley-Marling, 1985; Schwartz, 1982) and the "process" or "workshop" model of writing instruction (Calkins, 1986; Graves, 1982) which immerses students in the creation of texts so they can "learn how to write by writing." Teachers in Indian schools should be particularly interested in the possibility of combining these developments into an alternative approach to improving writing abilities. The need for such improvement is obviously acute for many Indian students, especially for those who go to college; and a reliance on conventional methodologies has not, over the past hundred or so years, produced many results which would justify its continuation if better alternatives were available.

For all of the attention which computers and the process model have received separately, however, little has been written about attempts to combine them into a unified approach to writing instruction. Also, although microcomputers are coming into increased use in Indian schools, studies of their potential for improving Indian students’ writing abilities are still rare (Diessner et al. 1985). This article is based on a study conducted at an Indian school which has instituted process-based writing instruction at all levels K-12 and at the same time provided all students from 4th grade on up with frequent and regular access to a computer writing lab.

Teachers at Pine Hill School, on the Ramah Navajo Reservation in west-central New Mexico, became interested in the process model and the use of computers for writing instruction and wrote a successful Title IV-B grant proposal to the Indian Education Programs Office of the U. S. Department of Education. This grant was to be used for improving writing abilities of all students in grades 4-12. As a result of the grant, Pine Hill had a small computer lab dedicated to word-processing by the middle of the 1984-85 school year, and all teachers involved in language arts were teaching writing through daily "writing workshops" based on the process model.

The lab consisted of eight Apple IIe computers, each with its own printer, as well as supplies for a small book-making center. Students in the 4th and 5th grades used the Talking Screen Textwriter program, which is capable of reading back what the computer monitor displays through earphones and otherwise operates like a normal word-processing program. From the 6th grade up, students used Bank Street Writer. Each student was supposed to have at least one hour per week on the computer in addition to regular classroom writing hours. Students received minimal technical training before using the computers; the important thing, it was felt, was to get them writing.

The stages through which a text passes in the process-model consist of (1) prewriting activities, (2) writing a first draft, (3) revising this draft after a conference with a reader, (4) editing and proofreading the revision, and then (5) making a final version which is published in some form or other. Although students spent more time writing in their classrooms than they did in the lab, they were encouraged to use computers at all stages and not just to see them as a means of getting a presentable copy from a handwritten text. Since the members of a class might be at any of the various stages when they were in the lab, this meant that at any given time some students might be engaged in a prewriting activity, others might be writing a first draft, others might be revising or editing, and still others might be turning a final version of their writing into an illustrated booklet or a newspaper article. The teacher, meanwhile, might be assisting individual students at any of these tasks, while at the same time students would be assisting each other in various parts of the room. The apparently common fear that introducing computers means transforming schools into places where students are isolated in front of individual monitors, and teachers feel guilty because they no longer have much interaction with students, could not have been more strongly disconfirmed.

Of the process stages, revising and editing were probably the most difficult to teach. It was generally easy, and productive, for teachers and students to engage in prewriting activities like brainstorming, list-making, and answering who-, what-, where-, why-, when-questions; and, nearly all students appeared to enjoy the process of getting their words onto the screen and from there into print. Although it is true that word-processing makes it easy for writers to alter what they have written, however, most students were used to not changing their writing once they had written it and did not spontaneously take advantage of this benefit. The process model relies on conferencing as a method for helping writers see how to make improvements in their first drafts; but in practice, it is difficult for a teacher to hold many writing conferences in a single class period. The alternative of turning every student into a helpful reader of other students’ texts sometimes worked, but teachers quickly became aware that there is nothing easier about turning students into such readers than there is about turning them into writers.

It is fortunate, therefore, that the computer-and-printer combination so nicely supports the final stage of the process model, which is publication. Having a goal like the production of a book or other document provided a motive for revising and editing that would otherwise have been absent; and the ease of printing and reprinting texts made such publication seem an achievable goal. Efforts to encourage students, teachers, and community members to read and appreciate the writings which students produced were an integral part of the project.

The beehive level of writing activity which the project fostered was encouraging, but there was still the question of whether students’ writing abilities were genuinely improving as a result of it. Writing assessment checklists which teachers filled out for individual students fairly uniformly indicated that students had improved since the time the project began; but of course they could be faulted as reflecting subjective judgments. When the opportunity came to request another grant for 1985-86—which also allowed the writing lab to acquire additional computers and printers—provision was made for evaluating students’ writing improvement in a more objective manner.

The method chosen was a pre- and posttest comparison of Pine Hill students and students from the same grade levels at Zuni, a nearby Indian school district which relied for the most part on conventional approaches to writing instruction. The "tests" were writing samples collected in the fall and spring of the school year, with approximately eight months of intervening instruction. To produce these writing samples, students described or wrote stories suggested by a photograph. The photographs were in pairs: students who wrote on photograph A in the fall of 1985 would write on photograph B in the spring of 1986, and vice-versa. All students wrote their samples by hand. The writings were then copied—with nothing corrected—on a computer and printed in such a way that readers would have no way of identifying when, where, or by whom any given sample was produced. Readers were trained in the method of holistic scoring and instructed to consider four different qualities—fluency, clarity, correctness, and maturity—in their assessments. Then they assigned an overall score of 1 to 6 to each sample they read. Two readers scored each sample. A third reader scored samples on which they had not agreed, and these samples received the average of all three scores.

There were 81 students in the experimental (Pine Hill) group and 90 in the control (Zuni) group, making a total of 171 students and 342 samples. The sampling included grades 4-8. Samples were gathered from high school students as well, but too fewer of the 9-12 students from Zuni who had produced writings in the fall were available to write again in the spring. It was not possible, therefore, to make experimental vs. control group comparisons beyond the eighth grade.

The three tables below summarize the scores for the sampling as a whole, for the two groups, and for the separate grades within each group.

TABLE 1
Score Summary for Entire Group

TABLE 2
Score Summary for Experimental and Control Groups

TABLE 3
Scores by Grade Level

The patterns of improvement were very different for the two groups. As Table 3 indicates, the general tendency of scores in the control group was relatively changeless between the fall and the spring. Scores of students in the experimental group, on the other hand, tended to increase by substantial margins. The two grades that performed most similarly between the two groups were the control group 6th grade and the experimental group 7th grade. That is, the control group grade which showed the largest gain closely resembled the experimental group which showed the smallest gain; it might be relevant that this class of 7th graders was also the one which spent the least amount of time during the year in the writing lab.

The pretest scores, however, also suggest the possibility that the writing skills of the experimental group were as much as a year behind those of the control group to begin with. In the fall, the experimental group 5th, 6th, 7th and 8th graders resembled the control group students one grade behind them more than they did the control group students at their own levels. The progress they made during the year, therefore, would reflect the catching up they had to do as well as their forging ahead. In order to establish that they performed differently from the control group in a significant way, it was necessary to equalize the two groups by subtracting the effects of how well they were writing before instruction from how well they were writing after it. The technique used was analysis of covariance (ANCOVA). In order to establish its applicability, a test for homogeneity of regression coefficients was run. This was found to be nonsignificant, showing that ANCOVA was an appropriate analysis technique.

The pretest (covariate) was significantly related to the posttest, F (1, 160) = 12.83, p <.001, accounting for 5.07% of the variance in the posttest. The disordinal interaction was significant, F (4, 160) = 7.89, p <.001, accounting for 12.48% of the variance in the posttest scores. The adjusted posttest means are shown in Table 4 and the graph (Figure 1).

TABLE 4
Adjusted Posttest Means

As Figure I shows, the experimental group did not always finish ahead of the control group (control 6th grade higher than experimental 6th grade) and also did not always improve with grade level (experimental 4th grade higher than experimental 5th grade). These anomalies make interpretation of the adjusted posttest means problematical. An additional statistical test, simple main effects, was performed to further explain the significant interaction; the probability level was adjusted to maintain an overall alpha-rate of .05. This test showed that only two results were statistically significant: the experimental group 8th grade (M = 3.61) performed significantly better than the control group 8th grade (M = 1.70), F (1, 160) = 73. 11, p <.0l; and the experimental group’s scores increased significantly with grade level while those of the control group did not, F (4, 160) = 7.31, p <.05.

The conclusion that the performance of the Pine Hill Students improved as they advanced whereas that of the Zuni students did not reinforces the initial impression made by fall and spring scores: not much changed in eight months for one group of students; a great deal changed for the other group. It is hard to know what to make of the striking disparity in scores between the two 8th grade groups. The fact that one group’s scores climbed at the same time the other group’s fell is surely part of an overall picture of improvement vs. levelling off, but in order to be sure of its importance, we would have to be able to compare scores from the two high school groups as well.

The question of which aspects of writing improved the most remains. Pine Hill samples from the spring tended to be longer than Pine Hill samples from the fall, and longer than Zuni samples from both spring and fall. Fluency, interpreted simply as length of text, may therefore have been of major importance. Readers were trained to give other features—clarity, correctness, and maturity—just as much weight as overall length, however, so as to avoid identification of the complex question of quality with the simple one of quantity; so presumably the writers who received higher scores for their spring than for their fall writings changed in some of these respects as well.

The general impression produced by this study of writing instruction is not so much that both approaches were doing the same things and that the new approach did them better, but that the new approach created a different pattern of growth in writing ability. Implicit in traditional forms of writing instruction is the assumption that writing improvement occurs in a linear fashion, from simple to complex, from less correct to more correct. The assumption implicit in the process model, however, is not at all the same: improvement is more like growth outward from a center, with different areas of writing ability developing at different rates. The success of such an approach is essentially dependent on the writer’s taking responsibility for the quality of the text he or she produces. An assessment of the writing abilities of students trained by this method, therefore, should reflect not only how well they have learned one or another skill, but also the extent to which they have taken this kind of responsibility—the extent, in other words, to which they have started behaving like writers.

If the approach used at Pine Hill works, it does so by creating enthusiasm about writing in students and teachers alike, and, in the process, by altering the setting in which writing takes place. The language skills which American students are expected to acquire from their schooling remain pretty much the same whatever the characteristics of the communities in which their schools are found; but a community which depends on and encourages the use of English and print-media probably reinforces school-taught language skills very differently from a community which is attuned to its native language over English, and to speaking and hearing its messages over writing and reading them. When a school is, like Pine Hill, located in a community of the second kind, its efforts at improving literacy skills may meet with little success until it finds a way of incorporating into itself some of the reinforcement functions that in other communities can be taken for granted. It may have to subsidize, in other words, a "literacy industry" within its own walls.

The constant activities which characterize a process-oriented school with a computer writing lab—students coming and going with their texts between computers, printers, teachers and their fellow students, and texts being printed, changed, printed again and made into books or posters—are outward signs of such an industry. Much work, particularly in the areas of curriculum development, teacher training, and evaluation studies like the present one, remains to be done before schools can feel confident they have established their long-term ability to direct such activities so as to achieve chosen educational objectives. The point to focus on, though, is that this instructional model clearly seems capable of energizing schools in such a way as to make genuine improvement in the writing abilities of Indian students an attainable goal.

Calkins, Lucy M. (1986). The art of teaching writing. Exter, NH: Heinemann Books.

Diessner, Rhet, Rousculp, E.L., & Walker, J.L. (1985). English fluency via computers at Yakima Tribal School. Journal of American Indian Education, 25, 17-24.

Dudley-Marling, Curtis C. (1985). Microcomputers, reading and writing: Alternatives to drill and practice. The Reading Teacher, 38, 388-391.

Graves, Donald (1982). Writing: Teachers and children at work. Exeter, NH: Heinemann Books.

Schwartz, Mimi (1982). Computers and the teaching of writing. Educational Technology, 22,27-39.