Effective Strategies for Teaching Writing
WRITING IS A HIGHLY complex process that writers ultimately apply independently (Bereiter & Scardamalia, 1982). Conceivably, writing is one of the most complex human activities (Bereiter, 1980; Hillocks, 1987; Isaacson, 1989; Scardamalia, 1981.)
The inherent complexity of writing suggests that acquiring writing proficiency might prove to be difficult for many students—a speculation borne out both by descriptive research and the experience of many teachers.
Applebee, Langer, & Mullis (1986) reported that students in fourth, eighth, and twelfth grades taking the National Assessment of Educational Progress (NAEP) performed poorly on measures of non-fiction writing: approximately half wrote adequate or better narrative and informative pieces, and only about a third wrote adequate or better persuasive pieces. Eleventh-grade students performed equally as poorly on the 1990 NAEP: they did not write much, and what they did write was of poor quality (National Assessment of Educational Progress, 1990). The NAEP data and other research (Applebee, Langer, Jenkins, Mullis, & Foertsch, 1990; Christenson, Thurlow, Ysseldyke, & McVicar, 1989; Flower & Hayes, 1981) suggest that students in general education experience many writing difficulties.
The students who experience the greatest difficulties with writing are those with learning disabilities and emotional and/or behavioral problems (Englert & Raphael, 1988; Graham, 1982; Graham, Harris, MacArthur & Graham, 1986; Morocco & Neuman, 1986; Montague, Maddux, & Dereshiwsky, 1990; Nodine, Barenbaum, & Newcomer 1985; Thomas, Englert, & Gregg 1987). Given the increasing diversity of children in classrooms (see section on learner characteristics in Chapter 2), there is a need to identify elements of writing instruction that are likely to be most effective at helping teachers improve the writing of the broadest possible range of students. That is, given the practical limitations of the classroom, which characteristics of a single writing curriculum are likely to contribute to improved performance for the majority of students?
In this chapter we describe a few fundamental characteristics of writing instruction that can contribute significantly to a single writing curriculum that is effective with a broad range of students at various performance levels. First, we briefly describe some current issues in writing instruction. Then we turn to the specifics of instructional design.
CURRENT ISSUES IN WRITING INSTRUCTION
Opportunity To Learn
We stress a single effective writing curriculum because frequently little or no real writing instruction takes place in regular classrooms (Applebee et al., 1990; Bridge & Hiebert, 1985; Langer & Applebee, 1986). Therefore, it seems quite impractical to advocate the implementation of two or more writing curricula in diverse classrooms as a means of accommodating the needs of diverse learners.
It goes without saying that the minimal requirement for adequate writing achievement is that effective writing instruction be made available to students at all.
In general, opportunity to learn has long been considered one of the major factors influencing achievement (in addition to pedagogical practice and aptitude; see Carroll, 1963). Students probably will not become better writers if they do not spend a relatively substantial part of most school days engaged in productive writing activities. Graves (1985) states, for example, that students should write for at least 30 minutes a day, at least four days a week, as opposed to a national average of writing one day in eight.
Author versus Secretary, or Author, and Secretary
Not just any time allocated generally to “writing” is likely to result in notable writing improvement. For example, neither “free writing” nor instruction on grammar and writing mechanics have proven, by themselves, to be effective means of improving writing performance (Hi1locks, 1984). The elements of meaningful allocated writing time are the principal subject of this chapter (See Isaacson, in press, for a full discussion of academic learning time and writing instruction.)
Smith (1982, cited in Isaacson, 1991) characterizes writing as a complex undertaking in which the writer works both as author and secretary throughout the process of writing. The writer-as-author is concerned primarily with matters of content, including the origination and organization of ideas, levels of diction, and so on. Simultaneously, the writer-as-secretary is concerned with the mechanics of writing. Sometimes the secretary role is characterized as a concern related almost solely to the revision phase of writing, but for students with learning difficulties, mechanical skills such as handwriting and spelling can present severe obstacles to participation in all authoring processes (Graham, 1990).
The author-as-secretary characterization provides a framework for identifying vastly different orientations toward writing instruction. The first is the skills-dominant approach, in which instruction focuses primarily on the mechanics of writing: secretarial concerns. Based on several descriptive studies, this approach has been the predominant one in American schools for many years (Applebee, 1981; Bridge & Hiebert, 1985; Langer & Applebee, 1986; Leinhardt, Zigmond, & Cooley, 1980). Within this approach, composition activities are minimal, often limited to writing short answers or transcribing.
Even less emphasis seems to placed on composition in skill-dominant approaches used with lower-performing students. Such students receive a great deal of skills instruction Englert et al., 1988, Graham et al., 1991; Isaacson, 1989; Roit & McKenzie, 1985), and even that instruction is poor, in that it occurs in isolation and is unconnected with its presumed eventual use (Graves, 1985).
Interestingly, skills-dominant approaches are polemical ghosts, in that they have little if anything to recommend them. To our knowledge, no one advocates skills-dominant approaches in the literature. If there is a rationale for such approaches at all, we can only speculate as to what it might be.
Perhaps someone believes that attention to mechanical skills will somehow result in improved composition. Perhaps someone believes that lower-performing youngsters are incapable of creating coherent text without first acquiring a full complement of mechanical skills. Or perhaps no rationale for skil1s- dominant approaches exists at all: for example, a teacher who is not comfortable with his or her own composition ability might inadvertently slight composition instruction in the classroom. We know for certain only that skills-dominated approaches are in widespread use, without the benefit of empirical or theoretical support.
A second, nearly opposite approach to writing may have evolved in reaction to skills-based approaches: the composition-dominant approach, concerned primarily with authoring aspects of writing. Advocates of this approach generally argue that instruction mechanics should be restricted to those concerns students raise themselves in connection with the polishing stage of composition (DuCharme, Earl, & Poplin, 1989; Graves, 1983).
We have some general concerns with composition-dominant approaches. First, there is little research to support the hypothesis that the mechanics of writing will take care of themselves in the context of authentic writing experiences. Many of the gains reported anecdotally for students in composition-dominated programs could possibly be the result of maturation. In addition, some measures of collaborative efforts may mask individual achievement, or lack of it.
Second, there is strong evidence that mechanical difficulties can effectively preempt many students from meaningful participation in far more rewarding authoring roles (Graham, 1990; Morocco, Dalton, & Tivnan, 1990). The kinds of general difficulties experienced by many students with learning problems strongly suggest such students are not likely to acquire knowledge of any sort casually (Isaacson, 1991). In our well-justified haste to distance ourselves from skill based approaches to writing, we should be cautious and thoughtful mechanics are an integral part of writing.
If we envision writing as an interweaving of complexities involving both author and secretary roles, then perhaps parallel instruction is one means for resolving
dominance conflicts. Within a parallel framework, instruction includes all aspects of writing from beginning to end - from conceptualization to “publication” - with a concerted focus on the integration of writing knowledge. (See Isaacson, 1989, for an in-depth discussion of this issue.)
PRINCIPLES FOR IMPROVING
WRITING INSTRUCTION STRATEGIES
The implication of time allocated to writing instruction (or the lack of time allocated to writing instruction) is clear and, it seems, unanimously advocated; more time needs to be allocated. Any controversy that exists relates to different approaches to such allocation of time.
In this chapter we apply the six design principals highlighted in Chapter 1 to both the author role and the secretary role of writers. Although we separate the roles for the sake of instruction, we wish to re-emphasize that the roles co-exist and intertwine in
authentic writing. Although the principles and applications we describe are research-based, we caution our readers that much of the substantial research conducted on writing in recent years is descriptive, anecdotal, quasi-experimental, or otherwise questionable as the basis for making broad generalizations about effective writing instruction (Graham et al., 1991). Still, data from a few very good studies, coupled with knowledge of diverse learners and instructional design research, provide the basis for cautiously identifying some important aspects of effective writing instruction for students at diverse levels of writing proficiency.
Designing instruction Around Big ideas
Big ideas and the Author Role in Writing In general, big ideas for writing instruction are those that seem to recur across successful writing programs. However, the notion of big ideas in general is not based as much on empirical evidence as on our intuitive analysis of the alternative: teaching small, inconsequential, or marginal aspects of writing.
Writing Process One well-known big idea in writing is usually referred to as the writing process. The idea that writing instruction should center on the stages through which writers most frequently work goes back at least 25 years, when Herum and Cummings (1970) wrote on the writing process for college students. Those educators may have been ahead of their time, since the widespread acceptance of their approach in public schools is usually credited to Graves (1983).
Presumably, professional writers and those for whom writing is a major part of their profession have always reiteratively planned, drafted, and revised their work, dating back to the classical Greek rhetoricians. Surely it is past time for school children learning to write to be let in on this fairly public “secret” of good writers.
Text Structures An awareness of the writing process by itself, however necessary to writing instruction, appears to be insufficient for consistent results, particularly for students with learning disabilities and with other difficulties (Englert et al., 1991). Text structures is another big, authoring idea which has resulted in impressive achievement gains when combined with process writing. Each writing genre can be identified by its own set of structural characteristics. Stories, for example, always have a protagonist, a crisis, developing incidents, and a resolution. Students who are unaware of such common recurring elements might write “stories” that are more like rambling narratives than true stories.
Several studies have shown solid promise for teaching text structures and process writing in conjunction with one another (Graham & Harris, 1989; Hillocks, 1986; Meyer & Freedle, 1984). The work of Englert et al. (1991) is especially promising in terms of effective writing instruction for diverse learners in that it demonstrates how writing can be effectively taught simultaneously to mainstreamed learners with disabilities and their average-achieving peers.
Englert et al. (1991) also have shown a distinct advantage of focusing on big
ideas: their instructional program taught less, but students learned more. That is, the program they developed taught only two text structures within a school year, but those structures (explanations and compare/contrast) were important to future schooling success, students learned them well, and the results on measures of transference were good. In contrast, our informal analysis of language arts texts reveals that between a dozen and two dozen text structures are typically “taught for exposure” within a single school year. When too much material is “taught for exposure” or merely “covered,” many students appear to learn and retain little. The study by Englert et al. suggests that “less is more” when the content chosen is truly important.
Peer Interaction Finally, peer interaction appears to be important for improved composition performance. Collaborative work has proven to be an effective instructional tool in many subject matter domains, but it has a particular benefit to writing instruction: when working in cooperative groups, each student has the opportunity to participate in authoring, editing, and reading. Although the act of writing is often a covert and solitary endeavor for mature and able writers, those learning to write benefit from many opportunities to talk about writing with peers.
Big Ideas and the Secretary Role in Writing
We can conceive of several potential big ideas related to writing mechanics, ideas that promote understanding and reduce the learning burden for students. Morphology may be a big idea for spelling instruction (Dixon, 1991; Henry, 1988). The idea of combining manuscript and cursive writing into a single system, as in the D’Nealian writing program (Brown, 1984), promises substantial efficiency for teaching handwriting. Effective keyboarding instruction also might help to reduce the burden of simply setting print to page (Brown, 1984) and promises substantial efficiency in teaching handwriting.
Hillocks’ (1984) widely known research review of effective writing practices suggested that although sentence combining alone is not the most effective way to improve writing, it was more effective than other approaches examined (teaching grammar, free writing, using good models of writing). Sentence combining and manipulation, then, might be considered a significant but non-dominant big idea for teaching writing mechanics. (We illustrate this possibility more fully in later sections of this chapter.)
The notion of big ideas is less an “instructional design characteristic” than a foundation on which to build successful instruction. We do not design big ideas, we uncover them through a careful and complete analysis of content area literature. Big ideas do not guide us on how to teach; they are a major factor in determining what to focus on as we design instruction.
Designing Conspicuous Strategies
Conspicuous Strategies and the Author Role in Writing Most students with learning difficulties, and many average-achieving students, do not automatically benefit from simply being exposed to big ideas, such as the steps in the writing process or text structures. A substantial body of research is accumulating that supports the teaching of conspicuous strategies for using those ideas (See Deshler & Schumaker, 1986; Pressley, Symons, Snyder & Cariglia-Buli, 1989).
A teacher once described to the first author of this chapter the difference between the “old” and “new” ways of teaching writing for one of her students: “He used to sit, unable to get started, when trying to write about his summer vacation. Now he sits, unable to get started, when trying to plan what he is going to write.”
A conspicuous “planning strategy” could clarify for students some specific steps for starting and successfully completing their planning. The steps in such strategies derive from the best efforts of subject matter specialists to uncover or emulate cognitive processes that are normally employed covertly by experts. However, teaching just any set of steps to follow does not necessarily constitute a good strategy.
The best strategies appear to be those that are “intermediate in generality” (Prawat, 1989). If a strategy is too general, it is not likely to lead to reliable results. For example, “think before you write” is a general strategy and a good idea, but is too general to be of much practical value for many learners. On the other hand, a strategy that is too narrow is likely to result in the rote acquisition of some bit of knowledge with little potential for transference.
Conspicuous strategy instruction has been used with promising results to teach all phases of the writing process: planning (Harris & Graham, 1985), text structure (Englert. et al., 1991; Graham & Harris, 1989), and revising (MacArthur, Graham, & Schwartz 1991). Such strategies have appeared to meet the “intermediate in generality” criteria.
For example, Graham and Harris (1989) taught students to generate and organize story ideas by asking themselves questions related to the parts of the story: “What does the main character want to do?” (p. 98). On the one hand, the strategy was not too broad. The questions taught in the study all involved parts of stories. On the other hand, the strategy was broad enough: it directed students’ attention to the elements common to all stories.
Conspicuous Strategies and the Secretary Role in Writing Assume that a student is puzzling over the following sentence while attempting to edit and revise a draft:
All of we young people seem to like ice cream.
Is it, the student wonders, we or us young people? In terms of grammar, the answer can involve a complex array of spiraling knowledge: nominative case, objective case, objects of verbs, objects of prepositions, predicate nominatives, appositives. It is little wonder that many teachers would choose to forgo a grammatical approach in favor of nearly almost any other option, such as telling the student the answer or suggesting that the student rewrite the sentence to make the problem disappear.
Yet the problem can be attacked via conspicuous strategy instruction, with relatively little effort and complexity, and with relatively high potential for transference. The strategy is to decompose or simplify the sentence in question, then examine the results:
All young people seem to like ice cream.
All of we/us seem to like ice cream.
A native speaker of English who does not have a severe language disorder will instantly recognize us as the correct choice in the simpler sentence and realize that it is therefore the correct choice in the original sentence.
The same general strategy can be applied to far different instances of pronoun case, and to difficulties not involving pronoun case at all.
pronoun case: compounds
John gave Mary and I/me a new book.
John gave Mary a new book.
John gave me a new book.
Original sentence: None of the boys was/were on time.
First simplification: Not one of the boys was/were on time.
Second simplification: Not one was on time.
Designing conspicuous strategies is challenging. We tend to readily recognize good strategies that others have designed, but most instructional designers agree that designing a good strategy from scratch is no simple matter. The best we can do is suggest that anyone attempting to design conspicuous strategies begin with something, then evaluate the early attempts critically, using “intermediate generality” as the principal criterion for analysis. Whenever possible, promising strategies should be field-tested with students.