Is Writing Across the Curriculum Effective?

Though there appears to be no shortage of anecdotal evidence of program successes—from improving writing scores to helping students make richer connections across subject areas—few high-quality quantitative studies of writing across the curriculum currently exist. The enormous variety among programs, not to mention the number of variables likely to affect student writing and learning, present serious obstacles to designing valid, large-scale studies of WAC's effects. As Harris and Schaible (1997) note in their review of the research on college WAC programs, "Investigators... are often unable to randomly assign students to control and study groups. They are not able to completely eliminate nonwriting variability in instruction. Nor can they assure that all instructors are equally skilled at the evaluation of writing assignments" (p. 31). Changes in school policy, funding, and staff also make long-term studies of program effectiveness difficult to carry out.

Lilyanne van Allen's (1992) study of 10 Texas middle schools is one of the few studies that attempt to sort through these variables. Comparing writing score data collected from five middle schools with WAC programs to data from five comparable schools without, van Allen concluded that schoolwide writing across the curriculum efforts do appear to produce significant improvements in student writing ability. Over a five-year period, "the five WAC schools increased the percentage of their passing composition scores by thirteen percent while the five non-WAC campuses increased their percentage only eight percent. Teachers in all five [WAC] schools said that students improved in fluency and organization and wrote for more purposes and audiences" (van Allen, 1992, p. vii).

Based on her visits to schools and interviews with teachers, van Allen also concluded that "a first and major effect of a WAC program is a change in attitudes toward writing" among teachers and administrators as well as students (p. 116): "Through writing, every student does have an opportunity to respond, to participate, to learn actively rather than passively, and to think independently…. A WAC program, founded on a process approach to writing instruction, also provides every teacher an opportunity to take a fresh look at writing" as more than just a product to be graded for punctuation and other surface-level errors (p. 116).

Given the difficulty of accurately assessing relationships between student learning and schoolwide programs, many researchers have focused instead on the effectiveness of particular writing activities employed by individual teachers. Bangert-Drowns, Hurley, and Wilkinson (2004) conducted a meta-analysis of 48 such studies, elementary through college, that examined relationships between classroom writing-to-learn activities and student achievement. Only studies with quasi-experimental, control group designs were included. Based on their analysis, the researchers concluded that writing-to-learn activities "can have a small, positive impact on conventional measures of academic achievement" (p. 29). The use of metacognitive prompts in which students could "reflect on their current knowledge, confusions, and learning processes proved particularly effective" (p. 50). It also appeared that the longer the period of time over which writing activities were spread and repeated, the greater their effect on student learning. The use of writing-to-learn strategies in grades 6–8 and lengthy writing assignments, however, appeared to predict reduced effects. The researchers speculated that this negative effect of writing-to-learn in the middle grades may be related to developmental issues, or to the changing demands placed on middle school students new to discipline-specific classes and genres of writing. It may also simply have been an anomaly, as other studies of middle school students do link particular writing-to-learn practices to student achievement (see Cantrell, Fusaro, & Dougherty, 2000, for example).

Coming at the issue from another direction, the National Writing Project (NWP), which strongly endorses a process approach to writing across the curriculum, has commissioned a number of studies that link their professional development model to improvements in student writing scores. One study, conducted by the Academy for Educational Development (AED), assessed the writing achievement of third- and fourth-grade students whose teachers had participated in NWP summer institutes. Results indicated that students in these classrooms "showed significant improvement in writing achievement…. In response to timed writing assignments, 89 percent of third-graders and 81 percent of fourth-graders reached adequate or strong achievement for effectiveness in persuasive writing on their follow up assessment in spring 2001" (NWP, 2002).

To be certain, readers searching for "gold standard," experimentally designed studies of writing across the curriculum will find many limitations to the existing body of research, as well as a number of important questions that remain to be addressed. As Ackerman (1993) notes, for example, connections between students' home literacy practices and the effectiveness of activities often associated with writing across the curriculum demand greater attention: "all writing practices carry cultural values, and instruction consisting of informal, expressive, and exploratory writing practices may be an affront to some writers as much as an invitation to personal discovery learning…. Writing to learn is a literate practice that assumes cultural norms" (pp. 350–351). How students from different ethnic, linguistic, educational, and socioeconomic backgrounds benefit from particular writing strategies has not been studied in nearly enough depth.

It also remains unclear, as Bangert-Drowns, Hurley, and Wilkinson (2004) note, "how student ability is related to the impact of writing to learn" (p. 53). While some writing activities may indeed be a good match for more print-oriented students who already have strong reading and writing skills for their grade level, it's possible that these same activities could limit subject-area learning for students who perform better in other ways. Writing activities that are poorly designed or that ask students to do more than they are developmentally ready for may also have a negative impact on learning.

In the following sections, we draw largely from practitioner literature to provide examples of current practices and programs. Teaching strategies supported by a wide body of evidence are noted whenever possible. In many cases, however, too little experimental research has been done to offer clear evidence that a particular approach will lead to a particular outcome for students or schools.