Researchers Cast Skeptical Eye on Efficacy of 'Writing to Read'
While governors in three states have moved recently to mandate or encourage the use of the computer-based "Writing to Read" literacy program in schools, some education researchers caution that the growing popularity of the software may be due less to its educational effectiveness than to creative marketing by the International Business Machines Corporation.
Gov. Gaston Caperton of West Virginia this summer announced a statewide effort that would lead to increased use of the program in his state's elementary schools. (See related story, next page.)
In April, Gov. Buddy Roemer of Louisiana made a controversial proposal to mandate the placement of a Writing to Read laboratory in each elementary school in the state.
And last year, Gov. Ray Mabus launched a $13-million project to encourage the use of personal computers and Writing to Read in every elementary school in Mississippi.
But despite policymakers' and educators' growing reliance on the ibm package to boost students' reading and writing skills, some critics assert that the program has never been proved more effective than traditional, less costly teaching approaches.
Mr. Roemer's proposal, which would have required every Louisiana elementary school to use the program, immediately roused the ire of local educators and lawmakers and that of Apple Computer Inc., ibm's major competitor in the K-12 education market. (See Education Week, April 25, 1990.)
The uproar stalled the proposal in a legislative committee and led the Governor to rethink his plan. But almost unnoticed in the political fray was a state official's pointed criticism of the effectiveness of the popular and highly promoted program.
"There are no valid studies to indicate that this program is effective," wrote John R. Rombach, the legislature's fiscal officer, in a memorandum to the chairman of the House appropriations committee.
In his letter, Mr. Rombach cited an article in the March issue of Educational Leadership, the journal of the Association for Supervision and Curriculum Development, critical of Writing to Read.
James F. Lytle, executive director for research and evaluation for the Philadelphia public schools, and Pamela Freyd, one of the district's computer teachers, argued in the piece that their review of 17 studies of Writing to Read revealed that the program "has yet to be subjected to rigorous critique" and that "no study has found long-term benefits to participating students."
Demand for the program, they suggested, could be attributed in part to an aggressive ibm marketing program featuring "good packaging and high-pressure sales."
The Lytle and Freyd article is one of several recent papers that question whether the long-term effectiveness of Writing to Read has ever been objectively evaluated.
Critics argue that other methods of instilling literacy in the early grades have largely been ignored by districts eager to appear sophisticated by using a technologically oriented approach.
"Ibm's measure of success is that the market has developed and they've been able to sell the thing," Mr. Lytle said in an interview. "But they have not done long-term evaluations of the product."
'Pays for Itself'
Since 1984, when ibm began a national demonstration project of the program developed by the education researcher John Henry Martin, more than 7,000 schools have purchased Writing to Read laboratories, according to Virginia Nelms, who oversees the program for the company.
The program, designed for kindergartners and 1st graders, uses a "multi-sensory" approach. Children move among five stations, learning to write with their own "invented" spelling, listening to stories on tape, filling in the blanks in workbooks, and typing out short stories, usually on ibm personal computers. The computers also drill students in a phonetic alphabet devised by Mr. Martin.
Pupils work in pairs and may spend as much as an hour a day in the Writing to Read classroom throughout the school year.
The labs are generally supervised by a teacher and at least one teacher's aide, who help students move through 10 lesson modules in the phonetic-spelling segment of the program.
A typical Writing to Read laboratory, designed to serve approximately 24 students at a time, costs $15,000 to establish, according to ibm That estimate includes all of the necessary hardware, but does not cover salaries and other overhead costs.
Ms. Nelms says that the program produces results that are popular with teachers, students, and parents and help improve students' self-esteem.
"We have looked long and hard at costs and when you consider some ... aspects such as students who have a very positive self-image and parents who feel very positive about the program, we feel it's a very minor expense," she argues.
"A lot of school systems," she adds, "feel that Writing to Read pays for itself in terms of getting students off to a head start" on their educations.
The program's effectiveness, she asserts, has been demonstrated by a two-year study conducted in 1983 and 1984 by the Educational Testing Service of Princeton, N.J., before the program was introduced to schools.
In that study, the researchers Richard T. Murphy and Lola Rhea Appel concluded that "'Writing to Read' is an effective educational program."
According to the ets study, kindergartners and 1st graders in Writing to Read labs wrote better than those in comparison groups. In addition, kindergarten students read much better than, and 1st graders read at least as well as, their peers.
Ibm still uses the findings of the ets study to support its claims of Writing to Read's educational effectiveness. According to Ms. Nelms, the company has no plans to commission an update of the six-year-old study.
But the ets study almost immediately drew criticism from other experts, who maintained that its conclusions were not supported by hard data.
In 1985, Walter E. Hathaway, writing as the director of the research and evaluation department for the Portland, Ore., school district, argued in a lengthy, unpublished critique to ets that the data actually culled by the testing service's researchers on improvement in reading and writing skills indicated only a "small, but positive effect" of the program on students' writing and "an inconsistent, positive effect" on their reading.
Mr. Hathaway asserted that the authors of the e.t.s. study "seem to have drawn conclusions that are more positive than are supportable by the data."
Mr. Lytle of the Philadelphia schools, who is also a visiting professor at the University of Pennsylvania graduate school of education, says he decided to take a critical look at Writing to Read when he discovered at numerous national forums that many of his colleagues seemed to question the program's effectiveness.
"Repeatedly, people would start talking about Writing to Read and about [unfavorable] evaluations of the program that had been suppressed [by district officials] for one reason or another," he recalls.
Many of his fellow research directors, he says, expressed "doubts about the efficacy of the program."
Unfavorable evaluations of the program by local educators frequently are downplayed, some critics maintain, because superintendents and other policymakers are awed by ibm's corporate stature.
But spokesmen for ibm's Atlanta-based Educational Systems division are quick to cite the ets study and to refer to the enthusiastic endorsements of school districts stretching from Volusia County, Fla., to Imperial Beach, Calif., that are included in press releases about the software program.
"We have 7,000 schools and over 2 million students using Writing to Read," Ms. Nelms says, "and we have many reports from those people ... that they feel very positive about the program."
Ms. Nelms, whose response to the Freyd and Lytle article was published in the same issue of Educational Leadership, maintains that the two researchers used a haphazard approach to criticizing previous studies that overlooked the educational benefits of Writing to Read.
"There were many more positives from those studies than there were negatives," she says. "There were more significant outcomes than there were nonsignificant outcomes in the studies."
But, she concedes, ibm does not have the sort of longitudinal data describing long-term effectiveness of Writing to Read that many critics argue would bolster the program's educational credibility.
Lack of 'Rigor' Seen
Doubts about Writing to Read have led several other researchers to take a critical look recently at the ets study and several studies conducted by school districts using the program.
In an article to be published in the Summer 1990 issue of the Journal of Computer-Based Instruction, Kathy A. Krendl and Russell B. Williams, respectively an associate professor and a graduate student in the department of telecommunications at Indiana University, conclude that most studies of Writing to Read are "characterized by a general lack of scientific rigor."
They argue that many of the reported benefits of the program can be attributed to the additional time that students in the program spend on instruction; to the students' enjoyment of a novel classroom situation; and to the increased attention children receive from teachers, principals, and other adults when they enroll in the program.
"Though there is some evidence of direct benefits from the program,'' they conclude, "there are also clear indications that these benefits may well be less a function of the ... program with its technological orientation and its multi-sensory experiences than a function of devoting more instructional time to, and putting greater emphasis on, structured writing experiences at these early grades."
And although spokesmen for ibm say they have not seen some of the unpublished studies that are critical of Writing to Read, they argue that the product's popularity4and its supporters in the education community speak for the program's effectiveness.
But they also note that the company has made no claims that Writing to Read is the only effective way to teach young children to read.
"I don't know that ibm is saying that [effective language instruction] can't be done without Writing to Read," says Tom Wall, a company spokesman.
Less Expensive Options?
In fact, contends Robert E. Slavin, a researcher at the Center for Effective Schooling for Disadvantaged Students at Johns Hopkins University, the aims of Writing to Read can be achieved more economically, and at least as effectively, in other ways.
Mr. Slavin reanalyzed 29 separate studies in 22 different school districts that compared Writing to Read with traditional teaching methods, including the ets studies.
In an unpublished paper completed this summer under a grant from the U.S. Education Department's office of educational research and improvement, he concludes that "there is no evidence to suggest that Writing to Read has a positive effect on the reading achievement of 1st graders, even if they have also been in the program in kindergarten."
In addition, he maintains, "there is no evidence to suggest that any effects of Writing to Read on reading achievement can be detected after the year in which students were in the program."
Mr. Slavin argues that many of the studies that purport to show the program's dramatic effects on reading compare kindergartners in academic programs with those in non-academic ones.
Moreover, he notes, five kindergarten reading programs that have been shown to be at least as effective as Writing to Read are already a part of the Education Department's National Diffusion Network. Those programs, he notes, "involve costs on the order of $100 to $150 per class plus training."
Mr. Slavin concludes that "Writing to Read is a very expensive means of producing very modest and short-lived improvements in the test scores of kindergarten students."
The Riordan Connection
Although educators have long believed that ibm has yielded the school market to Apple, Writing to Read is the one ibm educational application that has consistently been popular among educators and policymakers. The recent moves in Mississippi and Louisiana have done nothing to hurt the program's popularity.
A common element in both states has been the involvement of the Riordan Foundation, a Los Angeles-based philanthropy.
In Mississippi, the foundation, in cooperation with the New York-based rord Foundation, pledged $1.5 million to be matched by the state to purchase Writing to Read laboratories. (See Education Week, April 26, 1989.)
In Louisiana, the Riordan Foundation, acting alone, offered to help the state raise the money needed to equip schools with the labs.
Richard Riordan, the lawyer who established the foundation, is a strong supporter of Writing to Read; he currently is helping the Roman Catholic Archdiocese of Los Angeles equip each of its schools with the program.
"We believe in the program," says Calvin Morrell, the foundation's president. "[Mr. Riordan would] be interested in something that could be proved [as effective as] Writing to Read, but nothing has been tested as well."
But Mr. Morrell adds that he is troubled by the perception that the foundation is acting as a middle-man for the computer manufacturer.
"We're not selling this program," he says. "If somebody wants something else, we'll go look at it. But right now that's the only program we really believe in."
Mr. Morrell says that, when asked, the foundation has supplied some districts with funding to purchase other programs. Meanwhile, he adds, the philanthropy is funding a longitudinal study through the University of Mississippi on the effectiveness of Writing to Read in that state.
"We are interested in documenting [the program's benefits] ourselves," Mr. Morrell says.
Whatever the outcome of that study, Mr. Riordan's belief in the program's effectiveness has already had an effect on the computer-buying patterns of at least one state.
Since Governor Mabus first announced the availability of Writing to Read through the auspices of his office, "the response has been overwhelming," says Olan Ray, education adviser to the Mississippi chief executive.
In the program's first year, 320 of the state's 530 elementary schools applied for Writing to Read labs; within three years the state will have completed its program of buying labs for any school that requests one, Mr. Ray says.
Unlike Governor Roemer's proposal in Louisiana, Mr. Ray adds, Mississippi deliberately avoided mandating that schools purchase the labs because it was thought that voluntary cooperation was more likely to engender enthusiasm for the program.
"It's more effective if people come in as interested, willing, eager participants," he observes.
Ms. Krendl, the Indiana University researcher, argues that research does not yet exist to support such sweeping policy decisions as have been made by Mississippi and may yet be made in Louisiana.
"You can't make policy changes like that based on the kind of data we've got right now," she says.