Researchers and Publishers Explore Ways To Correct Deficiencies in Reading Texts
Rye, NY--Two years after a group of researchers first began telling them about the fundamental problems of basal reading texts, education publishers say that although they are very concerned about the situation, correcting it is more easily said than done.
In the third meeting designed to foster a dialogue on the subject, publishers and reading researchers gathered here last week to discuss the issues identified in the research, and to consider possible solutions.
Sponsored by the Association of American Publishers, the two-day meeting brought together representatives of 44 education publishing companies and about 30 reading researchers from institutions around the country.
The problems the two groups addressed have emerged in research on basal, or beginning, readers and the workbooks and manuals that accompany them. The studies were conducted by researchers from the federally supported Center for the Study of Reading at the University of Illinois in Urbana-Champaign.
Over the past six years, the researchers have identified numerous deficiencies in the various series of books used to teach reading in virtually all elementary schools.
The problems, according to a number of studies, include the misuse of "readability" scores and formulas, which can lead to oversimplified texts; teachers' manuals that include very little "comprehension" instruction; and poor workbooks. (See Education Week, Nov. 23, 1981.)
At last week's meeting, the publishers agreed that much of the criticism is valid. But they pointed out that publishers can produce better materials for teaching reading only if those who buy and use textbooks understand both the implications of the research and the limitations of standards currently used to judge reading books.
Any long-term solution, they said, will probably include changing the processes school districts use to select texts and informing educators of the reasons for the changes.
Currently, the publishers said, the pressures of the marketplace, the demands of state textbook-adoption committees for "objective" information on the difficulty of textbooks, and other factors all slow the rate at which they are able to change their products.
Although some of the issues will probably not be resolved for a long time, the two groups identified several broad strategies to improve the situation:
Educators, including members of state textbook-adoption committees, should be told of the limitations and abuses--such as using the formulas as the sole criterion for textbook selection--of "readability formulas." Researchers should work on developing new criteria that educators can use to judge the relative difficulty of texts.
Publishers, researchers, and educators should all work to change the current "unwieldy" adoption processes.
Publishers should continue to address the problems researchers have identified in teachers' manuals and workbooks.
And finally, publishers, if they are going to make constructive changes, need models of good materials. Researchers who offer criticism should be prepared to make concrete suggestions for improving the texts they dislike.
Fundamental Premises Questioned
The researchers first sought--and got--the publishers' attention in February 1981, when the two groups met in Tarrytown, N.Y. At that meeting, they described study results that raised questions about the fundamental premises on which books for beginning readers are based. But the researchers emphasized that they were aware that publishers shared responsibility with educators and others.
"We have been and we will continue to be highly critical of school materials as they are now," said Richard Anderson, director of the Center for the Study of Reading and professor of educational psychology at the University of Illinois. "But we do not single out publishers as the villains who do not do their job."
Hence, the discussion focused on two issues: Considering the demands of the market, what can publishers do now to improve basal reading materials? And how can the two groups work together to inform educators of the issues?
The overuse and abuse of "readability formulas," both researchers and publishers agreed, is one problem that educators should be made aware of promptly.
Although not all reading researchers agree, most of those attending the meeting concurred that such measurements should be applied only in very narrow circumstances. Some were highly critical of the formulas, which one researcher described as being "about as scientific as Alice in Wonderland."
Based on theories developed in the 1920's and 1930's, the various formulas are designed to quantify the level of difficulty of a particular piece of writing, according to Andee Rubin, a cognitive scientist with Bolt Beranek, a private research firm based in Boston.
Although they vary in their precise design, all of the formulas are based on the same premise: longer words and longer sentences make a piece of writing more difficult to read, Ms. Rubin said. The "readability score" assigned to a given piece of writing is usually calculated on the basis of a weighted formula. Word difficulty may be measured against a list of "familiar words"; sentence difficulty is generally measured by the length of the sentence.
The score that results from the calculations identifies the grade level for which a given text is appropriate, Ms. Rubin said. But she and other researchers emphasized that the score is intended to indicate a range--that is, the material could be suitable for students several grades higher or lower.
The formulas are used mainly in two ways, Ms. Rubin said: to "match readers with appropriate texts" and to "produce appropriate texts through [adaptation] or construction." Virtually all publishers use them in developing texts, and many teachers use them to assess material, according to Ms. Rubin.
The formulas, according to Ms. Rubin, have inherent problems, principal among them the fact that they are constructed on a "shaky statistical basis." Part of the process of determining "readability" involves comparing material with "other independent measures" of difficulty. But because the other measures are no more statistically valid than the formulas, a "kind of circularity" is introduced into the process, Ms. Rubin said.
In addition, studies have shown that "the variables that readability measures are not the things that make a text hard or easy," she said. Other factors, which are not incorporated into the formulas, are equally important, many researchers say.
For example, the inferences that a reader must draw, the complexity of the ideas presented, and the structure of the material all affect the ease with which a reader can understand it, the researchers argue.
Educators and publishers compound the problemmatic quality of the formulas by misusing them in several ways, Ms. Rubin said. Most often, someone uses the formulas to modify material to make it more "readable."
For example, a teacher or text developer may wish to adapt a story--perhaps a chapter from a respected children's book--for use in the first grade. He or she may attempt to lower the readability level by shortening the sentences, replacing connective words with periods, deleting nonessential words, and the like. When a readability test is rerun, the results may show that the level is two grades lower than it was to begin with.
But that result, Ms. Rubin said, does not mean that the material is easier to read and to understand. In some cases, she and other researchers stressed, the material may be more difficult because it is choppier and lacks a cohesive structure.
The formulas are also misused in constructing reading texts, Ms. Rubin said. Text developers may choose words to teach students about a particular vowel sound, not because they are the best words to convey meaning. "The formulas are used even more freely in constructing texts for poor readers," she said. "The result is okay phonetically but frustrating on a comprehension level."
To minimize misuse, Ms. Rubin suggested, educators and publishers should use readability "only as a rough guide to text selection and ordering, only when all other pedagogical considerations have been applied."
She also recommended that readability formulas not be used for adapting existing texts, constructing new texts, or as "dominant selection criteria."
Heavy Reliance on Scores
The publishers are acutely aware of the problems of readability formulas--but they are equally aware that textbook selection committees rely heavily on the scores when choosing which reading texts to buy, according to Nancy Sargeant, executive editor for reading for Houghton Mifflin Company.
Since many states approve texts on a statewide basis, the failure to provide adequate readability information can jeopardize sales worth thousands of dollars, she said.
"All of us are confronted with the issue of readability somewhere in the publishing process," said Ms. Sargeant.
"I agree with many of the criticisms. But I'm going to talk about the reality of the situation--the marketplace."
"There is a growing emphasis on scores" in every facet of education, not only in the area of readability, Ms. Sargeant said. "A number is what's important."
"We all know that shortening sentences and removing connectors lessens comprehension," she said. But if publishers try to explain that syntactic difficulty or other factors also should be considered when assessing materials, textinued on Page X