New Test Measures Students' Reading 'Power'
In 1982, the Boston Public Schools began administering a reading test that yields no complex subscores of "skills" and makes no mention of the grade level at which the students are reading.
That first year, officials gave the test to only a small number of students. Earlier this year, all 3rd through 12th graders in the system took the test. After the scores are analyzed, they will be given to the students' current teachers, and later to their teachers l year. Both are expected to use the scores to help devise strategies to improve students' reading comprehension.
The test is called the Degrees of Reading Power (drp). Sponsored by the College Board, it is a relative newcomer to a field that is frequently criticized for tests that tell teachers little about the progress of individual students or how well students understand what they read. The latter issue has received more attention lately in light of national surveys showing that students' abilities to interpret material lag far behind their "basic skills."
Boston was the first school system to adopt the drp, and although College Board officials can provide no statistics, they say use of the program is growing. New York State, Connecticut, and, most recently, Washington State, have begun to use it on a statewide basis.
The drp differs significantly from the other reading tests on the market in several ways, program officials say. Students are given short passages of increasing complexity with a word left out. Each of the four words provided as possible answers would make sense grammatically and within the meaning of the individual sentence. But only one word will make the sentence mesh with the meaning of the rest of the paragraph.
It is this format that allows those scoring the tests to tell whether the student understands the passage. Once the student receives his or her score, the teacher can match it to the scores of books and other materials whose "readability" has been calculated on the same scale. Both the student's level and the book's score are given in "degrees of reading power." The scale goes from 15 to 100 degrees.
The three basic components of the program are the tests, the analysis of materials by the College Board, and a recently developed inservice effort, in which board consultants work with local teachers. The board issues an annual catalogue that gives the drp rating of books and other materials; about 2,000 textbooks have been analyzed to date. The board charges $375 to analyze a book for a school district or a publisher.
"It was designed to answer fundamentally one question: 'What can you read and how well can you read it?"' said Stephen H. Ivens, the program's executive director at the College Board. "It also has norms, but that's not its purpose. We wanted to measure directly the most difficult text a student can read with comprehension."
"The key thing is that nobody has ever been able to develop a test where you could report results on a text-difficulty scale," Mr. Ivens said. "That was the technical breakthrough."
"It doesn't give you the detail in diagnostics like some other tests do," said Oliver W. Lancaster, Boston's deputy superintendent for curriculum and instruction. "But if you put it together in the composite, it tells you whether you understand what you've read. At first, it tells you whether you understand the test. But once you understand, there's not a lot of magical stuff. Can you or can't you read? It's about that simple."
Gaining Cautious Acceptance
The drp seems to be gaining cautious acceptance in the reading-research community, as well. "I think it probably does represent an incremental improvement in the measurement of reading comprehension," said Richard C. Anderson, director of the Center for the Study of Reading at the University of Illinois at Urbana-Champaign, who noted that he has not studied the test in depth.
"I can say I am guardedly enthusiastic," said Mr. Anderson, who is also the president of the American Educational Research Association.
"The problem is," he continued, "the nation has become so obsessed with testing. We're placing so much weight on these tests that no test is really up to what we expect from it."
The program has been criticized on some counts, including the issue of whether anyone besides the Col-lege Board can provide a sufficiently precise and consistent readability measure of any given book. But the key concern about the drp seems to be the possibility that teachers and publishers will treat it as an end rather than a means.
Researchers cite readability formulas--often accused of "driving" the development of basal readers--as an example of what can happen when an educational innovation becomes too strong a force in shaping the curriculum.
"We're already seeing evidence that districts are preparing drills on the degrees of reading power," Mr. Anderson said. "I heard from a colleague that a company is preparing materials to prepare for the drp test. I have to worry about whether that's educationally desirable."
Mr. Anderson also expressed concern about the possibility that textbook authors might write their books to suit a given drp level. "That's going to be a terrible thing if that happens," he said. "We're just starting to make some headway in getting rid of some of these simplistic methods. I think it almost sinful if people start putting in criteria like that."
"There's clearly potential for educational abuse," Mr. Ivens said. "Clearly, publishers could try to write texts to our formulas. The more serious abuse will come from teachers who think that once they understand drp, they can use it without interjecting their own judgment of the motivations and interests of the students and the quality of the texts.
"I worry that teachers will not exercise the care they need to in evaluating the text," he continued. "They can't use the drp as a substitute for all the other things they're supposed to do."
The key concepts in the College Board's reading initiative evolved from the work of several researchers. John Bormuth of University of Chicago developed the formula on which the drp is based in 1969. Much of the subsequent research was done by Burt Koslin of Touchstone Applied Science Associates in New York, under a contract with the state.
"The state was interested and anxious to find an effectiveness test in reading," said Winsor Lott, director of the division of educational testing for the state education department, describing the history of the program.
State officials knew of the drp when they started the state's competency-testing program in the mid-1970's, Mr. Lott said, but waited until it was further refined. Then, in the late 70's, the Board of Regents became "rather dissatisfied" with the competency standards and requested more rigorous tests, he said. State officials began phasing in the drp in 1979, and students now take the test in the 3rd, 6th, 8th, 11th, and 12th grades.
The process of refining the test was, New York officials point out, aided by funding from the Carnegie Corporation of New York, which since 1976 has provided about $2 million to the Board of Regents' research fund.
One of Several Efforts
The test was one of several efforts to develop alternatives to the standardized reading tests commonly used, according to Frederic Mosher, a program officer for Carnegie. It is also the most successful of these efforts, Mr. Mosher said.
The foundation's interest in the reading program grew out of its broader interest in elementary and secondary education, in particular efforts to help schools "do better by kids they do less well by," Mr. Mosher said. As part of that program, Carnegie officials looked at "outcomes"--the standardized-test scores.
"As we looked closely, we saw there was a problem in the kinds of outcomes that were being measured," Mr. Mosher said. "We focused on problems of standardized tests and possibility of finding better ways."
At that point, Carnegie encountered the work of Mr. Koslin. "I was very well impressed by what he'd produced, and it fit the line we'd begun to pursue in looking at what was wrong with conventional tests," Mr. Mosher said.
In 1980, with Carnegie continuing to fund research through the Board of Regents, New York gave the program to the College Board. The drp has yet to break even financially and is subsidized by the board, officials say.
Studies Not Yet Complete
The final round of Carnegie-sponsored studies, which are examining the extent to which the use of the drp improves reading comprehension, are not yet complete. But interviews with some of the school officials using the program suggest that it is regarded as a useful tool. School districts use the program in different ways; some test all students, while others may focus on students who need special help with reading, or those in a particular grade.
In Boston, Mr. Lancaster said, the district plans to use the scores to help students improve their reading comprehension. "For youngsters in urban settings, the area where we need to build the most strength is the area of comprehension--understanding and responding to inferences," he said. "The drp exam does speak directly to that, because you are given passages of increasing difficulty that you have to really understand in order to respond. It's not based on facts, or on what you're taught. We feel it's a much more realistic test. By developing a series of scores [over the years], we feel we'll get a better picture of what their needs are and what we'll need to do to improve their reading."
Teachers and administrators, he said, seem generally to like the program. He noted, however, that inservice was an important part of making it work.
A former administrator in the New York education department, Mr. Lancaster said that Boston probably avoided some of the confusion experienced by New York by making sure staff members understood the program.
In the Clark County, Nev., system, which includes the Las Vegas city schools, district officials are experimenting on "a very small scale," according to Jean Serum, who coordinates the program for the district's research and development department.
Next year, they will increase the number of students involved to 6,000, double this year's total. "We've been very happy with the information we get through the test," Ms. Serum said. "It's been an excellent tool in placing students." The district spent about $10,000 in materials for the 3,000 students.
New York officials are also enthusiastic. "I think it's a tremendous testing instrument," Mr. Lott said. "I think it's one of the soundest measures of reading comprehension I've seen and far superior to other reading tests."
"The other side is the instructional side," he continued, "and the importance of matching materials to level. That will have a really beneficial effect on the teaching of reading in this state, and is already having it."
Broad Instruction for Teachers
The major drawback, Mr. Lott noted, is that the program requires a lot of training. "It requires really broad instruction for teachers to acquaint them with the entire drp program nd how its used." The other drawback, he said, is that the drp requires extensive readability analysis of materials to be fully functional and effective. "Eventually, we'll reach that state but it's a big thing," he said.
Focus of Secondary Students
The District of Columbia is starting out on a small scale, restricting the program to 9th graders in eight schools, according to Helen Turner, supervising director of reading for the district. The district plans to focus on secondary students, she said, since those students tend to need the most work on comprehension.
"We found that many students have trouble at the secondary level," she said. "There are many higher-level skills that aren't taught." Ms. Turner praised the College Board's choice of inservice speakers, noting that their fees would otherwise exceed the district's price range.
But although school officials there seem to agree that the drp is a ''viable" program, Ms. Turner said, they plan to evaluate the results carefully before deciding whether to continue it next year. "The community is still interested in getting a grade level," she said. "And we don't want to just add another test on top of what we have."
The question about the drp that seems most likely to remain unanswered is whether school districts and publishers can adequately do the readability analysis of textbooks and other materials.
The College Board argues--strenuously--that they cannot. "If all we were doing was just analyzing texts for readability, how close the numbers come to each other would be less important," Mr. Ivens said. "But when you're measuring students' test performance, the extent to which someone else makes errors in the calculation of the score jeopardizes the validity of the test. The whole scaling of those test results to the text readability scale is based on the development that we did of the readability scale."
Idiosyncratic bits of texts, Mr. Ivens said, must be treated consistently. "Fundamentally, how you do all these things doesn't make any difference if you do it the same way every time it happens. That's one thing we do, is make sure the formula is applied consistently across texts," he said.
"We did publish the Bormuth formula," he noted. "Everyone has access to it, and they can try to do what we do. The real question is whether they can end up with what we end up with."
Others, however, respond that the College Board is overstating its case.
'Nothing But Market Puffery'
"They're regarded as a group in testing with some savvy, but to say that no one else can do it would be nothing but market puffery," Mr. Anderson said. "They're using a fancy new statistical model. It's probably an improvement on the classical psychometric model, but it's not magic. There are certainly other individuals, places who could do it."
"I don't think that there's anything that they do that is so sophisticated that it can't be taught to those of us out here," said Roger Farr, a reading researcher at Indiana University. Mr. Farr is currently working on a study of the drp
In Boston, school officials have opted to bypass the College Board's analysis in favor of one conducted by a principal in the district who uses a computer model.
Mr. Lancaster pointed out that the principal was given 15 books to analyze that, unbeknownst to him, already had College Board ratings. For 11 of the books, he came up with identical scores; four were one point off.