U.S. Ranks High In International Study of Reading
In a rare piece of good news from an international comparison of student achievement, students in the United States outperformed those from nearly every other country in a 32-nation study of reading literacy.
The study, the largest of its kind ever conducted, found that, among 9-year-olds, only those from Finland outperformed those from the United States, while American 14-year-olds were outscored by those from Finland, France, Sweden, and New Zealand.
Previous international comparisons had shown American students ranking at or near the bottom in mathematics and science, and had sparked widespread concern over the quality of education in the United States.
Alan C. Purves, the U.S. director of the International Association for the Evaluation of Educational Achievement, or I.E.A., which conducted the new study, noted that the United States' ranking in the reading results may have benefited from the fact that few limited-English-proficient students were tested.
Nevertheless, he said, the study highlights a success story in American education.
"American kids have always done pretty well in reading,'' said Mr. Purves, the director of the center for writing and literacy at the State University of New York at Albany. "We've done a pretty good job teaching reading.''
Marilyn R. Binkley, a senior associate at the National Center for Education Statistics and the U.S. national research coordinator for the reading study, said the results are consistent with those of the National Assessment of Educational Progress. Recent îáåð results, she said, have shown that most American students perform at the basic level of reading, the same level the I.E.A. test measured.
Other researchers pointed out that the I.E.A. study appears to support the view that recent trends in American reading instruction--including an emphasis on literature, rather than word-decoding skills--contribute to reading literacy.
Greater access to books in homes and school libraries, more time spent on reading instruction, and an emphasis on silent reading in class were all associated with high scores on the test, the study found.
But it also found that several factors thought to be associated with literacy--such as an early starting age of reading instruction, little television viewing, and a longer school year--are not necessarily linked with high achievement.
And it showed a weak relationship, particularly at age 14, between performance on the test and voluntary reading outside class.
"The message here is that teaching students skills in reading does not necessarily endow them with a desire to read,'' said John T. Guthrie, the co-director of the National Reading Research Center at the University of Maryland and a member of the steering committee of the I.E.A. project.
The study, "How in the World Do Students Read?,'' was based on a test given in 1990-91 to 210,000 9- and 14-year-olds in 32 school systems around the world. In the United States, 6,848 9-year-olds from 165 schools and 3,587 14-year-olds from 165 schools took part in the assessment.
Although previous international studies have been criticized for failing to compare similar populations of students or to reflect participating countries' curricula, the I.E.A. researchers took pains to insure that the reading-literacy study was valid, according to Mr. Guthrie.
"The methodological rigor was unprecedented in any international study,'' he said.
The researchers compiled a detailed set of standards for selecting samples of students, developed test items based on a consensus of all nations, and field-tested each one a year in advance, Mr. Guthrie said.
The test included multiple-choice questions--66 for the younger students, 89 for the older group--that tested students' understanding of written passages in three areas: narrative prose, ranging from short fables to stories of more than 1,000 words; expository prose, including brief family letters and lengthy treatises on smoking and lasers; and documents, such as maps, graphs, lists, and sets of instructions.
The 9-year-olds' test also included a word-recognition test, which required students to match simple, individual words with pictures, in order to determine if weaknesses in comprehension were due to students' inability to identify words.
The tests at both ages also included a handful of short-answer and open-ended questions, but because of concerns over the reliability of such test methods, the researchers did not include these in the international data.
The results of those questions, along with a detailed analysis of the results for the United States, are expected to be published in January by the N.C.E.S.
Finland at the Top
The international study found that 9-year-olds from Finland outpaced those from all other nations on all three dimensions of reading literacy.
The Finns' performance is particularly noteworthy, the report notes, since Finnish students begin school at age 7, and only half of the students can read at that age.
The report suggests that the highly regular symbol-sound relationship of the Finnish written language may be a factor facilitating reading instruction in that country.
But it notes that English has perhaps the most irregular orthography of any language, yet 9-year-olds in the United States performed well on the assessment.
Jerome C. Harste, a professor of language education at Indiana University, said the Americans' performance underscores the idea that a child's understanding of language involves meaning more than orthography.
"It only looks as though some languages are harder, because of what we think of as the codes that need to be broken,'' he said.
The I.E.A. study also notes that students from Singapore performed well, even though most of them speak Chinese, Malay, or Tamil at home but learn English at school.
Their performance, the report states, "must challenge the conventional wisdom that students should learn to read first in the language of the home.''
By contrast, it notes, Indonesian students--for nearly 80 percent of whom the national language is their second or third language--performed poorly on the assessment.
In the United States, 3.5 percent of the 9-year-olds and 3.8 percent of the 14-year-olds spoke a language other than English at home. Their achievement, particularly in the upper grades, tended to be lower than the national average.
But Ms. Binkley said the rules for excluding limited-English-proficient students were the same as those used for NAEP.
"I'm convinced this was an extremely representative national sample,'' she said.
Drop in Document Literacy
Like their younger counterparts, 14-year-olds from Finland also led the world in reading literacy, followed by those from France (of whom only public school students were tested), Sweden, and New Zealand.
Americans' performance dropped off by age 14, largely because of lower performance on document literacy.
"U.S. teachers totally ignore that source of information for students,'' said Mr. Guthrie.
But the study also provided evidence that there may be gender differences in that type of reading. Although girls outperformed boys in all countries among 9-year-olds, and in all but two countries among 14-year-olds, the smallest gaps were in the documents domain.
And, it found, 9-year-old boys outperformed girls in document literacy in the United States, as well as in France, Western Germany, the Netherlands, Spain, and Switzerland.
In examining factors associated with high performance, the study found, as expected, that countries that rank high on economic and social factors such as national wealth and life expectancy tended to be judged the most literate on the tests.
But it also found that, even after adjusting for economic differences, several education-policy factors were also linked to achievement.
Nations with large classroom libraries, that spent a greater amount of time on reading instruction, that provided more time for silent reading in class, and where teachers tested students often tended to perform relatively well on the 9-year-old test.
But several factors thought to affect students' literacy--such as attendance at preschool--appeared to have little relationship to performance.
And, in findings that appear counter-intuitive, countries with longer school years tended to perform less well than those where schools are open for fewer than 180 days per year, and countries with larger class sizes fared better on the assessment.
Examining television viewing, the study found that, in general, light viewers tended to score higher on the test. Americans far outpaced other the rest of the world in the amount of television they watch.
But it also showed that moderate viewing--up to 3 1/2 hours daily--has little effect on reading literacy.
Copies of the report are available for $4 each, plus postage, from
the I.E.A. International Coordinating Center, University of Hamburg,
Sedanstr. 19, D-2000, Hamburg 13, Germany.