The Wisdom of Grouping
Harvard statistician Frederick Mosteller made waves last year when he undertook a review of the major studies that looked at whether class size affects children's learning.
He found only one large-scale, long-term study that met his criteria. But it was strong enough, he concluded, to answer the class-size question definitively. In a report circulated by the American Academy of Arts and Sciences, he stated his conclusion: When it comes to class sizes, the smaller, the better. ("Less is More," July 12, 1995.)
This year, Mr. Mosteller, a professor of mathematical statistics emeritus, and his colleagues are at it again. This time, the target is the common instructional practice of grouping students according to their ability. "An overwhelming majority of American students are skill-grouped for math instruction," Mr. Mosteller says. But "are these wise decisions?"
The answer, he concludes, is that nobody really knows: "The appropriate, large-scale, multisite research on skill grouping has not yet been carried out, even though the issues have been debated as major public concerns within education for most of a century."
In reviewing the work done on skill grouping so far, the Harvard University researchers found that the pickings were paltry.
They came across only 15 studies that they considered true experiments. In other words, the studies involved field trials that compared skill-grouped classrooms with regular classrooms. But the studies were mostly small and short-term. What is more, they dealt with different variations of grouping practices.
Ten of the studies, for example, focused on "XYZ grouping," which, Mr. Mosteller says, is a lot like tracking. Students within a grade are grouped according to two or three levels of skill, such as high, medium, or low.
Another two studies examined cross-grade grouping. With this kind of arrangement, students in a reading group might be matched with pupils who are working at the same level but who might be in a different grade.
The remaining three studies focused on within-class grouping--the practice of breaking students in the same classroom into subgroups for reading, mathematics, or some other subject.
Predictably, the researchers found, some forms were more effective than others. Among the more successful was cross-grade grouping--putting students of similar abilities together regardless of their age. A typical child in a cross-grade grouping arrangement--in other words, a student whose test scores are at the median--scores better than 62 percent of students in more heterogeneously grouped classrooms.
Across all of the studies, however, Mr. Mosteller says, "we find that little clear evidence exists that skill grouping has a major impact, either positive or negative, on students' learning."
Mr. Mosteller and his colleagues, Richard Light and Jason Sachs, have completed a paper on their analysis, which is scheduled to be published this month in the Harvard Educational Review. The ability-grouping study, like the study on class sizes before it, was supported by the American Academy of Arts and Sciences.
Boys and Books
It's a common assumption in this country that young boys have more trouble with reading and writing than their female classmates do. They generally score lower on standardized reading tests and are referred in disproportionate numbers to remedial reading classes and special education programs.
One explanation for the differences, educators have long suspected, is that, academically speaking at least, boys mature more slowly than girls do.
But, if that's true, wondered researcher Donald D. Pottorff, why don't young girls in other countries have a similar edge in reading? He says studies conducted in Germany, Nigeria, Canada, England, Israel, and Finland have all shown, to the contrary, that boys do just as well or better than girls in reading.
So Mr. Pottorff, an associate professor of reading at Grand Valley State University in Allendale, Mich., set out to test a different explanation for the gender differences in reading between children in the United States. His theory: Boys see reading as a female-dominated activity.
Mr. Pottorff and his colleagues, Deborah Phelps-Zientarski and Michele E. Skovera, surveyed 730 students in grades 2, 4, 6, and 8 from a cross-section of rural, urban, and suburban schools. Students were asked: Who is better at reading books, reading to younger children, and writing stories--boys or girls? The answer, in all grades surveyed, was girls.
Then the researchers asked students whether mothers or fathers are more likely to read books, magazines, and newspapers, or read books to children. Children picked mothers for all but one category. Fathers, they said, were more likely to read newspapers.
"If reading and writing are seen by boys as gender-inappropriate, then boys may well avoid these activities as much as possible, or at best simply tolerate them," the three researchers concluded in a report on their survey published in the Summer 1996 issue of the Journal of Research and Development in Education.
Efforts to change those perceptions, they conclude, should begin at home: "Fathers, grandfathers, and other significant male role models need to be persuaded to get involved in reading to children at an early age."
Down, But Not Out
Most students who fail a minimum-competency test in high school are no more likely to drop out of school than classmates who pass the test the first time around.
So say researchers Bryan W. Griffin and Mark W. Heidorn in a report published in the Fall 1996 issue of Educational Evaluation and Policy Analysis. Mr. Griffin, an assistant professor at Georgia Southern University, and Mr. Heidorn, the program director for K-12 assessments with the Florida Department of Education, studied data on 76,000 students in grades 10-12 from 14 Florida districts.
They found that among low-achieving students there was no statistically significant difference in dropout rates between those who passed the state-mandated test and those who failed. Likewise, minority students were no more likely to drop out as a result of flunking the tests.
Surprisingly, the researchers found one group of students who, it appeared, were adversely affected by test failure: those who get good grades.
"Following a competency-test failure, students may experience a substantial drop in self-esteem, or they may feel embarrassed in front of their peers," the authors write. "And such experiences might be especially acute for students with a proven record of academic success."
But these students numbered no more than "a few dozen," Mr. Griffin said, and it was likely that other things were going on in their lives about which researchers had no clue.
The two researchers say their study is important because critics of minimum-competency testing often argue that failure on such tests will push minority and low-achieving students out of school.
"But I don't see this as a stumbling block for students at all," Mr. Griffin said in an interview. In Florida, he noted, students can retake the tests several times. Only 1 percent of students continue to fail the tests in their senior year.
The tests, which require students to perform tasks such as finding the main idea in a paragraph or selecting the better grocery-store bargain, are hardly a challenge, Mr. Griffin added. "It's amazing to me that you have students failing these things at all."