Smaller Classes Aid Test Scores, Results Show

Students who participated in the state’s class size reduction program fared modestly better on last spring’s state reading and mathematics tests than those who did not, according to data released Monday.

In third grade, for instance, 41% of students in classes limited to 20 pupils scored at or above the national average in reading, compared with 33% in larger classes, preliminary analysis of statewide test data shows.

The provocative findings, released in response to a public-records request by The Times, were encouraging to advocates of a state program that has consumed billions of tax dollars since 1996 to reduce the size of classes from kindergarten through third grade.

State Supt. of Public Instruction Delaine Eastin, a longtime proponent of smaller classes, called the results from the Stanford 9 achievement tests “a positive sign that we’re moving in the right direction,” said her spokesman, Doug Stone.

But how much of the performance gap resulted from the lower student-teacher ratio, and what role such factors as poverty and lack of English fluency might have played, remain unknown.

Schools in urban areas have faced many more obstacles to reducing class size--including shortages of classrooms and qualified teachers--than their suburban counterparts.

Bruce Fuller, co-director of Policy Analysis for California Education, which is participating in a more in-depth, state-funded study of class size, said the results released Monday raise as many questions as they answer.

“It suggests that class size reduction may be paying off,” Fuller said. “However--and this is a big however--schools that have been slow to implement class size reduction likely differ from schools that have rapidly implemented [it] in all sorts of ways.”

The new data are emerging as Gov.-elect Gray Davis, who takes office next month, prepares to call the state Legislature into special session to tackle problems facing the nation’s largest state public school system.

The size of public school classes has become a national issue in recent years at the urging of teachers who say they could do a better job if they were given fewer students. The newly approved federal budget includes $1.1 billion to cut class size in cities across the country.

In California, the state has allocated more than $4 billion since 1996 to class size reduction in an unprecedented initiative to improve reading and math skills. The program is extremely popular among parents, teachers and politicians. More than four out of five eligible students now participate in the program.

State officials emphasized that the data released Monday are preliminary and do not establish a definitive link between smaller classes and higher student achievement.

But the data show:

* In second grade, the earliest in which students statewide were tested last spring, 41% of the students who were in smaller classes scored at or above the national average in reading, compared with 35% who were not in such classes. Average is defined as the 50th percentile of a nationally normed test group.

* In math, 44% of the second-graders who were in smaller classes scored at or above the national average, compared with 38% who were not. For third-graders, the math difference was 43% to 36%, again favoring those in smaller classes.

* In both grades, students who spoke English only or were fluent in English and another language may have benefited more from the smaller classes than their counterparts who were not fluent in English. Limited-English students in smaller classes posted markedly smaller gains. For example, 10% of limited-English students in smaller third-grade classes reached or surpassed the national average in reading; 9% of such students in larger classes did as well.

Indeed, English fluency--or the lack of it--is almost always a factor in the public school system. Students who have not mastered English account for one-quarter of the state’s total enrollment of 5.7 million.

The newly released data showed in more detail than ever before the stark difference between how English-fluent students and non-fluent students performed on the Stanford 9 exams.

Consider the scores for eighth-graders.

In reading, the data show, 56% of eighth-graders who are native English speakers scored at or above the national average, while 7% of those classified as limited-English-proficient did as well.

Substantial, though slightly narrower, achievement gaps appeared in the mathematics test scores.

Educators say that those discrepancies reflect the stiff challenges facing a state in which students speak a profusion of languages at their homes. Many students were simply unable to follow tests given only in English. What’s more, fewer than 2% of the students in the national sampling group against which the California students were judged had limited English skills.

State officials said Monday that they still do not have a complete picture of how students from various language backgrounds performed on the tests. Mistakes in the handling of score sheets fouled up the demographic profiles of as many as 18% of more than 4 million students from grades 2 to 11 who took the tests last spring.

The state had planned to release detailed reports last summer on student performance broken down by gender, ethnicity, English-language proficiency, poverty and other factors. But a delay arose when state education officials pressed the company that administers the test to plug the 18% hole.

Some of the missing demographic data were later recovered, although not in time for the report released Monday.

Harcourt Brace Educational Measurement last week sent the corrected data files to the state. Gerry Shelton, a testing administrator for the state Department of Education, said more complete analyses of student performance will probably be made public starting in late January. The testing program in the last school year cost the state about $30 million, Shelton said. A new round of state testing is to begin in March.

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

Better Scores

Preliminary test score data indicate that students in smaller classes perform slightly better on standardized tests. The figures below show the percentage of California third-grade students scoring at or above the 50th percentile, which is the national norm.

READING

Classes capped at 20 students: 41%

Classes not capped: 33%

*

MATH

Classes capped at 20 students: 43%

Classes not capped: 36%

Source: State Department of Education