69传媒

Over the past decade, the state of Nevada has spent almost $500 million to keep class sizes under 20 students in the early grades. While four evaluations of the program have been conducted since 1993, officials at the state education department still have little proof that the long-running initiative is really improving student achievement.

鈥淭he problem is, they鈥檝e never given us enough money鈥� for a comprehensive study, says Mary Snow, an evaluation consultant at the Nevada Department of Education. Even the studies that were conducted were paid for by school districts, not with funds appropriated by the legislature.

State officials, though, are still proposing a detailed research plan, and Gov. Kenny Guinn has asked lawmakers for $374,000 to do the work.

In recent years, reducing class sizes in the early grades has risen to the top of the nation鈥檚 school improvement agenda. At least 20 states, as well as the federal government, now have class-size-reduction initiatives, and teachers鈥� unions tout the approach as the best alternative to private school vouchers.

Already this legislative season, several governors have called for more spending to keep such programs going or to launch new ones. For example, in Georgia, Gov. Roy E. Barnes wants to spend $18.9 million to extend his year-old class-size-reduction initiative into the 4th and 5th grades. And in Colorado, Gov. Bill Owens wants class-size reduction to be paid for with a portion of the revenue from a recent ballot measure that is expected to bring in $4.5 billion over 10 years.

To find out whether their investment in smaller classes is paying off, some states鈥攕uch as California, Tennessee, and Wisconsin鈥攁re also spending millions of dollars on research.

But Nevada has not been alone in displaying little enthusiasm for conducting formal evaluations of its class-size initiative. And even in states that are conducting such studies, some observers wonder whether researchers are asking all the right questions as they examine the impact of fewer students per class.

Research Money Hard To Find

New York state is in its second year of implementing a voluntary class-size-reduction plan that was passed as part of a comprehensive school improvement package in 1997. More than 300 of the 700 districts in the state are now participating in the program, which is targeted first at low-income communities.

Funding for the effort, which brings class sizes to an average of 20 pupils in grades K-3, stands this year at $140 million. Gov. George E. Pataki wants to raise that to $225 million in fiscal 2002.

Money for research, however, has not been as easy to obtain.

Gary Pollow, a supervisor of education programs with the New York education department, says he has had to piece together some funding from within his office to get a research project going. He has even asked the U.S. Department of Education for a waiver to use some of the money from the federal class-size-reduction program for the study, which will be conducted by Syracuse University.

Researchers will compare achievement levels鈥攖o be measured by scores on state tests鈥攐f students in the smaller classes with students鈥� performance in the years before the program began.

Another focus of the project will be to examine the qualifications of the 4,000 teachers who have been hired for the initiative.

Increasingly, both educators and some researchers say they are convinced that teachers, and what they do in smaller classes, are the real key to whether class-size reduction improves student achievement.

Simply reducing the size of a class may not be enough. The real payoff appears to come when teachers shift their practices to take advantage of having fewer students.

Perhaps the strongest research support for that viewpoint so far came earlier this year from Wisconsin. Researchers investigating the state鈥檚 Student Achievement Guarantee in Education, or SAGE, program found that some teachers in reduced-size 1st grade classrooms were more effective than others in the smaller classes.

Specifically, pupils performed at higher levels when teachers used one-on-one contact to focus on strengthening basic skills, provided frequent feedback, and asked children to discuss and demonstrate what they knew, the researchers found. The less effective teachers spent more time on problem-solving and project-oriented activities, and gave students more choice.

鈥淩educing class size is the necessary precondition,鈥� says Alex Molnar, one of the three principal investigators for the SAGE study and the executive director of the Center for Education Research, Analysis, and Innovation at the University of Wisconsin- Milwaukee. 鈥淭hen you follow up with staff development.鈥�

Shift in Strategy Urged

Administrators trying to carry out California鈥檚 class-size- reduction initiative,which sets the maximum pupil-teacher ratio in the early grades at 20-to-1, are also learning that lesson.

鈥淗aving smaller classes certainly provides more opportunities for students to be known and to have better communication with families,鈥� says Neil Schmidt, the superintendent of the 12,500-student Santa Monica-Malibu Unified School District. 鈥淏ut if it ends up doing that depends on whether something parallel happens, and that is staff development of teachers and strategies being changed.鈥�

California is spending a lot of money to find out whether its $4 billion class-size-reduction initiative is making a difference.

The RAND Corp., a research organization based in Santa Monica, Calif., and the American Institutes for Research in Palo Alto, Calif., are leading a four-year, $2.5 million research project that is focusing on a range of issues associated with the effort to reduce class sizes. The state is contributing $2 million to the research, while the rest of the money comes from a combination of federal and foundation grants.

While slight gains in student achievement have been reported since the initiative began in 1996, most of the findings have focused on the problems districts have had in putting the program in place. A shortage of qualified teachers and a lack of adequate classroom space have led some observers to label class-size reduction鈥� at least in California鈥攁 failure.

鈥淭hey rush in to reduce class size everywhere, and then they have no ability to say anything about its impact,鈥� says Eric A. Hanushek, a senior fellow on education policy at the Hoover Institution at Stanford University.

A more responsible approach, Hanushek says, might be for states to phase in such programs so they can first conduct controlled experiments to find out if smaller classes are worth the expense.

Hanushek also argues that reducing class size is not nearly as effective as improving the quality of teaching.

Class-size reduction, he adds, is popular because it鈥檚 a 鈥渟imple, easy program鈥� to communicate to state legislators. And while policymakers often include staff development as a part of such policies, states aren鈥檛 giving that aspect of smaller classes serious thought, Hanushek suggests.

鈥淐alifornia鈥檚 law says that teachers should have staff development, but there鈥檚 never been any clarity about what that means,鈥� adds Joan McRobbie, the California liaison at WestEd, a federally financed education research laboratory in San Francisco and one of the organizations involved in the California study.

鈥楢n Important Question鈥�

One researcher in California says he hopes his team will eventually be able to take a deeper look at the issue of teaching practices, including whether some teachers are more effective than others at making the most of smaller classes.

鈥淚 think it is an important question,鈥� says Brian M. Stecher, a senior social scientist at RAND. 鈥淏ut we鈥檙e limited by what we鈥檝e promised to do.鈥�

So far, Stecher and Cathleen Stasz, a colleague of his at RAND, have found few major differences between the practices of teachers in reduced-size classes and those with more students.

Teachers in both situations generally covered the same material in mathematics and language arts, and covered those topics for the same length of time, their research has found. But teachers in smaller classes spent a little more time working individually with children who were identified as poor readers. As similar comparison studies in other states have shown, the California researchers also found that teachers in the smaller classes spent less time disciplining students.

Some researchers maintain that simply having fewer students automatically makes teachers more effective.

鈥淭eachers do things differently in small classes, and they do some of the more important things more differently,鈥� says Barbara Nye, a professor of education at Tennessee State University and one of the principal investigators for the best-known study of class-size reduction, Tennessee鈥檚 Project STAR.

Getting to know parents, because there are fewer of them, is one thing teachers in smaller classes can focus on more intensely, Nye says.

The researcher agrees, though, that another frontier of research remains to be crossed: looking at the teachers who have the best gains in student achievement in smaller classes, and then sharing that knowledge with others who are also teaching in those classes or will be in the future.

Both Nye and Molnar, however, note that it鈥檚 unrealistic to expect the lessons learned from studies on reduced class sizes to be very useful to teachers in larger classes.

The impact of class-size reduction on teachers will also be part of a study on the federal government鈥檚 effort to lower class sizes. That Clinton administration initiative aims to bring classes in the early grades down to an average of 18 students nationally by providing money to help hire additional teachers. For the current fiscal year, Congress appropriated $1.6 billion for the effort, which was launched in 1998. Researchers are collecting data for the study鈥攚hich is being conducted this winter and spring鈥攁nd hope to release their findings next fall.

Abt Associates Inc., a Cambridge, Mass.- based research organization that is conducting the federal study, intends to 鈥渄ocument changes in instructional practices through extensive classroom observations and self-reports of those teachers observed,鈥� according to a summary of the project. The researchers will also look at efforts to train teachers.

Other Angles To Pursue

One state is pursuing a slightly different research question about teachers鈥� practices.

Because Indiana鈥檚 initiative, which is called Prime Time, pays for classroom aides to bring down the pupil-teacher ratio in K-3 classrooms, researchers at Ball State University in Muncie, Ind., are looking at how teachers are using those assistants鈥攚hether to handle clerical duties or to provide instructional help for students. The researchers will present their findings in April at the annual meeting of the American Educational Research Association.

An early review of the data shows that teachers are using their classroom aides, in most cases, to instruct students.

鈥淲e鈥檙e not finding that they are just collecting the lunch money,鈥� said Daniel K. Lapsley, the chairman of the educational psychology department at Ball State.

Prime Time, which began in 1984, was one of the earliest efforts of its kind in the country. Just a few years ago, the state legislature reduced the 鈥渢arget ratio鈥� of pupils to instructors (which includes teachers and teachers鈥� aides) to between 15-to-1 and 18-to-1. The average ratio had been as high as 20-to-1.

In addition to conducting a survey of teachers throughout the state, the researchers studying Prime Time intend to spend time observing classrooms, Lapsley says.

Other projects under way include one in Georgia, where instead of focusing just on children entering public school, the researchers at Georgia State University鈥檚 Applied Research Center will be tracking the progress of children who first attended the state鈥檚 publicly financed prekindergarten program.

Past studies have shown that the achievement benefits of that preschool experience begin to fade in the 2nd grade, especially for children in what the state used to call its Special Instructional Assistance Program, which served children performing below grade level.

鈥淥ur pre-K kids were in classes of about 18 kids,鈥� says Gary T. Henry, the director of the Georgia State center. 鈥淭hen many of them went into classes that were larger than 24 students.鈥�

The researchers will begin with a sample of 500 prekindergartners this coming fall and then follow some of them into regular classes in the early elementary grades, which have also been reduced to fewer than 20 pupils per instructor. They will also follow other children who are participating in the state鈥檚 new Early Intervention Program, which is replacing the Special Instructional Assistance Program. The average student-teacher ratio in the EIP is 11-to-1.

Whatever direction states take with class-size reduction, Nye of Tennessee State says, the public shouldn鈥檛 necessarily expect the same results from classes of around 20 students as those of 15 or even lower. 鈥淚t鈥檚 taken a lot of time to get that message across,鈥� she says.

That point became important in recent years as policymakers sought to extrapolate from studies of Tennessee鈥檚 student-teacher achievement ratio鈥攐r STAR鈥攑rogram. That controlled experiment began in 1985 and included 79 elementary schools in 42 districts. 69传媒 were randomly assigned to small classes of 13 to 17 pupils, large classes of 22 to 26 students, or large classes with a full-time aide.

After four years, researchers found that the smaller classes were associated with higher academic achievement than the larger classes, especially for minority students. Ongoing research has found that children who were in the smaller classes continued to outperform their peers from larger classes, even after the students entered classes with larger pupil-teacher ratios.

When it comes to the future of class-size research, much will depend on whether states spend enough money to pay for sound studies, says McRobbie of WestEd. 鈥淧eople don鈥檛 study it if funders aren鈥檛 funding the studies,鈥� she observes.

Some educators and researchers worry that without further studies yielding evidence that smaller classes improve academic results, state leaders could begin wondering whether the money directed to cutting class sizes might be better spent elsewhere.

Yet McRobbie predicts that even without more research, the popularity of class-size reduction is unlikely to fade anytime soon. 鈥淚f there was no change in test scores, there would still be people who would argue that it was a good idea,鈥� she says.

The Research section is underwritten by a grant from the Spencer Foundation.