Cooperative Learning and Class Size: Impact on Student Achievement

Introduction and Research Rationale

As educators in middle school and high school classrooms, content specialty teachers often work with a variety of class sizes. Yet, with such an assortment of class sizes, there are also extraneous variables that each teacher must consider in order to foster individual achievement. Participation and achievement are variables of individual students that weigh heavily on class success and are affected by class size. Educational mandates, as well as individual school district policies and requirements, also influence class size and affect individual achievement. Middle school and high school content specialty teachers are frequently searching for new ways to prevent individual achievement and participation from declining when class sizes rise. The varying number of students that content specialty teachers encounter from class to class drives them to seek out alternative methods to meet the needs of their learners in order to maintain individual success and achievement while promoting participation. Therefore, smaller class sizes may be an effective way to encourage students to participate and promote individual achievement.

Owing to rapidly growing enrollment, as well as towering tuition becoming a major concern for learners, teachers, and administrators at many academic institutions in the American education system, the search to determine the most successful means of providing high-quality education at an affordable cost continues. The reduction in government funding affecting numerous public schools, colleges, and universities nationwide has, time and again, given rise to a reduction in the number of course sections offered and an increase in class sizes within existing programs (Weimer, 1990).

Concurrently, over the past two years, class sizes have gradually climbed within the school district in which this researcher is currently working. The school district primarily includes students from low socioeconomic backgrounds. Like many similar urban schools, this district suffers from poverty, a large minority population, low test scores, a high number of discipline referrals, and many students who are not learning to read. Union contracts often stipulated maximum class sizes; however, even before the expansion of collective bargaining took place, there was widespread agreement that having more than thirty students in a class is a heavy burden for a conscientious teacher — especially for a middle or high school teacher who has five or six classes. This is the first year that the district has no cap on the number of students permitted in a classroom each period. Because of this new guideline, this researcher sees as many as thirty-two students one period and as few as twenty-one another period. Such numbers present difficulty in maintaining classroom management, teaching style, assessment, classroom configuration, and materials. But by far the greatest concern with classes as large as thirty-two students is participation and individual achievement. The challenge is to hold the interest of and promote the achievement of thirty-two students in a classroom that is at times overcrowded and lacks sufficient seating.

There is a rising concern regarding the preparation of high school graduates upon entering college. Subsequent to the 1983 declaration that America was at risk because of poor educational achievement, billions of dollars were committed to the educational system (Weimer, 1990). Regardless of this effort, minimal results were produced, discouraging public confidence in the instructional capability of American teachers. It has been noted, however, that there are additional factors involved — one such factor currently under assessment in both the political and academic arenas being the influence of class size on student achievement in school and college education.

Taking into consideration the rapidly growing enrollment in many public schools across the United States, administrators are burdened by the concern of increasing class size and the possible reduction of academic standards. Researchers emphasize that the "quantitative product" — financial gains provided by increased enrollment — are overshadowing the "qualitative product" — skillfully educated and well-informed high school graduates. It is, consequently, of immense significance that research studies are carried out to present compelling evidence as to whether or not students, faculty, staff, and possibly the nation as a whole may be suffering negative consequences because of the increase in class size (Weimer, 1990).

It is essential that all students receive individualized attention in some form. Whether it is through participating and sharing ideas, positive reinforcement, a conference, or working one-on-one, none of these are easily accomplished within a large or oversized class. It is essential for successful learning to maintain small class sizes in order to achieve individual student participation.

Cooperative learning is a term representing a variety of interactive group arrangements working toward a common goal. Each student in a cooperative group is individually accountable for the entire group's success. Cooperative learning groups contain students of mixed abilities, different genders, and different cultural backgrounds. For research purposes within this study, a small class consists of fewer than 25 students and a large class consists of more than 30 students. Classes are heterogeneously mixed, containing students of mixed learning abilities, different genders, and varied cultural backgrounds.

This action research project investigates the influence of cooperative learning on class size. Specifically, this study examines how cooperative learning impacts smaller and larger class sizes.

Review of Related Literature

The purpose of class size reduction is to raise student achievement. Classes of varying sizes have presented teachers with the challenge of providing appropriate opportunities for participation while attempting to maintain achievement. A reduction in class size alone does not always lead to high student performance, because the teacher is an essential part of the equation and must practice effective teaching strategies.

There are three factors that determine teacher effectiveness and qualities of a less effective teacher: instructional orientation, management style, and individualized focus. Instructional orientation includes the type of content the teacher emphasizes in lessons and how that content is taught. Management style encompasses how teachers discipline students and organize their lessons. The final factor, individualization, is comprised of how much time and energy teachers spend on individual, one-to-one instruction.

There are many research experiments nationally known for supporting a reduction in class size. The federal government and 20 states within the United States have launched programs to lower the average class size (Zahorik, Halbach, Ehrle, and Molnar, 2003). One well-known research project is Project STAR — Student Teacher Achievement Ratio — a controlled experiment conducted in Tennessee that has been widely recognized and acknowledged by a variety of educational researchers, economists, and statisticians. It is assumed that small class sizes are more effective because there is an improvement in morale and enjoyment of teaching among teachers of small classes (Finn, 2002). Project STAR has been used as a model by groups such as the American Academy of Arts and Sciences and the U.S. Department of Education, as it was conducted in a complex setting represented by public schools.

SAGE, Wisconsin's Student Achievement Guarantee in Education, is another well-recognized research project regarding class size. The SAGE findings are reviewed over a period of five years. Overall, SAGE students scored higher than the comparison group on the reading, language arts, and mathematics subtests. These results showed a 25–30 percent higher level of academic achievement than their counterparts in larger classes, results that were maintained for three years or the length of the program.

Several major questions were raised, notwithstanding an overall pattern of research pointing to the positive effects of class-size reduction on student learning and teaching behaviors. These questions include: (1) How large is the SAGE effect on achievement? (2) Does SAGE reduce the achievement gap between African Americans and white students? (3) Are the benefits of SAGE limited to disadvantaged students? (4) How much does SAGE benefit students with poor attendance? (Smith, Molnar, and Zahorik, 2003).

SAGE affects student achievement. On the basis of the norm groups' predicted performance, the difference translates into a 25–30 percent gain over a year's growth — a significant increase that supports SAGE's claim of improving student achievement. Class size reduction benefits all students, but its effects are especially powerful for African American students. African Americans entering small classes had lower reading and math scores than African Americans entering larger classes in comparison schools. But by the end of the school year, their achievement scores were significantly higher than those of African American students in larger classes. African American students appear to profit more from the SAGE experience than white students when compared with non-SAGE students. The achievement gap between African American and white students widens each year (Smith et al., 2003).

It has been noted that class size initiatives have enjoyed widespread support from parents, teachers, and the general public. People generally believe that smaller class sizes are a good idea, and teachers report experiencing lower levels of stress and job dissatisfaction with smaller classes. This is primarily because teachers are better engaged with each student, and therefore student motivation increases and discipline problems decrease. Parents believe that a teacher's individualized instruction leads to improvements in a child's academic performance. This is apparent because teachers with smaller classes have more time to interact with parents, and their knowledge of their students deepens within those interactions (Gilman and Kiger, 2003). In some districts, the economy is the deciding factor in maintaining the status quo or increasing class size. The cost is often too high for school districts struggling with budget cuts, although the research supports reducing class size.

The decision to reduce class size does not guarantee that qualified teachers and appropriate classrooms will be available. Policymakers face serious challenges presented by America's aging school buildings and the growing shortage of well-qualified teachers. State officials from California to New York have threatened to cut back substantial class size reduction programs in the face of declining state revenues. The National Governors Association estimates that approximately 44 states currently face revenue deficits. In the past, declining revenues for elementary and secondary education have led to fewer fully qualified teachers and larger class sizes. Congress weighed in on the issue of reducing class size in 1998 when it funded a Class Size Reduction program that would reduce class size by hiring 100,000 new and qualified teachers over seven years.

Calculating the cost of a statewide Class Size Reduction program involves considering several ingredients: initial average class size, the size of the reduction target, whether there is a rigid cap or flexibility in the number of students per teacher, teacher salary scales and experience levels, and the cost of constructing or repurposing facilities to create new classrooms. The National Education Association currently supports a class size of 15 students in regular education programs and even smaller in programs for students with special needs.

The American Federation of Teachers cites four necessary steps in order for class size reduction to be effective. The AFT suggests that the most effective classes should be between 15 and 19 students. Particular schools — especially those with low-achieving and low-income students — should be targeted. An adequate supply of qualified teachers and classroom supplies is essential. In addition to increasing student achievement, the AFT recommends that smaller classes improve the classroom atmosphere so that students receive more individualized attention and teachers have the flexibility to use different instructional approaches. With fewer students in a classroom, students are less likely to distract each other, and there will be a lower noise level. Smaller classes also enable teachers to know their students better and offer more extra help, recognizing learning problems, special educational needs, and achievements. According to the American Federation of Teachers, by spending less time on discipline, teachers report spending more time on instruction.

The benefits of class reduction in the early grades last throughout a student's educational career. In 4th, 6th, and 8th grade, students who attended small classes in the early grades were significantly ahead of their regular-class peers in all subjects. By 8th grade, they were still nearly a full year ahead of their peers. The Class Size Matters organization believes that smaller classes are a very cost-effective strategy to lower the number of students who must repeat grades. In the Tennessee STAR study, only 15 inner-city students placed in small classes in early grades were retained through the 9th grade, compared to 44% of those from similar backgrounds in regular-size classes. In high school, students who had been in smaller classes in the early grades had significantly lower dropout rates, higher grades, and higher scores on college entrance exams.

The Class Size Matters organization reiterates what many other studies have also found: with smaller class sizes, behavior problems are significantly reduced. In New York City, a principal in East Harlem reported that disciplinary referrals dropped 60% in her school in the year when smaller classes were instituted. In Burke County, North Carolina, disciplinary problems and interruptions declined by more than 25% after class sizes were reduced. Lower rates of disruption and behavior problems have also been reported in Indiana and California. A survey by Public Agenda shows that among teachers themselves, smaller class sizes are seen as the most effective way to increase the quality of instruction, far above raising salaries or providing more professional development. Smaller classes also lead to improved teacher retention. Reduced class size leads to more parent volunteers in the classroom and more parent involvement overall.

In measuring the potential influences of class size on achievement, the question of how to properly measure academic achievement has been called into question. Kennedy and Siegfried (1997) affirm that the influence of class size on achievement depends upon the measure of achievement used, and the outcomes of a number of reviewed studies suggest that when degrees of knowledge are measured, large-class approaches are as successful as small-class approaches (Alex, 2000). On the other hand, when measures of transfer of knowledge to novel settings, retention of information, critical thinking, problem solving, and attitudinal change or motivation are employed, small-class discussion techniques are preferred. It is evident that employing unsuitable or overly simplistic measures of achievement may lead to invalid conclusions. It has also been noted that grading systems in educational institutions have been affected by grade inflation, resulting in a ceiling effect, with grades within a class being more heavily distributed in the "A" or "B" range than in lower ranges. Research studies relying on class grades as the sole measure of attainment may therefore be problematic due to the resulting restriction of the data range (Alex, 2000).

Eight important studies are examined in this review. Of these eight, six are quantitative in nature, while two are qualitative. All studies utilized a non-experimental design. The class topics covered a number of educational subjects, including one study in statistics, one in electrical and computer engineering, three in economics, and three comprising more than one subject.

Of the eight studies assessed, seven clearly identified class size in measurable terms. Relevant variables such as sample size and sample characteristics were loosely described in these studies and were, on a number of occasions, not addressed at all. Of the six quantitative studies assessed, four defined "achievement" solely in terms of course grade.

The first study was carried out by Gibbs, Lucas, and Simonite (1996). A number of classes across different subjects were evaluated over a span of nearly ten years from 1984–85 to 1993–94. Data collected included the module area, phase, year, course enrollment or class size, and the number and percentage of students achieving grades of "A," "B+," "B," "C," and "F." Class size was described as "small" if it held 30 or fewer learners and "large" if it held more than 70 (as referenced in Alex, 2000). Achievement was defined as the mean module score, expressed as a percentage, for each module, and data were collected on 6,075 learners over the ten-year period. It was hypothesized that a negative association between module enrollment and mean module scores exists, predicting that modules with larger enrollments would have lower mean scores. A Linear Regression Analysis established the first hypothesis but failed to substantiate the second, yielding inconclusive findings (as referenced in Alex, 2000).

Another study, carried out by Esposito, Raimondo, and Gershenberg (1990) at a university, investigated the association between class size in introductory economics courses and student achievement in subsequent intermediate economics theory courses. One hundred forty-six learners who had completed both courses formed the research sample, which included 11.6% Asians, 32.9% women, 3.4% Hispanics, 89.7% full-time students, 9.6% Black students, and 52.7% economics majors. A two-stage procedure was employed to identify learners who had taken the introductory course in either a large (250–350 learners) or a small (25–35 learners) class. Of the 146 learners in this sample, 58.9% had previously taken the introductory course in a large section. The mean score for the large introductory economics course was 2.51 (SD = 1.08) and the mean for the small economics course was 2.66 (SD = 1.00) (as referenced in U.S. Dept of Education, 1999). The outcomes of a linear regression model produced mixed results. Enrollment in a large introductory course did not significantly influence student performance in the intermediate microeconomics course, but did have a negative and statistically significant impact on student achievement in the intermediate macroeconomics course (as referenced in U.S. Dept of Education, 1999).

In 1999, Borden and Burton investigated the influence of class size on learner achievement in introductory college courses across a number of different subjects. Class size was defined as: (a) "small" — 5–30 learners, (b) "average" — 31–90 learners, or (c) "large" — more than 90 learners. Achievement was defined in two ways: (a) the learner's course grade and (b) whether the learner completed the course successfully with a "C" or higher (as referenced in U.S. Dept of Education, 2000). Outcomes of an ANCOVA revealed a generally negative impact, with learners in larger classes not performing as well as learners in smaller sections of the identical course. In addition, the negative influence of large section size was considerably greater among less skilled learners (as referenced in U.S. Dept of Education, 2000).

The influences of class size on learner achievement in statistics were observed by Hancock (1996) in a study in which a teacher of sophomore statistics taught one section each semester in a lecture hall with approximately three times the standard enrollment — around 120 learners. Class size was defined as "standard" sections (approximately 39 learners) and "mega-sections" (approximately 118 learners). Achievement was measured using standard semester examinations. Following three semesters, learner achievement in the "mega-sections" was compared to six standard-sized sections taught by the same teacher using the same textbook, examinations, and grading criteria (as referenced in U.S. Dept of Education, 2000). A Chi-Square Test for Independence produced a test statistic that was not significant, leading to the conclusion that the grade distribution in this statistics class was not influenced by class size (as referenced in U.S. Dept of Education, 2000). A strength of this study was that the researcher controlled for teacher, examination, textbook, and grading methods, though weaknesses included the small sample size, the restricted definition of achievement, and failure to describe sample characteristics.

A pre-test/post-test approach was employed by Kennedy and Siegfried (1997) to study the influence of class size on achievement in economics while holding constant a variety of factors considered to affect learning, such as learner ability and study time. This study used a national economic education database (TUCE III), aggregating data so that each observation corresponded to a class rather than an individual learner, thereby reducing judgment difficulty. Sixty-nine classes ranging in size from 14 to 109 learners were employed, along with one class of 278 (as referenced in U.S. Dept of Education, 2000). Two key conclusions emerged: first, that once proper control was made for aptitude, none of the variables in this dataset over which teachers or department chairs had control affected achievement; and second, that large class size does not reduce the ability to learn the fundamentals of economics. Kennedy and Siegfried caution, however, that these results should not be interpreted as a call to place all learners into one enormous class as a cost-cutting measure, but rather to offer a mix of class sizes and allow learners to choose the class size that best matches their learning style (as referenced in U.S. Dept of Education, 2000).

A study conducted by Hou (1994) on class size and determinants of learning efficiency in economics employed a highly quantitative approach. This study used a procedure initially developed for bias and discrimination assessment to decompose measured learning differences in managerial economics into a component attributable to differences in learner characteristics and a residual component attributable to class size. Surprisingly, results indicated significantly higher achievement in the larger class. Hou warns, however, that because of the size of the model, the strength of some conclusions is uncertain (as referenced in U.S. Dept of Education, 1998).

A qualitative study carried out by Hofmann, Posteraro, and Presz (1994) investigated students' perceptions of what supported their continued enrollment and eventual graduation. The sample consisted of 39 recent graduates of a small liberal arts institution across three academic years. Participants were interviewed by telephone using a semi-structured interview protocol. One of the most significant findings was that graduates cited teaching staff as the most important factor underlying their success in completing their degree. Small class size was another prominent finding, illustrating learners' perceived benefits of teacher interaction and reduced class size (as referenced in U.S. Dept of Education, 1998).

The final study reviewed was carried out by Kopeika (1992), a subjective account of an electrical engineering department's observation on the relationship of smaller class sizes to academic retention, achievement, and perceived quality of teaching as reported by subsequent employers. Starting in 1985–1986, all required departmental courses were divided into two comparatively equal classes of 65–70 learners each in order to reduce class size. Kopeika declared that when class size was reduced, there was an increase of 30 learners in the number graduating that year. Feedback from industry and graduate schools indicated that the quality of graduates' academic preparation had increased. Kopeika concludes that the increased number of graduating learners generates a profit increase that more than offsets the increase in teaching costs associated with smaller class sizes (as referenced in U.S. Dept of Education, 1998).

Additional research has found that class size reduction does not have a substantial influence. Tomlinson investigated trend data from the 1950s to 1986 in America and did not find a reliable association between class size and standardized examination grades, concluding that the current research did not justify a policy to reduce class size given the costs involved and the possible negative influence on the quality of the teaching workforce. Critics noted that this study combined learners from all grade levels simultaneously, that reliance on the student/teacher ratio was an insufficient measure of class size, and that Tomlinson overlooked many intervening issues and social changes that may have obscured the association (as referenced in U.S. Dept of Education, 1999).

Odden assessed the existing research and argued that a system-wide class reduction strategy would generate only modest gains in learner achievement at an excessively high cost. He instead advocated for targeted class reduction plans in conjunction with a series of other interventions, maintaining that his proposals could produce greater benefits at lower costs (as referenced in U.S. Dept of Education, 1999).

An examination of the association between class size and learner achievement for Florida learners using 1993–94 school-level data found no association between smaller classes and learner achievement. The study's authors, however, expressed concern about drawing conclusions from the study given the limitations of the available data (as referenced in U.S. Dept of Education, 1999).

More constructive conclusions were drawn from an examination of Texas education data. Using data from more than 800 districts with over 2.4 million learners, Ferguson established significant associations among teacher quality, class size, and learner achievement. For grades one through seven, using the student/teacher ratio as a measure of class size, Ferguson found that district learner achievement declined as the student/teacher ratio increased beyond 18 to 1. Levels of teacher quality — particularly teacher literacy proficiency and professional experience — were even more strongly associated with higher learner grades (as referenced in U.S. Dept of Education, 1999).

Current Experimental Research on Class Size

Hanushek has frequently reviewed existing research comparing various school resource inputs — including class size reductions — with student outcomes and has concluded that reducing class size should not be expected to produce enhanced learner achievement. His investigations found the associations between various school expenditures and learner achievement to be extraordinarily weak. Other researchers, however, using somewhat different analytical methods on the same data, have challenged Hanushek's conclusions, arguing that the data do show significant influences on learner achievement, including the effect of smaller classes (as referenced in U.S. Dept of Education, 1999).

Data from several more recent initiatives have added significantly to the research evidence regarding class size reduction in the United States at the early primary levels of education. Efforts in Indiana, Tennessee, North Carolina, and Wisconsin have contributed important findings, with the Tennessee projects currently providing the most comprehensive and carefully designed study of class size reduction effects.

Tennessee's Project STAR (Student–Teacher Achievement Ratio) and two related data-collection efforts have made significant contributions to the quality of research evidence regarding class size reduction. STAR was a four-year longitudinal study of kindergarten, first-, second-, and third-grade classrooms in Tennessee, beginning in 1985. STAR compared classes of 13–17 learners with classes of 22–26 learners, both with and without an additional instructional assistant in the larger classes. Contributing teachers did not receive any professional training focused on teaching in reduced-size classes (Charles, 1999).

STAR was notable because it incorporated fundamental features of a controlled research experiment intended to generate reliable evidence regarding the effects of reducing class size. The project integrated 79 schools, more than 300 classrooms, and 7,000 learners, with learners monitored across four years. Teachers and learners were randomly assigned to the three different class types to ensure the study was not biased by who was placed in which type of class. All participating schools implemented at least one of each of the three class types in order to negate possible effects arising from differences in school quality (Charles, 1999).

The evidence from learner testing in STAR demonstrated that learners in smaller classes outperformed learners in larger classes, regardless of whether the larger class teachers had an assistant. Specifically, Project STAR established that: smaller-class learners considerably outperformed larger-class learners on both standardized (Stanford Achievement Tests) and curriculum-based examinations (Basic Skills First); the positive achievement effect of smaller classes for minority learners was initially double that for majority learners and then became approximately equal; a smaller percentage of learners in small-sized classes were retained in-grade; and there were no significant differences in academic achievement for learners in larger classes with or without an extra instructional assistant (Charles, 1999).

Subsequent research presented significant additional evidence on the constructive influences of class size reduction. In 1989, the Lasting Benefits Study began a follow-up to observe whether the effects of smaller class size continued when learners returned to standard-size classes. Research findings to date include: in fourth grade, learners from small-sized classes still outperformed learners from larger classes in all academic disciplines; in fourth grade, learners from small-sized classes showed better classroom behavior than learners from larger classes; and through at least eighth grade, a diminishing but still significant academic achievement advantage for learners from smaller classes continued (Charles, 1999).

In Project Challenge, Tennessee sought to apply the Project STAR findings by implementing smaller class sizes in sixteen of the poorest school districts in the state. Beginning in 1990, the state phased in small-sized classes at the kindergarten through third-grade level in districts with the lowest per capita incomes and highest percentages of students in the subsidized school lunch program. The Project Challenge districts moved from near the bottom of school district achievement in Tennessee to near the middle in both reading and mathematics for second grade. In-grade retention also declined in the Project Challenge districts after smaller classes were implemented (Charles, 1999).

Altogether, the Tennessee studies have been considered landmark research. Finn concluded that this research leaves no doubt that small-sized classes have an advantage over larger classes in school achievement in the early primary grades. Mosteller, Light, and Sachs described this body of research as one of the great experiments in the history of U.S. education (Charles, 1999). Krueger, in an external re-examination of the Project STAR data, reaffirmed the primary finding that learners in small-sized classes performed better on standardized examinations than learners in standard classes, even after accounting for school influences, attrition, re-randomization after kindergarten, non-random transfers, and variability in actual class size (Charles, 1999).

Beginning in 1984, Indiana's Prime Time program allocated funds to support the reduction of class size to 18 in first- and second-grade classrooms, and subsequently in kindergarten and third grade. The outcomes of Prime Time were diverse. An assessment of the program examined achievement grades for first- and second-grade learners, comparing mean class scores in reading and mathematics from ten school districts for tests administered in the year immediately before the program with tests administered in the first year of the project (Charles, 1999). In these districts, the average first-grade class size was reduced from 22 to 19 learners, and in second grade from 21 to 20 learners. Reading scores for first-graders showed the greatest improvement, with smaller gains in mathematics (Charles, 1999).

In Burke County, North Carolina, a recent initiative to reduce class size has also yielded noteworthy data. Beginning in 1990, Burke County piloted and subsequently phased in a class-size reduction program in its public school district. In 1995–96, 1,193 first-graders and 1,125 second-graders participated in the program. The program's objective was to reduce class size to 15 learners in all first-, second-, and third-grade classrooms. The Burke County initiative also incorporated professional development activities covering teaching and assessment (Charles, 1999). Assessment of the program produced the following findings: learners in the small-sized classes outperformed the comparison group in first, second, and third grades on both reading and mathematics achievement tests; and the proportion of classroom time dedicated to instruction in the small-sized classes increased from 80% to 86%, while the percentage of time dedicated to non-instructional activities such as discipline decreased from 20% to 14% (Charles, 1999).

Beginning in 1996–97, Wisconsin launched a class size reduction plan entitled the Student Achievement Guarantee in Education (SAGE) Program. The SAGE program's purpose is to phase in class size reduction in kindergarten through third grade in public school districts serving learners from low-income families. Its goal is to reduce class size in the applicable grade levels to a student/teacher ratio of 15 to 1 or fewer (U.S. Dept of Education, 2000).

In the 1997–98 school year, approximately thirty schools from twenty-one school districts participated in the SAGE program, and fourteen schools in seven districts provided comparison student background and achievement data. SAGE and comparison school learners' academic performance was measured at the start and end of the first-grade year and again at the close of the second-grade year (U.S. Dept of Education, 2000). Learners' scores were compared with those of learners in corresponding comparison schools serving similar student populations, with the following findings: SAGE first-grade learners performed consistently better than comparison learners in reading, mathematics, language arts, and total scores on the comprehensive test of basic skills; the achievement gap narrowed between white and African American learners in the SAGE small-sized classes in first grade, compared to a widening of the gap between white and African American learners in the larger classes of comparison schools; and second-grade SAGE learners' academic achievement remained higher than that of comparison school second-graders, though the difference did not increase significantly (U.S. Dept of Education, 2000).

These conclusions are consistent with the findings of Project STAR; however, two important qualifications apply. First, these are second-year assessment data from an ongoing study, so conclusions may change as the program is fully implemented and further data analyses are completed. Second, the SAGE class size reductions were supplemented by additional program components: participating schools were also required to implement a rigorous academic curriculum, provide before- and after-school activities for learners and community members, and implement professional development and accountability programs (U.S. Dept of Education, 2000).

The higher learner achievement produced by class size reduction may result from the ways in which reducing class size naturally changes the classroom environment. Teachers assigned to smaller classes report an improved classroom atmosphere, the ability to provide more individualized attention, and greater flexibility to use diverse instructional methods (Alex, 2000). One unexpected effect of the Burke County reduced class size program was that teachers found themselves with more classroom space, since they were using the same classrooms with fewer learners. Class size reduction also transforms learning opportunities beyond the classroom, as teachers have a greater proportion of time to dedicate to working with each student's parents (Alex, 2000).

Class size reduction transforms many features of the classroom environment. There are fewer learners to distract each other. Each learner in a reduced-size class receives more attention from the teacher, on average, and more time to speak while others listen. Reduced class size also decreases the noise level in a class. One hypothesis offered to explain the positive influences of class size reduction on learner achievement simply argues that in smaller classes each learner receives a greater share of the learning resources represented by the teacher's instructional time, and therefore learns more (Alex, 2000).

Researchers have also drawn attention to the quality rather than the quantity of teaching in smaller classes. The SAGE assessment study used teacher interviews, classroom observations, and additional data-collection methods to examine what occurs in small classes, suggesting that learners benefit from more individualized attention. Teachers are better acquainted with each of their learners and can track what each is accomplishing on learning tasks. This information enables the teacher to intervene more effectively to assist individual learners in making progress (Alex, 2000).

Researchers have also suggested that smaller classes are more likely to be "welcoming" places where learners develop better relationships with their peers and with the teacher, encouraging learners to become more engaged in classroom learning activities. The smaller the class, the more difficult it is to avoid the positive influence of the classroom learning experience (Alex, 2000).

The rationale for why reduced class size is particularly advantageous in the early grades may originate from the fact that in early grades children are learning how to be learners in classroom environments where the number of people is larger than the number of people in their families — they are learning a new set of customs. This hypothesis is also consistent with the research findings in both Project STAR and SAGE that the largest increase in learner achievement occurs in the first year of a learner's experience in a small-sized class (Alex, 2000).

These various explanations for why class size reduction increases learner achievement are not mutually exclusive, and they do not preclude other potential influences. It may be that class size reduction improves learner achievement because there is more time for learning, more individualized attention, and a better introduction to classroom routine. If teachers were trained to adapt their teaching in ways suited to smaller classes — such as initiating different kinds of learner activities possible only in smaller classes — this might produce an additional benefit (Alex, 2000).

The emphasis on class size reduction in the early grades also suggests that small-sized classes represent a preventive rather than a remedial approach. If smaller classes help learners get off to a good start in adapting to the classroom environment and becoming engaged in learning, then learners avoid the more difficult educational trajectory of falling behind and catching up with their classmates (Alex, 2000).