The most important aspect of a youth's development is the social map that he or she draws of the world, for this map is both the product and the cause of experience (Garbarino, 1997).

Introduction

Since 1940 the size of the average US school district has risen from 217 to 2,637, and the size of the average school has risen from 127 to 653 (Walberg, 1994). In many cases these size increases occur incrementally as schools fill and are repeatedly replaced by new ones with slightly larger capacity. In other cases the increases are the result of school consolidation, often a devastating experience in those cases where "the local school may be a focal point of the community's identity" (Ornstein, 1993). Therefore it is rare that a community has the opportunity to select between large and small school models. Such a decision should be made with the best possible understanding of the educational and social implications for the children whose lives will be affected.

There is remarkable consistency among the research studies that have been reported on school size; smaller is better. To understand these findings one must appreciate the pressing need of children, especially the younger ones, for structure, social stability, and community support. It appears that smaller schools strengthen interpersonal relationships and sense of community. Smaller schools are also associated with stronger parental committment and have higher rates of parental involvement. Here again, it is consistently reported that this improves educational efficacy, no matter what its form (Henderson, 1987). Noting that many rural schools tended to be smaller, researchers asked whether it was the ruralness or the size that was important. Once again it was determined that it is the size that counts (Lindsay, 1982).

It is not hard to see how alienation and loss of identification occurs in larger schools. Consider a class play organized in a smaller school. Mary may have marginal acting talents but is welcomed to participate, simply because there are few students in her class and she is needed. In a larger school, only the best participate, and Mary is superfluous. Since Mary has no stake in the play, neither does her family, with the result that the entire family is distanced from the school. Garbarino and Asp (1981) write,

We must always remember that the school is usually the first social system that children encounter. In fact, it may well be the only social system, other than the family, that they know well at all. We must therefore be very attentive to its structure and behavior as a context for socialization.

Garbarino argues that contemporary schools are large because the focus on "cognitive academic curricula" has caused decision makers to ignore social dynamics. Large schools contribute to depersonalization, negativism, alienation, and ultimately truancy and drop-outs. "School size affects student participation and satisfaction independent of the effects of SES and academic ability" (Lindsay, 1982).

The dilemma for decision makers looking only at cost figures and tax rates is that the numbers do not take into account the long term costs of making the wrong decision, real costs that are hard to estimate. They arise from crime, vandalism, joblessness, and related community problems attributable to asocial behaviors acquired through an inadequate social context in the childhood and adolescent years.

One of the few contrary results comes from an Illinois study over a ten-year period that concluded that the lowest high school achievement (on three standardized tests) came from schools with fewer than 495 students. Highest achievement was found in high schools with 495 to 1200 students (Ornstein, 1993). On the other hand, for elementary school students the situation was reversed, and smaller schools produced superior achievement.

Another result that emerges is that the benefits of small schools are greater for minority students or those of low socioeconomic status (SES). Thus larger schools are discriminatory against such students and those who are academically marginal (Barker and Gump, 1964; Fowler and Walberg, 1991).

Factors Affected by School Size

The following section is taken largely from Cotton's extensive metaanalysis of the research literature (Cotton, 1996). The factors considered are hers, and the details and literature citations are readily available on the WWW.

Quality of the Curriculum
Some educators argue that large schools are advantageous because they can offer a broader curriculum. Upon examination, researchers have found that doubling enrollment produces only a 17% increase in the variety of offerings, and that only 5-12% of the students in larger schools avail themselves of these extra courses. Similarly, five of six studies report that smaller schools were just as effective as larger ones in preparing students for college entrance.

Cost-Effectiveness
Researchers find that up to a point, there are economies of scale and per-student costs decline with increasing size. Then as enrollment continues to increase, extra staff members are required to manage large numbers of students, and costs increase again. The school size that maximizes cost-effectiveness is variable and highly dependent upon the particulars of the community and the school system.

Academic Achievement
No research finds large schools superior to smaller schools in academic achievement, and smaller schools show positive effects on the achievement of ethnic minority students and those of low socioeconomic status.

Student Attitudes
Research on student attitudes strongly favors small schools over large ones, with minority and low-SES students again showing the most positive benefits from a small-school environment.

Social Behavior
Small schools have lower incidences of negative social behavior than large schools, wich minority and low-SES students showing the most positive effects from smaller schools.

Extracurricular Participation
Students in small schools are involved in a greater variety of activities than those in larger schools. Hamilton (1983) writes,

Students in the large schools were more polarized, with a group of active participants at one end of the continuum and a large group of students who did not participate in an extracurricular activities at the other. In the small schools there were few students who did not participate in anything.

Attendance
Attendance statistics again favor small schools over larger ones.

Dropouts
Nine of the ten reports reviewed by Cotton report lower dropout rates for smaller schools.

Belongingness/Alienation
Researchers report a greater sense of community and belonging among students of smaller schools.

Self-Concept
Personal and academic self-regard are stronger in smaller schools. As Rutter (1998) writes,

Evidence of increases in social bonding to teachers and school, self-esteem, academic self-concept, locus of control and sociocentric reasoning suggest that [small alternative] programs can respond constructively to students' underlying needs.

Interpersonal Relations
Researchers report that interpersonal relations among students and teachers at smaller schools were more positive at smaller schools.

Teacher Attitudes
Cotton reports fewer studies of teachers and administration in schools of different sizes, but those that were found favor smaller schools. As Gottfredson (1985) writes,

Large schools appear to promote negative teacher perceptions of school administration and low staff morale.

References

Comments to:
Roger W. Ehrich
Department of Computer Science
Virginia Tech