Blueprints for School Success

It is sometimes assumed that the quality of educational facilities has no impact on academic achievement. Mayor John Norquist of Milwaukee, for example, was quoted this spring during debate over a Milwaukee Public Schools construction referendum as saying that he believed there was “no clear relationship between how well kids do in school and the facilities they occupy…none of this [school building renovation and expansion] will necessarily improve education.”

School referendum issues have been defeated across Wisconsin and in many other states. A partial interpretation is that people do not believe that improving school buildings will lead to improvements in performance.

We beg to differ. There is significant evidence that a number of design characteristics of schools make a difference to educational performance. An excellent review of the role of the physical environment of schools by Carol Weinstein in the 1979 Review of Educational Research concluded, “There is considerable evidence that the classroom environment can affect nonachievement behaviors and attitudes” (her emphasis, refers to such student and teacher attitudes and behavior, as increased social interaction or decreased aggression). But newer evidence since 1979 shows that school and classroom size make substantial differences in learning achievement, while location and the provision of secluded study spaces contribute to other educational outcomes.

School Size

Between the early 1960s and 1980, 344 articles were published pertaining to the effects of school size on academic achievement and other achievement-related variables. In the now-classic Big School, Small School, environmental psychologists Roger Barker and Paul Gump conducted a study of a sample of large (over 2,000 students) and very small (100-150 students) high schools in Kansas. They concluded that small schools offered students greater opportunities to participate in extracurricular activities and to exercise leadership roles. In particular, participation in school activities, student satisfaction, number of classes taken, community employment, and participation in social organizations were all superior in small schools. A review of some of the subsequent studies appeared in an article in the 1980 Journal of Youth and Adolescence. Small schools (those on the order of 500 students) also have lower incidence of crime and less serious student misconduct. Larger schools discourage a sense of responsibility and meaningful participation, particularly among students who have academic difficulty and come from lower socio-economic backgrounds.

But how about academic achievement? William Fowler reported in a 1992 U.S. Department of Education paper that the academic effects of school size at the elementary school level are conclusive. Controlling for socio-economic differences, studies of small schools (100-200 students) versus large elementary schools (1,500-4,000 students) have found: (1) negative relationships exist between math and verbal ability tests and elementary school size; (2) smaller elementary schools particularly benefited African-American students’ achievement in Philadelphia; and (3) negative relationships between school size and student performance are most prevalent in urban schools, based on 4,337 K-6 schools in California. The impact of school size on academic performance is clear — smaller schools lead to a range of beneficial effects, with the evidence suggesting even more pronounced effects of small elementary schools for African-American students and urban locations.

Class Size

Many other studies over the past 10 years have looked at classroom size and class density and their impacts on children’s performance. High density conditions have been found to lead to increased aggression, decreased social interaction, and non-involvement. Smaller class sizes lead to better scores on learning achievement.

Perhaps the most famous study is Project STAR (Student/Teacher Achievement Ratio Project), a $12 million, four-year random-ized longitudinal experiment in Tennessee, which followed some 6,500 children from kindergarten through third grade in 79 different schools from 42 state-wide school districts (see Rethinking Schools, Jan./Feb. 1992). Across the board, children in small classes (13-17 per room) outperformed those in traditionally sized classes (22-25 per room), as measured by tests such as the Stanford Achievement Test. In the early grades, children in small classes outperformed children from regular class sizes in all subjects, especially on various reading and mathematics achievement measures (average improvements of 15%). While this was true across all locations — rural, urban, suburban, and inner city — smaller classes were especially helpful for children in inner-city schools. And while the improvement was immediately clear in small kindergarten rooms, the benefits increased in first grade and remained stable over second and third grades.

A follow-up study using the same schools, pupils, and tests, called the Lasting Benefits Study (LBS), has been looking to see if there are any long-term effects of small class size. What happens for students who benefited from small class sizes during the K-3 years when they return to larger classrooms (25:1) in grade 4? The LBS analysis is yielding clear and consistent results across 4,500 of the students tracked from the earlier STAR study. Students previously in small-size classrooms demonstrated statistically significant advantages two years later over students previously in regular size classrooms and even those with an extra teacher’s aide. Performance gains ranged from 11-34%. The results were consistent across rural, urban, suburban, and inner-city schools. The greatest achievement advances appear to be for inner-city and suburban classes, and for minority students.

The Project STAR results are buttressed by studies from other states and from Canada. In an initiative called Prime Time, Indiana reduced some K-2 classes from an average of 23 students for each teacher in 1981 to 14 to 18 per teacher in 1983. The results were impressive, with 14% more students in the small classes scoring above average on standardized reading and mathematics achievement tests than students from larger classes. Finally, as reported in the 1992 Phi Delta Kappan, elementary students in the small classes (like 15:1 classrooms) participate more than those in large classes, holding all other things equal. Student participation is essential for learning to occur and is also linked to staying in school longer.

A number of studies were reviewed in 1982 by Glass and associates in School Class Size: Research and Policy. The authors concluded, without qualification, that “reduced class size can be expected to produce increased academic achievement” (p. iv). The analysis indicated that reducing class size from 30 to 20 can yield a gain of 6 percentage points on achievement scores, whereas a reduction from 20 to 10 students per classroom yields another 13 percentage points in achievement. Reductions in class size begin to make substantial differences in learning achievement at around 15 students to a class.

Changes in teaching strategies are crucial, too — active learning and participation mean that teachers can’t just lecture to a class of 15. While changing class size affects academic performance, more profound effects would be expected by changing both class size and teaching methods.

Location and Noise

Other physical planning and design variables indirectly affect learning. For instance, the location of new schools is now known to be critical. A series of studies between 1980 and 1986, reviewed by our colleague Gary Evans in New Directions in Health Psychology Assessment, concluded that there are significant increases in systolic and diastolic blood pressure in children in schools near noisy urban streets, and abnormally high blood pressure in children residing around airports. This can be expected to be the case also for suburban expressways and noise hazards near rural schools. Exposure to traffic noise at elementary schools also has been associated with deficits in mental concentration, leading to more errors on difficult tasks, and a greater chance of giving up on tasks before the time allocated has expired.

Furthermore, as found by Sheldon Cohen and his colleagues in Los Angeles, elevated blood pressure does not habituate or decline with continued noise exposure over time — children don’t get used to noise.

While blood pressure, concentration, and task persistence are not academic achievement outcomes, they indirectly affect educational outcomes. The appropriate location of new schools and their proper architectural design can alleviate these noise-related problems.

Study Spaces

Secluded study spaces within classrooms are also important to child development, and have been found to be related to performance. Creating small learning centers within classrooms reduces classroom visual and auditory interruptions, makes learning materials more accessible, increases privacy, and leads to more questions asked by students. A study in the 1982 Elementary School Journal reported that structured reading areas significantly increase literature use by students. Some of our own research, reported in the 1986 Journal of Environmental Psychology, has shown that attention span is longer for pre-school children in architecturally well-defined activity settings within classrooms than it is in more traditional classrooms.

We mention these four areas of research as there is a widespread — but quite mistaken — impression that there is no relationship between how well kids do in school and the school buildings they occupy. To the contrary, there now is substantial scientific evidence reported in the past 15 years that the location of schools, school size, class size, and the provision of secluded study spaces within classrooms all lead to substantial improvements in the learning environment and to significant increases in school performance.

The evidence strongly suggests that educational planners, school boards, and the public look atthose aspects of school facilitie that directly affect student performance.

Barker, R. and P.V. Gump. Big School, Small School. Palo Alto, CA: Stanford University Press, 1964.

Evans, G.W., W. Kliewer, and J. Martin. The role of the physical environment in the health and well-being of children. In H.E. Schroeder (Ed.), New Directions in Health Psychology Assessment. New York: Hemisphere, 1991.

Finn, J.D. and C.M. Achilles. Answers and questions about class size: A statewide experiment. American Educational Research Journal, 1990, Vol. 27, pp. 557-577.

Fowler, W.J., Jr. What do we know about school size? What should we know? Paper presented to the American Educational Research Association Annual Meeting, San Francisco, April 1992. Available from the Office of Educational Research and Improvement, National Center for Educational Statistics, U.S. Department of Education, Washington, DC.

Gabarino, J. Some thoughts on school size and its effects on adolescent development. Journal of Youth and Adolescence, 1980, Vol. 9, pp. 19-31. Glass, G.V., L.S. Cahen, M.L. Smith, and N.N. Filby. School Class Size: Research and Policy. Beverly Hills, CA: Sage, 1982.

Miner, B. Students learn best in small classes: Tennessee study follows 6,500 children for four years. Rethinking Schools, January/February 1992, Vol. 6, No. 2, p. 15.

Morrow, L. and C.S. Weinstein. Increasing children’s use of literature through program and physical design changes. Elementary School Journal, 1982, Vol. 83, pp. 131-137.

Pate-Bain, H., C.M. Achilles, J. Boyd-Zaharias, and B. McKenna. Class size does make a difference. Phi Delta Kappan, November 1992, pp. 253-256.

Weinstein, C.S. The physical environment of the school: A review of the research. Review of Educational Research, 1979, Vol. 49, pp. 577-610.