But a new study by two Johns Hopkins University researchers says this popular middle school reform should be approached with far less enthusiasm and far more caution. The higher student achievement in math and reading associated with K-8 schools has more to do with student demographics, grade size and the school transition issue than with the K-8 grade structure, they say.

"Much of the old K-8 advantage clearly resides in the different student populations that are served by old K-8 schools and middle schools," the researchers conclude.

Their conclusion is based on a five-year longitudinal study of 40,883 eighth-grade students in the Philadelphia City School District which is implementing a policy of converting its middle schools into K-8 schools to improve math and reading achievement. The study was recently published in the American Journal of Education.

While students in older K-8 schools in the Philadelphia district did indeed perform better in math and reading than their middle school counterparts, students in newer K-8 schools with high-poverty, high-minority populations show only some advantage in reading and none in math, according to this rigorous, large-scale study.

January 2008 issue Less than ideal interventions for ADHD still productive Reading intervention targets motivation to read What Wal-Mart can teach us about data-driven decision-making Rapid assessment most cost- effective way to raise achievement View recent issues Subscription information

"A K-8 conversion policy alone does not represent a 'silver bullet' reform for closing the achievement gap and improving student achievement," write researchers Vaughan Byrnes and Allen Ruby. "Administrators must ask themselves if such a massive reform is truly worth the resources given the likely impacts. They must also compare it to other possible reforms and decide if they are getting with K-8 conversions the best possible 'bang for their buck' in terms of reform finances."

Demographics important
The neighborhoods and communities of K-8 schools also may play a role in the observed halo effect of these schools.

"We would also believe that the stronger community and relationships that exist in old K-8 schools, which foster student achievement and social outcomes, are not entirely the result of their smallness and continuity into the middle grades but also due to the demographics of their student populations and parents, the community members themselves," the researchers say.

"So long as the new K-8 schools consist of the same high-minority and high-poverty student populations as the middle schools, it seems unlikely that they will develop the same sizable achievement advantage seen in the old K-8 schools."

Sign up to receive free research updates.

Of the 95 schools in the study, 14 were new K-8 schools. The researchers caution that the cohorts in the new K-8 schools were small compared to the number of students in old K-8 schools and that half of those 14 schools only added an eighth grade in the last year of the study, but they say their results nevertheless question the benefits of K-8 when demographic characteristics are not taken into account.

Students in K-8 schools have not only been found to have higher achievement in math and reading, but also higher rates of attendance and better emotional outcomes such as self-esteem, leadership, and attitudes toward school. Parents often praise the greater sense of community in K-8 schools, and some studies have noted stronger relationships among students, teachers and parents.

This K-8 advantage is attributed to several  factors:
• differences in the average size of K-8 schools versus middle schools;
• differences in teacher populations;
• the transition students must make from elementary school to middle school; and
• differences between the student populations of K-8 and middle schools.

Some research has suggested that as the "top dogs" in their schools, early adolescents may have greater feelings of confidence, maturity and leadership.

"If student demographics are the main reason for the different academic performances of the two school types," the researchers say, "then converting middle schools in K-8s may not lead to a significant improvement in student achievement if the student population remains unchanged."

Middle schools in general serve student populations with higher rates of poverty and larger proportions of minority students, the study observes. Middle school staffs have lower rates of teacher retention, fewer experienced teachers and lower rates of certification, not only because most teachers are trained to teach at either the elementary or high school level, but teachers with more experience and seniority are more likely to transfer out of middle schools.

The large size of middle schools has been found in many studies to be detrimental to academic achievement, attendance and social engagement, the authors report.

"With a touch of irony, smaller size may also enable K-8 schools to more effectively implement the very set of 'best practices' that were originally thought to be an advantage of middle schools," they write, "and the greater use of these practices may also be a reason why K-8 schools tend to perform better."

Rigorous large-scale study
Given how widely K-8 conversion has been adopted across the United States, the researchers say they wanted to examine the K-8 advantage with more rigorous and appropriate methods of statistical analysis, a substantially larger sample size, and a more diverse set of statistics controls.

Student performance was measured by results on the Pennsylvania State System of Assessment (PSSA) for school years 1999-2000 to 2003-2004. This is the high-stakes test used by the state to evaluate schools.

The researchers controlled for prior achievement with fifth grade PSSA scores and income of families with percentages of students on the free-reduced lunch program. For teacher characteristics, the researchers used measures for teacher absentee rates, percent of certified teachers at the school, average experience of teachers at the school and the student/teacher ratio of the school. Other variables were grade size and a proxy measure for student mobility.

The researchers used multilevel modeling, which is similar to regression modeling, but takes into account that students nested within the same schools will have shared similar experiences.

For administrators who are tempted by the academic and psychosocial lift of K-8 schools but leery of such a massive reform, an option for increased student achievement is to cultivate the set of best practices that were originally thought to be one of the unique advantages of middle schools, the authors write. In Philadelphia, many of the highest-performing middle schools, with achievement levels comparable to K-8 schools, are using outside partner programs designed to implement small learning communities, professional development, cooperative learning and other instructional strategies.

On a final note that is both sobering and hopeful for policymakers and administrators, the researchers point out that much of the variation in achievement in their pertains to the students themselves.

"It is likely that a good deal of that unexplained variation resides in factors pertaining to a student's parents and their home environment, factors that schools and school administrators cannot address on a schoolwide level," they conclude.

"Comparing Achievement between K-8 and Middle Schools: A Large-Scale Empirical Study," by Vaughan Byrnes and Allen Ruby, American Journal of Education, November 2007, Volume 114, Number 1, pp. 101-135.

Published in ERN January 2008 Volume 21 Number 1