This study analyzed the effect of school/grade configuration for regular and special education students in K-8 schools and middle schools. The analysis looked at the effect of grade configuration on two outcomes: student achievement and student attendance. The study followed a cohort of fifth graders (regular and special education) through sixth, seventh and eighth grades, in either a K-8 or a middle school. The analysis used multilevel modeling to account for student demographics, prior achievement and school variation in achievement.

Many factors including demographic features such as race, free and reduced meals (FARMS), and prior achievement can affect middle grade performance. Because students and schools differ in terms of some of these variables, multilevel evaluation was necessary to partial out their effects to determine the effect of school grade configuration on student outcomes.

Results indicated that students performed slightly better academically, as measured by the Maryland State Assessments, in the K-8 schools but only at a statistically significant level for regular education 6^th grade math and special education 6^th grade reading. Attendance results indicated that students in 6^th grade regular and special education performed statistically significantly better in K-8 schools but the magnitude was small.

Four major outcomes of this study had implications for policy and practice: (a) The results of this study do not support reconfiguration as a strategy for better outcomes; (b) Student performance in fifth grade was the best predictor of student success in the middle grades; (c) School performance had a significant effect on student performance, regardless of school configuration or placement in regular or special education; (d) These results suggest the importance of including special education students in high-performing schools, where they will have the potential to perform at higher levels.