Search

This report presents student-level achievement results for the four charter schools that began operation in Tennessee during the 2003-04 academic year. To conduct a rigorous and valid analysis of student achievement outcomes at these schools, we employed a matched program-control design at the student level, whereby each charter school student was paired to a comparable "control" student attending the same or similar school during 2002-03. The four charter schools were: Circles of Success in Learning Academy (COSLA), Memphis Academy of Health Sciences (MAHS), Memphis Academy of Science and Engineering (MASE), and the Smithson-Craighead Academy (SCA). The first three schools are located in Memphis and the fourth in Nashville. Participants in the study were 14 second graders from COSLA, 70 sixth graders from MAHS, 134 seventh graders from MASE, and 9 fourth graders from SCA. All had scores on the spring, 2003 (pre-implementation or "pretest") and the spring, 2004 (posttest) Tennessee Comprehensive Assessment Program/Achievement Test (TCAP/AT) in Reading, Language Arts, and Math. The analyses of 2004 TCAP/AT scores for the three Memphis cohort schools showed directional advantages for the charter school students over their matched control counterparts on all three subtests. However, only the effect for MASE in Math was statistically significant. The median effect sizes were +0.15 for test scores unadjusted for pretest differences and +0.12 for adjusted scores, indicating small to moderate effects. When the results for all three schools were combined, significant advantages for the charter school students occurred on both Reading (ES = +0.19) and Math (ES = +0.19), with a directional advantage in Language Arts (ES = +0.12). Attendance was also significantly higher at two of the Memphis schools and directionally higher at the third. For SCA, however, the comparisons to control students yielded comparable to negative results. Because only 9 student matches were available for SCA, these results need to be viewed very cautiously relative to those from the three schools and much larger sample sizes in the Memphis cohort. Overall, the first-year achievement outcomes seem highly suggestive and impressive in view of the: (a) early (first-year) implementation of the charter school programs; (b) charter school students' transition to new schools, a factor normally associated with lower achievement; and (c) comparable magnitude of the effect sizes in Memphis to those obtained in prior research on established comprehensive school reform (CSR) models. The present study will be replicated over the next several years to examine first-year results for newly established charter schools and longitudinal outcomes for the entire population of Tennessee charter schools. [This report also produced by Office of Research and Evaluation, Memphis City Schools.] (Contains 5 tables and 1 footnote.)

The present study examined outcomes on the Tennessee Comprehensive Assessment Program/Achievement Test (TCAP/AT) for the KIPP:DIAMOND Academy (KIPP:DA), which is in its second year of operation. Importantly, as in Year 1, a rigorous quasi-experimental research design was employed, in which each KIPP:DA student was individually matched to a highly similar counterpart who attended one of several geographically proximate neighborhood schools. The 2004 TCAP/AT was taken by three subsamples of KIPP:DA students: (a) fifth graders who had completed their first year at the school, (b) sixth graders who had also completed their first year, and (c) sixth graders who had completed their second year at the school. "Pretest" analyses revealed that KIPP:DA and control students scored almost identically on the TCAP/AT in the year prior to the former group's enrollment in KIPP:DA. Comparisons between the three KIPP:DA groups and their matched control counterparts on each of the TCAP/AT subtests (Reading, Language Arts, Mathematics), directionally favored KIPP:DA in 7 out of the 9 analyses (median ES = +0.14) unadjusted for pretest score, and in 8 out of 9 analyses (median ES = +0.16) adjusted for pretest scores. Inferential analyses showed the KIPP:DA advantages to be statistically significant for only the fifth-grade and sixth-grade longitudinal cohort subsamples. Supplementary descriptive analyses of the criterion-referenced (CRT) portion of TCAP/AT further revealed that, while not statistically significant, KIPP:DA students in all three subsamples were more frequently represented than the matched control students at the Proficient and Advanced levels, and less frequently at the Below Proficient level, in both Reading/Language Arts and Mathematics. These results are clearly suggestive of positive KIPP:DA effects in Year 2, especially in view of the doubling of school size and special unanticipated challenges faced during the year. Pretest School Attendance for Control Group Mates tabular data is appended. (Contains 4 tables.)