Milwaukee Independent Charter Schools Study: Report on One Year of Student Growth. SCDP Milwaukee Evaluation. Report # 21--Version 1.1

The general purpose of this evaluation is to assess the effectiveness of independent charter schools in promoting two desirable student outcomes: student achievement growth and educational attainment. Independent charter schools are authorized by non-district entities and are considered "independent" because they are not a part of the Milwaukee Public School District. The authors will estimate achievement growth of independent charter school students in grades 3-8 over four years in reading and math on the Wisconsin Knowledge and Concepts Examination (WKCE). Similarly, in later reports the authors will track student attainment, specifically whether upper-grade cohorts in their evaluation graduate from high school. Case studies of independent charter schools will help them to identify best practices in these schools and will also be addressed in future reports. This report provides findings comparing the first year of achievement growth (2006 to 2007) of students attending independent charters to the achievement growth of a group of matched comparison students attending Milwaukee Public Schools. The authors' next report, to be released in spring 2011, will examine two- and three-year achievement growth. These reports draw upon a panel of all 2,295 students attending 10 of the 14 independent charter schools in grades 3-8 in 2006-07 with test scores for that year. Using regression models that produce the most precise estimates of 2007 achievement, their comparisons of students in their sample of independent Milwaukee charters to matched MPS students exhibit few significant effects of attending a charter school on achievement growth in either math or reading. The exception is in one of their three models for mathematics gains. When they control for prior achievement, and not for student characteristics or switching schools, students in charter schools gain approximately 0.105 standard deviations more in math achievement than students in MPS. Further analysis reveals that the positive impact of independent charter schools on average in math is concentrated primarily at the lower end of the achievement distribution; these schools were estimated to improve the math achievement of students at the 25th percentile of the achievement distribution by 0.109 standard deviations. There are no differences in any models in reading. There are differences, however, when they disaggregate the charter impacts by charter school type. Conversion independent charters, schools which converted from private schools, hold an advantage in math and reading achievement. Prior to controlling for both student characteristics and if students switched schools, students in conversion charters make 0.170 standard deviations greater gains in math achievement compared to similar students in MPS schools. Once controlling for student characteristics and school switching, the effect is reduced to 0.114 standard deviations. Similarly, in reading, students in conversion charters make 0.124 standard deviations more gains than MPS students without controlling for student characteristics and switching schools. By adding these factors the effect is reduced to 0.054 standard deviations. At the same time, students in non-conversion, independent charter schools, schools which began as new charter schools or startups, achieve gains that are no different from their counterparts in MPS. Descriptive Statistics for Variables in the Achievement Model is appended. (Contains 16 tables, 4 figures and 10 footnotes.)