Magnet School Conlerence, December 4, 2014 1. lnterdistrict Magnet Schools (IMS) Operated by local school districts. regional education service centers and collegesluniversities Uses a combination of state and local funds An IMS generally enrolls students from multiple districts and is expected to meet statutory racial isolation target of at least 20% white students 2. Public Charter Schools (PCS) Public non-sectarian schools that are operated independently from the local or regional board of education by charter management organizations Uses a combination of federal, state, and private funds A PCS enrolls students from one or more districts 3. Open Choice Offers public school students mostly in urban cities the opportunity to attend public schools in suburban towns at no cost to the student's family - Utilizes a combination of state and local funds choice options include CT Technical High Schools and Vocational Agriculture programsSigni?cant outcome Control means Equivalence of treatment and control groups at baseline Pre: Year I Grade 1 Past: Year] Grade 3 3 Pretest Pasttest Means for Treatment and Controls Based on Multiple Non-equivalent Matched Groups Nan-Uram. Upper Limit Gap Blue lines represent meals of orange piper-formalin for 30 matched control groups, Treatment group Signi?cant treatment effects, in red, compared to grand of empirical control means at Outcome Urbm. tower Umh Gap Selection Bias 0! Treatment Program and Matde Over Urban Average Year I Baseline? "Foe?titan "out?t? 15 75% Rearment pretest matte? means 50% Fnhety Skni?unt? outperne due to large sample sizes mrpe ler'ror) (antral group pretest games! means 35% Nm-fnurvol'ence of mean: at baseline favors treatment group - We CAN NOT make causal attributions - We CAN test whether quasi experimental groups exceed untreated matched samples We CAN follow-up to identify programmatic or pedagogical differences that may be validated 18