Are Magnet Schools Attracting All Families Equally ? Naralys Estevez ’06 Cities, Suburbs, and Schools research project at Trinity College, Hartford CT.

Are Magnet Schools Attracting All Families Equally?

Naralys Estevez ’06Cities, Suburbs, and Schools research project

at Trinity College, Hartford CT

http://www.trincoll.edu/depts/educ/css

July 18, 2005

http://www.trincoll.edu/depts/educ/css

Sheff and School Segregation

1996 Sheff vs. O’Neill ruled that that the state must desegregate schools in metropolitan Hartford to address educational inequalities

Sheff and School Segregation

1996 Sheff vs. O’Neill ruled that that the state must desegregate schools in metropolitan Hartford to address educational inequalities

Finding a remedy remains a challenge to this day

Magnet Schools as a Solution?

2003 Sheff settlement focused on inter-district magnet schools as a key desegregation remedy

Popular voluntary approach, with 19 magnet schools for 2005-06

Magnet Schools as a Solution?

2003 Sheff settlement focused on inter-district magnet schools as a key desegregation remedy

Popular voluntary approach, with 19 magnet schools for 2005-06

Not Pictured: MLC (Bloomfield), EHG & CT IB Acad (East Hartford), GPA (Manchester)

Learning Corridor

Magnet School Controversy

Magnet schools are designed to attract a special mix of students



But critics charge that magnets “cream” the best students from neighborhood schools

- race

- socio-economic status

- achievement



But critics charge that magnets “cream” the best students from neighborhood schools

- race

- socio-economic status

- achievement

Although designed to equalize society, magnet schools may create an unexpected second tier of inequality

Research Question

Are magnet schools attracting all families equally?

Research Question


Research Question


My study uses Geographic Information System (GIS) to conduct spatial analysis of the “creaming” controversy by using smaller geographic units of analysis.

Research Question


My study uses Geographic Information System (GIS) to conduct spatial analysis of the “creaming” controversy by using smaller geographic units of analysis.

GIS training funded by National Institute for Technology & Liberal Arts Education (NITLE) and Trinity College Academic Computing

Four Levels of Investigating Creaming1. Individual magnet school


2. Categories of inequalityRaceSocio-economic statusAchievement



3. Geographic unit of analysisSchool DistrictNeighborhood




4. Stages of magnet admissions processApplication (Which students apply?)Acceptance (Which applicants are accepted?)Enrollment (Which students actually enroll?)

Sample Statistical Test of Creaming

Sample Statistical Test of Creaming1. Individual magnet school

Montessori Magnet School (2000-2005)2. Categories of inequality

RaceSocio-economic statusAchievement




District Enroll (Bloomfield 5yr)

Total White Students

Observed Percent

12,560 630 5%


District Enroll (Bloomfield 5yr) MMS Applicants (Blmfd 5yr)

Total White Students Total White Applicants

Observed Percent Observ Pct

12,560 630 5% 105 8 7%




Observed Percent Observ Pct Expected

12,560 630 5% 105 8 7% 5.25





12,560 630 5% 105 8 7% 5.25

Calculate Chi-Square goodness of fit test to determine statistical significance

Chi-Square = 1.516 P>.05





12,560 630 5% 105 8 7% 5.25

Calculate Chi-Square goodness of fit test to determine statistical significance

Chi-Square = 1.516 P>.05

Difference in percentage of White Bloomfield MMS applicants, compared to Whites in the district, is not statistically significant


White MMS Applicants (2000-2005)Districts Observed Expected SignificanceBloomfield 8 5.2New Britain 18 15.9

East Hartford 21 39.1 * W lessGlastonbury 18 25.9 * W lessManchester 25 42.2 * W lessWest Hartford 46 85.4 * W lessWethersfield 29 43.4 * W lessWindsor 27 56.3 * W less

Hartford 84 49.7 * W More


White MMS Applicants (2000-2005)Districts Observed Expected SignificanceBloomfield 8 5.2New Britain 18 15.9

East Hartford 21 39.1 * W lessGlastonbury 18 25.9 * W lessManchester 25 42.2 * W lessWest Hartford 46 85.4 * W lessWethersfield 29 43.4 * W lessWindsor 27 56.3 * W less

Hartford 84 49.7 * W More

Non-White Applicant Results are reciprocal (e.g. when Whites are less likely to apply, then Non-Whites are more likely to apply)

Purple = Significant

White Dots = White Students

< less likely to apply> more likely to apply

Significant Differences in MMS Applications and School District Enrollment, by Race, 2000-2005


Future Statistical Tests

Future Statistical Tests1. Individual magnet school

2. Categories of inequality

RaceSocio-economic statusAchievement





Census Block Group = 1500 residents

With Census 2000 demographic data for race, SES, educational attainment

Requires magnet applicant addresses


Limitations and Factors to Consider


Limitations and Factors to ConsiderCensus data – only available every decade

Application form data – some magnets do not ask raceApplication process

– depends upon marketing (varies by magnet management and individual school)

- student eligibility lottery requirements (eg GHAA audition)

- type of lottery system (open to all vs. selected districts only)Achievement analysis- Requires compiled CMT scores

Are Magnet Schools Attracting All Families Equally ? Naralys Estevez ’06 Cities, Suburbs, and Schools research project at Trinity College, Hartford CT.

Documents

district magnet schools

individual magnet school

neighborhood schools

schools research project

day slide

tier of inequality slide

school segregation

creaming controversy