Start date: Spring 2016
End date: August 2016
Location: Philadelphia, PA
Partner: Mayor’s Office of Education
What’s a Community School?
When I asked Deputy Director of Community Schools Holly Gonzales that question during our first meeting, her answer surprised me. I knew that recently elected Mayor Jim Kenney had run on a platform that included expanding community schools and public pre-K in order to better address the non-academic needs of children that get in the way of learning. I figured the goal would be to equip each chosen school with certain critical social services that communities need. Mental healthcare, enrichment programs, adult education – these were all concepts I had heard mentioned in conversations surrounding the new initiative.
What Holly told me was that there isn’t a specific definition of a community school in terms of what it has, but rather in how it does. “A community school is a school that is actively engaged with parents and the community to make the school what the community wants and needs.” Philadelphia’s program would be centered on the creation of a full-time Community Schools Coordinator position at each selected school who would engage the teachers, students, parents, administration, community, and business or nonprofit partners to identify wants and needs, and then build partnerships to meet them.
So the selection process wasn’t going to be a formula and the plan for each school wasn’t going to be a checklist. A good candidate for a community school would have staff and administrative capacity and support, strong neighborhood buy-in, and established student and community needs. This project as well as the Stress Index aimed to provide context around those needs to help the Mayor’s Office of Education make these choices.
The final report for this project is not available publicly, but all of the data is public. The maps included here are provided for visual example only. We used public data on City-Owned Facilities, the School District’s Open Data portal, and ACS 2015 5yr Population by Sex by Age, and Families’ Ratio of Income to Poverty Level by Presence of Related Children under 18.
Starting with the school district’s data, we did some basic cleanup. Names and IDs aren’t always consistent between different data sources, so some editing was needed around that. Also, the Schools geographic layer itemizes facilities (i.e., buildings or grounds) whereas the District’s records reference institutions (schools and programs). Sometimes multiple schools might share a single site, so we had to add records to make sure we had a point for each institution, even if that meant duplicates right on top of each other.
We were interested in highlighting elements of the School District’s data that would characterize the attitudes students, teachers, and parents felt toward the school (parent and climate scores), academic performance and special needs (PSSA or Keystone test scores, ELL students and IEP students), functioning capacity of the school, behavioral problems, and safety (student attendance, teacher attendance, suspensions, police calls), as well as general facts about the school environment (enrollment type, gender ratios, racial makeup, grades served, school population). We tried to visualize these factors for each school in conjunctions that would be intuitive and useful (% of ELL students by school’s population size, parent climate rating by survey participation rate, teacher attendance relative to district average).
We mapped the existing resources near schools using the City Facilities dataset for city parks, libraries, and rec centers, as well as SEPTA stops, Indego bikeshare stations, Parks and Rec after-school programs, Police Atheletic League (PAL) centers, recognized commercial corridors, and community health centers.
With neighborhood data, applying the same logic of dasymmetric mapping used for the Stress Index, we identified the residential lots within the school’s catchment (defined as the neighborhood), as well as the Census tracts that intersect with that area. For each tract, we calculated the number (count) of Youth (under 18) living there as well as the number (percent out of families who have children) of families whose income is within 185% of the poverty line. We calculated a weighted average of each ACS value for each school’s neighborhood, assuming even distribution of population within each Census Tract amongst the area of residential parcels. We repeated this calculation for CDC BRFSS data modeled at the Census Tract level by TRF for PolicyMap for Asthma, Obesity, and Diabetes rates (not publicly available) for each school’s immediate neighborhood.
Together, these maps did not inform a set of criteria for Community Schools. Rather, they gave the program’s directors socioeconomic and academic context for the applications submitted by 31 schools in the first year. After Community School Coordinators were hired, they may refer to some of this information to help them in the process of working out what each community school wants to be.