Blog Archive

Archives by Month:

October 2014

National Network Spotlight

National Network Spotlight

My favorite time of the season!  Halloween?  Sure, it’s fun, and I’m looking forward to Friday, but how much data can we squeeze out of that day?  I’m just returning from my hands down favorite event in the fall – the National Neighborhood Indicators Partnership (NNIP) meeting.

For those who haven’t heard me rave before, NNIP is a national network of organizations that share a commitment to creating and maintaining neighborhood-level data systems and helping their communities use the information to make better decisions.  We meet up twice a year to talk about our data shops – new discoveries, advances in analysis and online tools, challenges we’ve encountered, and much, much more.  D3 has been a proud partner organization since 2009.  Partners take turns hosting the meeting, and this time we went to Denver, Colorado, where the Piton Foundation makes data come alive through their Data Initiative.

As always, I learned so much last week from all of our partners in the network.  Here are some of the highlights I took away…

Nonprofit does not mean we can keep going without funds!

(A quick aside: I’m using the term nonprofit here, but I agree with Doug Bitonti Stewart of the Max M. and Marjorie S. Fisher Foundation that we need find a new term for our sector such as the “for impact sector”.  Why should we define ourselves by what we are not, versus what we are for?)

As D3 approaches its sixth anniversary, we’ve been doing a lot of thinking here about how we fit into the local environment, how we can best serve our community through the products and services we provide, and how to ensure D3 can continue to serve our community over the long haul.  As we grow up, we are also turning our data-driven philosophy inward and discovering some really important information about ourselves.

Taking a data-driven approach on the inside as well as out is certainly not new among the national partners, but it seemed to be a common theme of our meeting this fall.  Many of us are working through formulating business models, diversifying funding streams, monitoring internal performance, and understanding external impact.  Check back in the near future to read more about what D3 is learning through our investigation and analysis.

Our partners are building awesome online tools

Rhode Island Community Profiles from the Providence Plan show us a creative and expertly crafted use of the new Census Bureau API.  And the code is on Github!

Greater Portland Pulse is demonstrating incredible potential for regional collaboration and metric tracking for a region comprised of seven counties in two states.

My hands down favorite illustration of the uncertainty we should feel when using American Community Survey (ACS) data: The Randomizer.  The tool generates random values for households in poverty by census tract between the range of possible values given by ACS, and a resulting total value of the households in poverty based on the tract numbers generated.  Thanks to John Garvey for sharing his creation with me, and to Urban Strategies Council in Oakland for always pushing the cutting edge.

We can all be powerful data advocates

As John’s tool depicts, American Community Survey (ACS) data can present some challenges when presenting data for small areas.  It seems I find myself engaged in multiple discussions at every NNIP meeting about ACS data: to use or not to use? Are the data reliable enough or just garbage?  Should we replace ACS with locally sourced data, data which may be more accurate, but less comparable to other places and much more time consuming to produce for more than one jurisdiction?

I didn’t walk away with any clear answers, but I do know one thing: if we appropriately fund the ACS as a nation the data would be more reliable than they are now, which is important not only for the annual $400 billion in federal and state fund disbursement which relies on the data, but also for all the community development and planning work that goes on every day at a local level across the country.  Congress appropriates ACS funding – let your congressperson know you care!  And I beg you – if you are lucky enough to receive an ACS questionnaire, please fill it out!  I’m still waiting for mine…

NNIP data advocacy work extends beyond government or administrative data use, it also includes maintaining open data sites and convening data user groups.  D3 only recently launched our Open Data Site, but some partners have been publishing open data in their respective cities for years.  Some have even developed sophisticated user groups, trainings, and conferences to more broadly engage their communities in the practical use and application of data.  Why is this so important?  William Gibson put it best: “The future has already arrived.  It’s just not evenly distributed yet.”

Baltimore, Columbus, Indianapolis, and the MacArthur Foundation (on behalf of the Chicago School of Data) shared with us their experiences with local user conferences during the last day of the meeting.  I still have much to digest and apply to D3’s trainings and advocacy for improving data literacy, but in the meantime, I’ll suggest readers refer to NNIP’s putting open data to work in communities.

Thanks NNIP for gathering us all together once again!

New Student Dispersion Tools

New Student Dispersion Tools

Just as longer commutes can have detrimental effects on adults, it reasonably follows that longer school commutes may have such effects as an increase in stress, tardiness and obesity rates on our youth. Conversely, in areas where housing patterns concentrate poverty and race in a neighborhood, longer commutes outside those neighborhoods might improve student outcomes. Though it is not yet clear just how school commute distance effects student performance, what is clear is that Detroit students and families are exercising their choice. Given the potential problems and benefits, where do these patterns exist and how might schools and families adapt?

For the first time, in partnership with The Skillman Foundation, Excellent Schools Detroit and Great Gains, Data Driven Detroit was able to perform our student dispersion analysis with data covering all publicly funded schools in Detroit*. The Michigan Center for Educational Performance and Information (CEPI), working with the Michigan Center for Shared Solutions (CSS) provided Data Driven Detroit unidentifiable enrollment data and census block codes approximating student residence location from the fall of 2013. CEPI also provided information for students attending schools in the Cities of Warren, River Rouge, Southfield, Hamtramck and Highland Park. This summary analysis is the most comprehensive student dispersion analysis D3 has done to this point in terms of breadth of schools, and includes data from more than 141,000 students and 300 schools!

Student Dispersion map of Chrysler Elementary School Student Dispersion 13-14_Western International High School

Student dispersion map examples: Chrysler Elementary School and Western International High School

The Detroit Context

Educational reforms in Michigan have broadly opened up public grade school options and resulted in a complicated school environment where families have many choices, near and far, in where to send their kids to school.

Detroit’s school context is dynamic, if not chaotic, where several schools may open and close year to year, and where there are a number of different public school systems serving students, including suburban schools of choice, charter schools (Public School Academies), the Detroit Public Schools (DPS) system, and the state-run Educational Achievement Authority (EAA). With more options, inconsistent transportation, school openings and closings, and other reasons, some students may find themselves, either by choice or necessity, with a more difficult commute. Because of this complex environment, it is not well understood just where youth from different parts of the city choose to attend school and how far they go.

The way students in Detroit are dispersed may have many important implications of interest for educators and planners.

  • Which schools have the longest commutes for students? Which schools fail to draw students outside their neighborhood, and why?
  • Are there parts of the city that have a stronger local draw?
  • Are there parts of the city whose students are more mobile?
  • Are certain students drawn to DPS schools vs. charter schools vs. the EAA schools?
  • How can future transit decisions be made that help students get to the schools that they attend?
  • If schools create neighborhood identity and community by causing local youth to attend classes together, are there neighborhoods that might be disproportionately affected because students aren’t attending local schools?
  • In schools that have further average commutes, how can educators help mitigate those students’ further commutes?

Data Driven Detroit is now making available a couple of new resources to help explore this data. First, our new interactive map allows users to view the locations of schools by type or level and then download digital maps showing the dispersion patterns of each individual school. In addition, we have calculated the average student commute distance for every public school in Detroit and summarized our results in a brief report.

While these data and map resources will not answer these questions on their own, we hope that these tools are useful to researchers, policy makers and educators as they make plans and formulate policy to improve the educational environment in the city.

New Resources

Student Dispersion Maps
D3’s student dispersion maps show the patterns of where students who attend certain schools live. Our tool allows users view maps of any publicly funded school from across the city by level, type, or by searching by name. These maps reveal spatial patterns that would be impossible to understand without this type of visualization.

 

Commute Distance Summary
The commute distance summary report helps to quantify just how far students from different schools are traveling, and how different types of schools compare against each other. This summary is meant to help educators understand the commute burden placed on their students and assist administrators in planning for school location. The summary lists the average distance for each public school. In addition, the school locations from our study, complete with average distance information, are available as a GIS shape file or table from Data Driven Detroit’s Open Data Portal  (Search for “commute”).

Next Up
D3 will be following up this blog post in the near future with two more blogs on this topic. The first will take a deeper look at a few interesting dispersion pattern maps.  The second will take a closer look at the average commute distance analysis.

Since this analysis elicits more questions than answers, D3 is hoping to continue its analysis of student location data. Through our partnership with ESD and CEPI, we are expecting to get a more complete data set including not just students that attend Detroit publicly funded schools, but also Detroit residents who attend schools outside the city. As a large percentage of Detroit youth do attend school in the suburbs, this next data set should draw a much more complete picture.

 

 * Ann Arbor Trail Magnet School was mistakenly excluded from the data set delivered to Data Driven Detroit.

Motor City Mapping Mini-Grants Come to a Successful Close

Motor City Mapping Mini-Grants Come to a Successful Close

By Stephanie Edlinger