Voices: Diving deep into SchoolChoice

Researcher Robert Reichardt has taken a close look at data gleaned from Denver’s new SchoolChoice process, and he has discerned some interesting trends.

This year Denver Public Schools implemented a new process to facilitate students’ ability to choice out of their home schools. The process has been blogged about here, and covered by the Denver Post here.   This article takes a deeper dive into the evaluation reports.

The new choice process consolidated over 60 different processes into one.  A computer program was used to assign students to schools based on student preferences, number of available seats, and school preferences (e.g. siblings, residents, or auditions for Denver School of the Arts). A second round  is open now through August 31, 2012 for students who are not happy with their current assignment or did not enter the first round.

A Transparency Committee of DPS administrators and principals along with community stakeholders was selected by A+ Denver to receive and interpret an evaluation report on the computer program used to make the assignments and a second on the information created by the choice process.  A+ Denver also provides spreadsheets of choice data by school. I helped plan the second project.

Choice is intended to improve parental satisfaction and improve schools. It creates pressure to improve through demand for seats in “good” schools and declining enrollment in less-preferred schools.   Demand is measured as either total requests, or first choice request per seat available in a school.

This demand measure is not perfect.  The analysis shows these measures are influenced by factors such as grades with available seats in a school and whether a school is new.

So what did we learn from the evaluations?

The choice process worked. DPS was able to collect over 20,000 hand-written choice requests and implement a complex computer program to assign students to schools.

There are huge differences in demand for schools. The differences are largest in high schools: Denver School of Science and Technology, Stapleton, had 8.2 first choice requests per available seat compared to Denver Online High School which had .01 first choice requests per seat.  This is a difference of 82,000%.

School ratings on the school performance framework (SPF) are related to student choices.  The points schools earned on the SPF are correlated with demand: More points equals more demand. Also, students in the lowest rated schools (On Probation) requested more schools (3.43) than those currently attending in higher rated schools (2.7 choices), suggesting they were more interested in finding other options.

Living close to good schools is important.  About two-thirds of students choose schools in their own region.  Student poverty (measured by free and reduced lunch status) was not correlated with making choices in different regions.  At the same time race was related to making choices in other regions: Hispanic students most often chose schools in their region. In other words, poverty may not be related to willingness/ability to travel to a school, but race may be.

Not surprisingly, parents of younger students are less likely to let their child travel to other regions for school while high school age students made the most choices in other regions.

The evaluation data highlighted that the availability of seats in higher rated schools varies significantly by region.  Equally important, Hispanic students tend to live in regions with the fewest highly rated schools. This probably explains why Hispanic students more often chose and were placed in lower rated schools. However, when requests for schools are controlled for, there are no differences by race in the schools where students were assigned.

The demand for early childhood education (ECE) was higher than available seats.  This was a particular problem in the near-northeast region where there were 1.9 requests for each ECE seat available. There was also higher demand than expected for kindergarten, where there were 1.25 requests per seat in the near northeast.  DPS guarantees a seat for all kindergarten students.

Finally, deep in the data is a worrisome suggestion of segregation.  At the elementary level only, demand was negatively associated with the proportion of non-white students in a school.  This does not mean that people are picking schools based on the race/ethnicity of students.  But, it does mean that changes in racial segregation within DPS should be monitored.

So what does this all mean?  The bottom line is DPS was able to implement a challenging new process that should incentivize school improvement and make choice transparent for parents.  Other districts should consider doing the same thing.

And DPS needs to keep working on having good schools, and slots for ECE throughout the district.

About our First Person series:

First Person is where Chalkbeat features personal essays by educators, students, parents, and others trying to improve public education. Read our submission guidelines here.