Increasing Economic Diversity at a Flagship University: Results from a Large-Scale, Randomized Trial

The intervention consists of sending materials to the students, their parents, and their school principals. The materials promise four years of free tuition and fees, conditional on acceptance to the Flagship University, and also provide information about the Flagship and the application process. Mailings are delivered in the late summer and early fall of senior year.

For all cohorts, we randomized at the school level. That is, every student in a school who meets the sample criteria is assigned the same treatment status. We also stratified randomization by the number of students eligible in each school, so that randomization took place within each of four school size groups (schools with one, two, three, or four or more students meeting our sample inclusion criteria). Assignment to treatment was done once per stratum (pure randomization within strata).

For the second year of randomization, all first-year treatment and control schools with HAIL-eligible students in the second cohort retained their treatment or control status. Schools with HAIL-eligible students in the second cohort that were not part of the sample in the first cohort (due to no students in the school meeting the HAIL criteria in the first cohort) were randomized according to the same cluster-stratified method as the first cohort. Similar processes were conducted for years three and four.

2015-09-01

2018-09-30

All outcomes are measured using administrative data from the Flagship and the state Department of Education. The main outcome of interest is attending a college at least as selective as the Flagship University. This outcome, as well as time spent in college, degrees earned, and major of any degrees, will be measured using data from the National Student Clearinghouse. For students who enroll in public colleges in Michigan, we have transcripts that include credits earned, grades and major. Intermediary outcomes include applying to the Flagship University and timing of application (we do not have application information for other colleges).

Cluster-stratified randomized controlled trial. The intervention will be clustered at the school level, and stratified by the number of students eligible for the intervention in each school.

Not available

Randomization done by a psuedorandom number generator in Stata.

Clusters at the school level.

Yes

Year 1: 529 schools, Year 2: 498 schools Year 3: 666 schools Year 4: 324 schools

Year 1: 2,108 students. Year 2: 1,806 students Year 3: 3,516 students Year 4: 1,181 students

Year 1: 1,057 treated students in 262 schools and 1,051 control students in 267 schools. Year 2: 1,806 students from 498 schools, with 879 treated students in 239 schools and 927 control students in 259 schools. Year 3: 3,516 students from 666 schools, with 1,833 students treated in 349 schools and 1,683 control students from 317 schools. Year 4: 1,181 students from 324 schools, with 737 treated students from 199 schools, and 444 control students from 125 schools.

Using data from 2013, the latest cohort for whom we have available outcome data, we identified students who met the eligibility requirements for the HAIL scholarship. There were 2,692 such students in 654 schools (4.12 students per school). Matching their data with college enrollment data, we see that schools had an average of 8.7 percent of eligible students enroll in a highly or most selective institution. This provides a baseline estimate of the outcome of interest prior to the implementation of the treatment. In the first year of the intervention, we have 529 schools with roughly half in treatment and half in control. We expect to have a similar sample in year two. We estimate the minimal detectable effect (MDE) size for the study of 0.048. This estimate of the MDE does not account for our strata. This estimate is consistent with other estimates in the literature. Hoxby and Turner (2013) estimate an effect size of 0.12 on a student enrolling in a “peer” college, which the authors define as a college where the median student scores within five percentiles of the student’s own score.

HAIL Preanalysis Plan