Results from the 2024 NYC School Admission Lottery Surveys

Amelie Marian
Algorithms in the Wild
10 min readMar 15, 2024

This post reports the results of the crowdsourcing survey for the 2024 NYC HS Admission Lottery. To learn more about the lottery and see the results for the three previous admission cycles you can read the first four posts in the series:
Part 1. Decoding the NYC School Admission Lottery Numbers (includes 2021 results)
Part 2. Gaining Insights from the NYC School Admission Lottery Numbers
Part 3. NYC High School Chances of Admission Predictions
Results. Results from the 2022 and 2023 NYC School Admission Lottery Surveys.

You can also read a more in-depth account of the study I presented at the 2023 ACM Conference on Fairness, Accountability, and Transparency (ACM FAccT’23): Algorithmic Transparency and Accountability through Crowdsourcing: A Study of the NYC School Admission Lottery, or watch the conference presentation.

This data is reported for informational purposes and can not be copied or reproduced without permission.

As I reported last year, the NYC DOE (now called the NYC Public Schools) has made significant improvements in the transparency of the process, implementing many of the changes I have been advocating for in my previous posts. In particular, families are now given their lottery number at application time, and the number of applicants reported on MySchools only shows the students who were considered by the school (i.e., who were not matched higher on their list) instead of all applicants who listed the school, and MySchools separates information about the number of applicants by priority groups. These changes have made the process easier to navigate for families, but information on the actual cutoffs at each school is still missing. Many families have told me how useful the results of past surveys were for helping them create their list of choices and managing their expectations; list strategizing is a frequent topic on the Applying to High School in NYC Facebook group. This post provides crowdsourced information from students who applied during the Fall 2024 application season.

Survey Results: Methodology

The data was collected through Google surveys in early 2024. The data is self-reported by parents, and errors in data entry are possible; the following results should be interpreted with this in mind.

  • As of April 30, 2024, there were 789 answers to the HS survey. I manually cleaned the data to remove obvious errors in data entry.
  • The survey recorded students’ lottery numbers, as well as their groups, Ed. Opt. category and assessment/audition score (if applicable). I extracted the highest (worst) lottery number, along with the priority category (or score) of all students who received an offer to a school. To identify the lowest (best) lottery number declined, I only looked at schools that were ranked higher in the student’s choice list than the one they matched to, as the algorithm does not consider students for schools lower on their list than the one to which they match. More details on the process can be found in my previous posts.
  • Survey answers are not representative of the whole applicant population: respondents are clustered in some geographical areas, skew higher income and higher achieving than the DOE student population. This does not impact the correctness of the cutoff information derived from the surveys but does impact the completeness of the information.
  • Survey participants were asked whether they qualify for FRL (free and reduced lunch, the DOE measure of low-income status) diversity in admissions (DIA) set-asides, and whether their student had a student with disability (SWD) designation. About 9% of answers reported qualifying for FRL, and 11.5% for SWD.

Historical Survey Data (new in 2024)

The above table reports on the statistics of the data generated by each year’s survey. The overall school systems statistics are given in the last column but only include data from 2021-2023 as the 2024 data will not be available until the Fall.

As noted above, the demographics of survey participants skew towards higher-income students and students who do not have disabilities, although the representation of these two categories improved from year to year. Approximately 440 high schools participate in the match (the exact number varies as some schools are often added or removed at the last minute), with several schools offering multiple programs. Over the years, the survey data has contained information for about a fifth to a quarter of the schools, showing the lack of completeness of the information gathered.

An interesting observation is the change in the rate of match of survey participants over the years. While the actual match rate for the whole DOE system (red column) stayed stable year to year, the results for the survey participants changed significantly. This can be explained by various factors:

  • For the first couple of years, the DOE did not routinely provide families with their lottery numbers. Families had to request it after the match. The families who were incentivized to go through the process were often those who were unhappy with their match (or lack thereof) and were organizing on social media to try to understand what had happened.
  • In the past two years (2023 and 2024), families have been provided with their lottery numbers at the time of application, which allowed them to strategize their list to ensure a match. Interestingly, access to the lottery number did not seem to improve the actual citywide rate of match significantly. In contrast, survey participants received on average much better outcomes.

This shows that the information collected from the surveys helps families. The survey results were made public starting in 2021. There is a high correlation between survey respondents and families who were aware of the survey results at the time of application. Many parents online have expressed their intent to “pay it forward” by filling out the survey for future applicants. This suggests that giving better information about the odds of matching at different schools helps improve outcomes. Hopefully, these results will convince the DOE to make all cutoffs public.

High School Admissions Results

I present the results by separating school programs into five different categories based on their admission methods. School admission methods were identified through the publicly available HS Directory for Fall 2023. For ranked screened (essay-based) schools, I used the list available on the NYC DOE website. To identify schools with DIA set-asides, I used the list on this DOE web page.

For the first time since the pandemic, the HS admission process has not changed compared to the previous year. This gave more predictability to families, and as we can see by comparing this year’s results with last year’s, school cutoffs stayed relatively stable.

Screened High Schools

For admission to screened schools in 2024, students were placed into 5 groups based on their final seventh-grade core course grades. Admissions to screened schools were done in group order, with ties broken by lottery number.

The following table shows, for each screened school, the highest Group and lottery number within that group that gained admission and the lowest (group, lottery number) that was denied admission. For instance, a Gen-Ed, non-DIA applicant in Group 1 with a lottery number starting with ‘18’ was admitted to 02M416: Eleanor Roosevelt HS, but a Group 1 applicant with a lottery number starting with ‘19’ did not get an offer. (This is very similar to last year’s data, where a Group 1 applicant with a lottery number starting with `17' got in, but one with ‘19’ did not get in.) The school sets aside 50% of seats for low-income students (FRL), making the odds of admission for a DIA-eligible Gen-ed student higher but students still need to be in Group 1 and have a lottery number lower than ‘49’ (possibly lower) to receive an offer.

The results are inconsistent for two schools:

  • 22K425 : James Madison High School (22K425) : K25B : Medical/Health Professions (K25B): this school offer several programs, the conflicting results may be due to the two students applying to different programs and that information being incorrectly reported in the survey.
  • 22K535 : Leon M. Goldstein High School for the Sciences (22K535) : K76A : Leon M. Goldstein High School for the Sciences (K76A) had a group 3 student with lottery ‘b3’ accepted, but another group 1 student with lottery ‘ac’ not accepted. This school has geographic priorities, which the survey does not record, which can explain the discrepancy.

Ranked Screened (Assessment-based) High Schools

Several schools used both assessments and grades. Students were ranked on a composite score based on their school-specific assessments (typically essays), and grades (based on a coarse mapping of their Group to a grade score). Details on the scoring weights for each school can be found on the DOE website. The composite score is out of 100.

As usual, ties are broken by lottery numbers. Results for these Ranked Screened schools are shown in the table below.

As was the case in previous years, some schools gave 100 to most students, while others used the whole grading scale.

  • To gain admission to 03M479 : Beacon High School, students who did not qualify for a set-aside needed a perfect 100 assessment score AND a lottery number that started with ‘87’ or lower (the actual cutoff being between ‘7b’ and ‘87’). DIA-eligible students could be admitted with a lower score, so the lottery number had less of an impact. Similar results are shown for 01M376 : NYC iSchool.
  • 03M541 : Manhattan / Hunter Science High School seems to have once again assigned a maximum score of 100 to all essays. The composite score is composed of 70% course grade (using the coarse mapping of groups to grade) and 30% essay. For non-DIA students, this resulted in a lottery process among all Group 1 students, with a number lower than ‘39’ needed for admission. There were not enough answers to determine a cutoff for DIA-eligible students.
  • Both Bard High Schools: 01M696 : Bard High School Early College and 24Q299 : Bard High School Early College Queens provided fine-grained grading for their assessments, which means that students with bad lottery numbers had a chance at getting an offer if their assessment scores were high enough. For instance, students with a composite score greater than 88 for 01M696 : Bard High School Early College were admitted regardless of their lottery numbers, students with scores of 88 needed a lottery number starting with a number lower than 3.

Audition High Schools

The audition school process is similar to that of the ranked screened schools above, except that the score is based solely on the audition and does not take into account the Group of the student.
Audition schools did use the whole scale to score students, making the lottery number far less important in the chances of admission.

The results are inconsistent for one school: 30Q501 : Frank Sinatra School of the Arts High School (30Q501) : Q40J : Fine Arts (Q40J), where an answer reported not getting an offer with a score of 80. This is contradicted by several other answers and is likely due to a data entry error with the respondent selecting the wrong program or wrong priority; for clarity, I did not include the data point in the table below..

Open Admission High Schools

These schools are purely lottery-based, all students regardless of grades are placed in the same lottery pool. Some schools have DIA set-asides or SWD seats; when available data for these is separated (Priority column).

Ed. Opt. High Schools

These schools place students into three pools: High, Middle, and Low, based on their academic performance. The results below show for each school the cutoffs for the three categories when the information is available.

Here too there are two conflicting data points for the High Priority group at 21K525 : Edward R. Murrow High School (21K525) : K57A : Communication Arts (K57A) and the Middle SWD Priority group at 02M580 : Richard R. Green High School of Teaching (02M580) : M23B : Liberal Arts Academy (M23B). Once again this is likely due to data entry errors.

ASD/ACES Programs

Here are the results for students admitted to ASD/ACES programs. Because there is little information as to how students are ranked for these programs, I report on the lottery numbers that were accepted and rejected, and on the group if that information seems to have been taken into account in the ranking.

Here too there is a conflicting data point for 30Q301 : Academy for Careers in Television and Film (ACTV) (30Q301) : Q01Y : Academy for Careers in Television and Film ASD Nest Program (Q01Y). Once again this is likely due to a data entry error.

D75 Programs

Here are the group and lottery numbers for students admitted to screened D75 programs. Here as well, it is not clear whether schools use group information to screen students.

List of High Schools where unmatched students were assigned

Finally, this is a list of schools to which unmatched students have reported being assigned.

  • 02M282 : Urban Assembly Maker Academy (02M282)
  • 02M500 : Unity Center for Urban Technologies (02M500)
  • 02M580 : Richard R. Green High School of Teaching (02M580) : M23A : Teaching Academy (M23A)
  • 15K464 : Park Slope Collegiate (15K464)

All material appearing on this blog post/website (“content”) is protected by copyright under U.S. Copyright laws and is the property of Amélie Marian. You may not copy, reproduce, distribute, publish, display, perform, modify, create derivative works, transmit, or in any way exploit any such content, nor may you distribute any part of this content over any network, including a local area network, sell or offer it for sale, or use such content to construct any kind of database. You may not alter or remove any copyright or other notice from copies of the content on this blog/site. Copying or storing any content except as provided above is expressly prohibited without prior written permission of Amélie Marian. For permission to use the content on this blog/site, please contact Amélie Marian.

--

--

Algorithms in the Wild
Algorithms in the Wild

Published in Algorithms in the Wild

Advocating for transparent and accountable decision processes. Studying the behaviors of algorithms when they get out of the minds of designers and take on a life of their own.

Amelie Marian
Amelie Marian

Written by Amelie Marian

CS Professor at Rutgers — I like to explain algorithms and advocate for accountable decision processes.