Sumner Academy of Arts & Sciences is once again the the national news based on the grades of the students there during 2017-2018...

How U.S. News Calculated
the 2020 Best High Schools
RankingsWe looked at thousands of public schools to identify
​ the top performers.
By Robert Morse and Eric Brooks April 20, 2020, at 9:00 p.m.
More


More
A GREAT HIGH SCHOOL educates all students from different social and economic backgrounds, exposing them to challenging coursework on the path to graduation. The highest ranked U.S. public schools in U.S. News & World Report's 2020 Best High Schools rankings are those whose attendees demonstrated outstanding outcomes above expectations in math and reading state assessments, passed a diverse array of college-level exams and graduated in high proportions.

Under U.S. News' 2020 rankings methodology, the vast majority of all public high schools are ranked.

In coordination with North Carolina-based RTI International, a global nonprofit social science research firm, U.S. News ranked approximately 17,790 public high schools out of more than 24,000 reviewed. This is the count of public high schools that had a 12th grade enrollment of 15 or greater, or otherwise had sufficient enrollment in other high school grades during the 2017-2018 school year to be analyzed.

We did this by summing their weighted scores across six indicators of school quality, then computed for each school a single zero to 100 overall score reflective of performance across these metrics. The overall scores depict how well each school did on a national percentile basis. For example, a school with a score of 60 performed in the 60th percentile among all schools in the rankings.

Ranks were assigned in descending order of overall scores. Schools below the 25th percentile have their scores concealed and display the entire bottom quartile's ranking range. Schools without a grade 12 or with very small enrollment are simply displayed as unranked.

Why Did U.S. News Take This Approach?

With students, parents, educators and the general public in mind, U.S. News made the rankings as thorough as possible by including as many schools as possible.

The Best High Schools methodology enables users to understand why some schools place ahead of others. Each school’s rank is linked to a single overall zero to 100 score. For example, a school with an overall score of 78 has placed in the 78th percentile nationally and thus is ranked more highly than a school with a score of 77. That score of 78 also means that 22% of all the ranked schools ranked higher and 78% of the schools ranked lower.

For transparency, each school's directory page includes indicator ranks that explain how it performed nationally and within the state on each ranking factor.

This is the second year the high school rankings have been calculated using the current methodology. Consequently, only rankings from the 2020 and 2019 editions are comparable to each other. As always, U.S. News does not display rankings that predate the latest edition.

Data Sources

U.S. News does not collect any information directly from high schools. The data used to produce the Best High Schools rankings and published on usnews.com came entirely from the following third-party sources:

• The Common Core of Data is the U.S. Department of Education website, updated annually, that contains basic data on enrollment, student ethnicity and other profile information on all public high schools in the U.S. The DOEd collects the data directly from the schools, school districts or state departments of education, which is reflected on usnews.com.

• Statewide math and reading level assessment tests and high school graduation rates data used in the 2020 rankings in almost all cases is from the 2017-2018 school year. This data in most cases is from each state's education agency websites or directly from state education agencies.

• The College Board was the source of the Advanced Placement examination data for each public high school, when applicable, that was used in rankings calculations. The AP exam data used in the analysis is for 12th grade students in the 2017-2018 school year.

• International Baccalaureate was the source of the IB examination data for each public high school, when applicable, that was used to create calculated values. The IB exam data used in the analysis is for the 12th grade students in the 2017-2018 school year.



Overall National Rankings
Below is the description of each of the six ranking indicators and their weights used to produce the overall score:

• College Readiness (30%)

The College Readiness Index, or CRI, is measured by the proportion of a school's 12th graders that took and passed AP or IB exams.

There are two elements. One is a participation rate – the number of 12th grade students in the 2017-2018 academic year who took at least one AP or IB test before or during their senior year, divided by the total number of 12th graders at the school. The other is a quality-adjusted participation rate, defined as the number of 12th grade students in the 2017-2018 academic year who took and passed – received an AP score of 3 or higher or IB score of 4 or higher – at least one AP or IB test by their senior year, divided by the number of 12th graders at that school.

Quality-adjusted participation was weighted at 75%, and the simple participation rate was weighted at 25% toward CRI. The maximum CRI possible is 100 if every 12th grader at a school took and passed at least one AP or IB exam by their senior year.

In cases where schools had both AP and IB results, U.S. News created a blended CRI that gave a proportionally larger weight to the program that had greater participation. All AP and IB subject tests were considered when determining whether a student took or passed at least one AP or IB exam.

Many schools without any AP or IB exam test takers scored a zero on this ranking indicator. But adjustments were made so that schools without APs and IBs would not score significantly worse than schools with very few APs and IBs even after the scores were standardized. In summary, not having any AP or IB exams was not enough alone to sink a school to the bottom of the rankings.
 

• College Curriculum Breadth (10%)

The College Curriculum Breadth Index, or CCBI, is calculated among a school's 2017-2018 12th graders from the percentage who took and the percentage who passed multiple AP or IB exams. Students who took and passed exams in four AP or IB content areas earned full credit. Those passing two or three AP or IB content areas were given partial credit – 50% and 75%, respectively.

For both exams, the percentage of students taking exams in multiple areas was weighted at 25% and the percentage of students passing exams in multiple areas was weighted at 75% to derive a CCBI score from which the natural log was taken to handle outlier cases. Altogether, high schools where the largest proportion of 12th grade students in the 2017-2018 academic year took and passed AP or IB tests in at least four AP or IB content areas scored highest.

The AP exam content areas measured were English; math and computer science; sciences; world languages and cultures; history and social science; arts; and AP capstone. For the IB exam, content areas measured were studies in language and literature; language acquisition; individuals and societies; sciences; mathematics; the arts; extended essay; and theory of knowledge.  

• Math and Reading Proficiency (20%)

Each state issues standardized tests measuring student proficiency in subjects related to mathematics and reading. The states often look closely at student performance on these tests to determine whether learning in core subjects is achieved and to review how well schools are educating their students. At least passing some of these assessments may be a requirement for students to graduate.

The math and reading proficiency indicator is a simple measure of schools' student performance on these assessments. For example, if a state's grading system assigns a 1-4 to every assessment and each student at a school earned a 4 for math and a 4 for reading, that school would achieve a perfect score.

Schools' total assessment scores were compared against other schools from their states. The distribution of scores within each state was then considered to compare how a school's relative performance in one state equated to a different school's relative performance in another state.

• Math and Reading Performance (20%)

This ranking indicator also is derived from math and reading state assessments. But in this case, the total assessment scores are compared with what U.S. News predicted for a school with its demographic characteristics in its state.

In all 50 states, there is a very positive statistical relationship between the proportion of its student body that is black, Hispanic and/or from a low-income household – defined as being eligible for free or subsidized school lunch – and a school's results on state assessments. Schools performing best on this ranking indicator are those whose assessment scores far exceeded U.S. News' modeled expectations.

• Underserved Student Performance (10%)

This is a measure assessing learning outcomes only among black, Hispanic and low-income students. This evaluates how well this underserved subgroup scored on state assessments compared with the average for non-underserved students among schools in the same state. Schools performing above the 50th percentile nationally in this comparison received the highest score while other schools’ scores decreased the greater the distance between their underserved students and their state’s median for non-underserved students.

• Graduation Rate (10%)

Among students who entered ninth grade in the 2013-2014 academic year, this measure is the proportion who graduated four years later by 2017-2018. Graduation rates are an important indicator of how well a school is succeeding for all its students. 

Methodology Continued
Schools were required to have assessment data available in order to be reviewed for a ranking. Some imputations were made, though, for missing data for student body, subgroup level assessment data and graduation. In some cases, data from 2016-2017 was used when 2017-2018 data was not available.

For all six ranking indicators, each school's scores were standardized about their means and divided by their standard deviations to account for statistical variance. For example, the difference between an 89% graduation rate and an 85% graduation rate may not seem like much on an absolute basis, but it matters significantly on a relative basis for the rankings because the national average graduation rate, according to the National Center for Education Statistics, is around 85% and many schools' graduation rates fall close to that proportion.

The weights were selected based on research and coordination between U.S. News and education experts at RTI on what factors matter most. State assessments contribute most to the rankings because data was used for every school. Additionally, states are known to place greater stock on these scores than on graduation rates when assessing schools. Math and reading proficiency and performance received double the weight of underserved student performance because they factor the absolute and relative performance of all students at those schools, including but not limited to underserved students. For college-level exams, three times the weight was placed on College Readiness Index versus College Curriculum Breadth Index because in preparing students for college, providing at least some exposure to AP and IB exams is believed to be of greater importance than whether students take the heaviest AP or IB course loads possible.

Assigning Ranks

After the six indicators that had been standardized were weighted, those weighed scores were summed and then transformed so that each eligible school receives an overall percentile score between zero and 100 at two decimal places, with the top performer scoring 100. The overall score is calculated as a percentile score that indicates what percentile position a school is in out of the nearly 18,000 ranked schools. For example, a school with a score of 90 means that 10% of the high schools are ranked higher and 90% of the schools are ranked lower. The overall score is published rounded to two decimal places like 99.96 to show how close schools are in the rankings. Finally, high schools are ranked against peers in descending order of their overall scores. The numerical ranking was based on the overall score carried out to many decimal places in order to prevent ties. In total, six schools were tied in the rankings. High schools placing in the top 75% display their individual rank on usnews.com. For lower-performing high schools in the bottom 25% of the rankings, U.S. News made an editorial decision to not display the actual rank. Instead, U.S. News is displaying the ranking range of the bottom quartile for those schools,#13,345-17,792 on usnews.com.

For the most detailed explanation of the methodology, read the Best High Schools Technical Appendix.

State Rankings

• The state rankings use the same methodology as the national rankings, but ranks are re-sorted among schools within the state. For example, the highest nationally ranked U.S. school located in Virginia is also the No. 1 school in the Virginia state rankings. Schools in the bottom quartile ranking range in the national rankings display a state-specific ranking range in their state rankings based on how many schools in each state placed in the national bottom quartile. Also, if a school is ranked in ranking range nationally, meaning it's ranked in the bottom 25% nationally, it will be in a ranking range in its state ranking. For example, a Virginia school in the national ranking range #13,345-17,792, would be ranked #278-320 in the state of Virginia rankings.

Metropolitan Rankings

For the second time, U.S. News ranked high schools by metropolitan area. The metro rankings use the same methodology as the national rankings but schools are grouped and re-ranked within their metropolitan areas when applicable. For example, the highest nationally ranked U.S. school located within the Chicago metropolitan region is also the No. 1 school in the Chicago metro rankings. High schools that have their ranking published as a ranking range in the national and state ranking also display a ranking range in the metro rankings. Their metro area ranking range is based on the total number of high schools in their metro region and the number of schools in their metro region with a ranking in the ranking range.

U.S. News strictly linked its metropolitan areas to Core-Based Statistical Areas, or CBSA, as defined in 2010 by the federal government’s Office of Management and Budget, or OMB, using U.S. census data. A metropolitan area may encompass parts of multiple states. For example, Chicago’s CBSA region includes the city proper, nearby parts of Illinois and neighboring Indiana and Wisconsin. For some of the largest metropolitan areas, U.S. News made an editorial decision to relabel for display purposes the government’s CBSA descriptor.

In 2020, in order for high schools to be ranked by metropolitan area, a metropolitan area had to have at least three or more ranked high schools in the national rankings. This is a change from 2019 when the minimum was one ranked high school. This year, 14,892 high schools, or 83.7% of all ranked schools, were ranked in 811 metropolitan areas, down from 933 in 2019.

Charter and Magnet Rankings

For the second consecutive year, all 1,882 charter and 931 magnet schools that were eligible will be ranked. The charter and magnet school rankings methodology looked at all public high schools nationally that were designated as charter or magnet schools, or both, as reported to the DOEd in the 2017-2018 Common Core of Data. From there, it uses the same methodology as the state rankings that are based on the national rankings.

• The charter and magnet rankings are based on whether a school is designated as a charter or magnet school, or both, and then sorted among charter and magnet schools, respectively, based on national rank. For example, the charter school that is third highest ranked in the national rankings is the No. 3 school in the charter school rankings, and the magnet school that is third highest ranked in the national rankings is the No. 3 school in the magnet rankings. Schools in the bottom quartile ranking range in the national rankings display charter-specific and magnet-specific ranges specific to those schools. For example, a school in the national ranking range #13,345-17,792 would be ranked #1,053-1,882 in the Best Charter High Schools rankings.

This same rankings methodology used to produce the charter schools ranking was also followed to produce the Best Magnet Schools.

• STEM rankings: To determine the top science, technology, engineering and math schools, U.S. News looked at its top 1,000 ranked high schools from the national rankings and then evaluated the participation and success of those schools' AP students on the science and math exams. The top 250 schools, unchanged from last year, were ranked in the 2020 Best High Schools for STEM. Read the full STEM methodology.
• State rankings and demographic profiles of high schools: U.S. News published a state-by-state ranking of the 2020 Best High Schools based on the percentage of all of each state's ranked schools performing in the top 25% of schools nationally. This ranking reflects which states are most successfully educating students, based on state test scores, graduation rates and the proportions of students participating in and achieving passing scores on AP and IB exams.
• School district rankings: For the first time, U.S. News will publish distinct rankings of high schools by school district. In other words, each school district with three or more ranked high schools will have a No. 1 ranked school for that district, and so forth. A school's rank in its district will be based entirely on its national rank. For example, if the highest-ranked high school in a school district is No. 30 nationally, then that school is also ranked No. 1 in that state; if the second highest-ranked school in that same district is No. 250 nationally, then that school is ranked No. 2 in that district. High schools that have their ranking published as a ranking range in the national and state rankings also display a ranking range in the school district ranking. Their school district ranking range is based on the total number of high schools in their school district and the number of schools in their school district with a ranking in the ranking range. In total, there were 7,277 high schools ranked in 1,018 districts.

For a more detailed methodology, see the technical appendix produced by RTI.