Methodology and Data

As the preeminent architect Robert Burns once noted, "the physical design of the campus makes a fundamental contribution to the pursuit of academic excellence."  Architectural beauty is determined by inclusion in Architectural Digest’s (2025) 64 most beautiful college campuses. Each college was assigned either a 1 (inclusion) or a 0 (exclusion).  

Historical significance is determined by the number of National Historic Landmark (NHL) structures on the school campus. If the entire campus (as opposed to an individual structure) is designated a NHL or a National Historic Landmark District, it was given a weight of 5. 

A college's endowment relative to the size of its student body is reflective of its capacity to accept students without regard to socioeconomic background, provide generous financial aid for those in need, offer outstanding research and travel opportunities, build state of the art facilities, and maintain its buildings and grounds. Data were obtained from the 2024 National Association of College and University Business Officers website and in some cases from school annual reports.

In an era of grade inflation, academic performance relative to one's classmates is a better metric of student achievement than GPA alone. Data were obtained from each college or university's most recent Common Data Set.** 

Adjusted SAT scores were calculated by multiplying the 25th percentile scores by the percentage of students who did not submit scores and by multiplying the 50th percentile scores by the percentage of students who did submit scores and then adding the results. SAT scores were scaled from 0–1 by subtracting the minimum possible score (400) from each school's adjusted median. Data for each school were obtained from the NCES college navigator site (Fall 2023) or the 2024-2025 common data set. 

A measure of the quality of a school's faculty. Students at an elite college or university should be taught predominantly by professors with the highest level of training in their respective fields. Data were obtained from each school's most recent Common Data Set.**

A measure of the accessibility of teaching faculty to students. Student to faculty ratios were obtained from each college or university's most recent Common Data Set.**

Student retention, defined as the percentage of first year students who return the following year, is an important and objective measure of student satisfaction. Data were obtained from each school's most recent Common Data Set.** 

An important measure of the financial success of graduates  relative to the price of their school and relative to those who chose not to attend college. We used Third Way's 2024 Price to Earnings Premium as the metric for return on investment. 

This metric represents the number of graduates from an undergraduate institution that earned a PhD in 2023 relative to the size of that school's undergraduate enrollment. It reflects the talent, preparation, and sustained academic achievement of the schools' alumni/ae. Data were obtained from the 2023 Survey of Earned Doctorates conducted by the National Center for Science and Engineering Statistics.

Graduation rate is an important first outcome of a college or university. Data were obtained from each school's most recent Common Data Set** and, with a few exceptions, reflect the Fall 2018 cohort of students.

We defined athletic excellence as the school's cumulative score in the Learfield Director's Cup Standings, which calculate a school's performance across its intercollegiate athletic program. Scores were relativized across NCAA Divisions (e.g., although Emory was ranked first in the D3 standings, its score was relativized against UTexas' better score in D1). For schools that have joint athletic teams (e.g., Pomona-Pitzer), we assigned each school an equal fraction of the total number of points.

Admission and enrollment likelihood at the best schools should not be unduly associated with family income. Socioeconomic diversity was calculated as the evenness of enrollment likelihood among 14 family income brackets, controlling for SAT scores. Data were obtained from Opportunity Insights. 

A few schools had missing data. For schools that did not report SAT scores, high school class rank, or number of instructional faculty with doctorate or terminal degrees in their most recent common data set (CDS), we used data reported in earlier CDS, when available, up to five years previously. 

 For the few schools that did not report class rank data even in earlier CDS, we assigned a score that was equal to the lowest score for this variable among comparable (i.e., size, region, public/private) schools that had equal or better median SAT scores.  

 For schools that did not report data on number of faculty with doctorate/terminal degrees even in earlier CDS, we assigned a mean score for this variable calculated from comparable (i.e., size, region, public/private) schools.

For colleges and universities that were not included in Third Way's Price to Earnings Premium tables, we assigned a midpoint (median) value for the variable to that school.  

We eliminated a school from consideration in our rankings if it had more than two missing variables that required estimated data.