Niche Hardest to Get In Ranking Methodology
The Hardest to Get In ranking assesses the difficulty of admissions at traditional four-year colleges and universities in the United States. It uses data sourced from the U.S. Department of Education.
A high ranking in Hardest to Get In generally indicates that:
- The college has a very low acceptance rate
- Enrolled students scored very highly on the SAT/ACT tests
Colleges Assessed by this Ranking
At the time of calculation, our database contained records for 2,245 public and private, traditional four-year colleges and universities across the United States. For the purposes of this ranking, a "traditional" college is considered to be any accredited, non-profit post-secondary institution that primarily offers four-year degree programs (as opposed to two-year or less). Some colleges were not included in this ranking if: (1) they were not located in one of the 50 U.S. states, Puerto Rico, or the District of Columbia; (2) they had fewer than 100 full-time undergraduate students; or (3) they had insufficient data (see below). The final ranking results in 1,349 colleges receiving a numerical ranking.
Statistics obtained from the U.S. Department of Education represent the most recent data available, usually from either 2012–2013 or 2013–2014, as self-reported by the colleges.
The process used to compute this ranking was as follows:
- First, we carefully selected the factors listed above to represent a healthy balance between statistical rigor and practical relevance in the ranking.
- Next, we evaluated the data for each factor to ensure that it provided value for the ranking. (The factor needed to help distinguish colleges from each other and accurately represent each college.) Because there are different factor types, we processed them differently:
After each factor was processed, we produced a standardized score (called a z-score) for each factor at each college. This score evaluates distance from the average using standard deviations and allows each college's score to be compared against others in a statistically sound manner.
With clean and comparable data, we then assigned weights for each factor. The goal of the weighting process was to ensure that no one factor could have a dramatic positive or negative impact on a particular college's final score and that each college's final score was a fair representation of the college's performance. Weights were carefully determined by analyzing:
- Factors built from student-submitted survey responses were individually analyzed to determine a required minimum number of responses. After this, responses were aggregated. We logically have a higher degree of confidence in the aggregated score for colleges with more responses, so a Bayesian method was applied to reflect this confidence.
- Factors built from factual information were inspected for bad data, including outliers or inaccurate values. Where applicable, this data was either adjusted or completely excluded depending on the specific data.
After assigning weights, an overall score was calculated for each college by applying the assigned weights to each college's individual factor scores. This overall score was then assigned a new standardized score (again a z-score, as described in step 3). This is the final score for the ranking.
With finalized scores, we then evaluated the completeness of the data for each individual college. Depending on how much data the college had, we might disqualify it from the numerical ranking or from the grading process. Here is how we distinguished these groups using the weights described in step 4:
- How different weights impacted the distribution of ranked colleges;
- Niche student user preferences and industry research;
- Each factor's contribution to our intended goal of the ranking described in the introduction above.
Lastly, we created a numerical ranking and assigned grades (based on qualifications discussed in step 6). Here is how we produced these values:
- Colleges missing the data for 50 percent or more of the factors (by weight) were completely excluded. They did not qualify for the numerical ranking or a grade. Note: This exclusion occurred before calculation of the final z-score.
- Colleges that had all of the factors were deemed eligible for both a grade and a numerical ranking. Colleges that did not have all of the factors were not included in the numerical ranking and received a grade only.
- The numerical ranking was created by ordering each college (when qualified) based on the final z-score discussed in step 5.
- Grades were determined for each college (when qualified) by taking the ordered z-scores (which generally follow a normal distribution) and then assigning grades according to the process below.
Grading Process for This Ranking
While our ranking shows the Top 100 colleges, we use grades to provide the user some context to those rankings and also to provide insight into colleges that did not make the Top 100. It's important to focus on more than just the number in the ranking. Given the high number of colleges included in this ranking, there may not be a large gap between the 15th and 30th ranked colleges. In reality, both are exceptional colleges when compared to the total population of all colleges nationwide. Grades are assigned based on how each college performs compared to all other colleges included in the ranking by using the following distribution of grades and z-scores:
||1.96 ≤ z
||1.28 ≤ z < 1.96
||0.84 ≤ z < 1.28
||0.44 ≤ z < 0.84
||>0.00 ≤ z < 0.44
||-0.44 ≤ z < 0
||-0.84 ≤ z < -0.44
||-1.28 ≤ z < -0.84
||-1.96 ≤ z < -1.28
||-2.25 ≤ z < -1.96
||-2.50 ≤ z < -2.25
||-2.50 > z
Note that we intentionally did not assign a grade below D- to any colleges.
Of the 2,245 colleges analyzed, 1,349 received a numerical ranking. The top ranked college was Harvard University, which had a final score that was over three standard deviations above the mean college. Yale University, Princeton University, Columbia University, California Institute of Technology, Stanford University, and Massachusetts Institute of Technology all scored very highly, just slightly below Harvard.
For questions or media inquiries, please contact us.