Methodology Note: To bring the four metrics together, we created a separate index for each metric. The school with the best measurement received an index score of 100. Every other school’s index score was then measured against the best to come up with an index score for that metric. For applications per seat, for example, Stanford led all schools with 17.9 application for every available seat in the class. Stanford was awarded an index score of 100. MIT which had the second best number, 11.7, was then divided by 17.9 which resulted in an index score for MIT of 65.0, or 65% of Stanford’s 17.9 applications. Then all the index scores across the four metrics were added together and an index was built on those numbers which account for each school’s numerical rank.
The benefits of this method are obvious. Schools that have significant, and statistically meaningful, differences in any one of the four metrics get full credit for that difference. Typically, ranking organizations do a simple rank order of such numbers from 100 to 0 which fail to take into account those differences. This system also prevents less meaningful differences in the data from having a bigger overall impact.