Please Don't Treat Ordinal Scale data like Continuous Data

Advanced Placement exams do have a scale [1, 5]. However, these values are represented on an ordinal scale, not on an intervallic or ratio scale.

See https://wbuchanan.github.io/nnerppDataVisualization/#/5/2.

The numeric types used by Python assume that the values used a fixed location for the value 0. This means this equality will hold:

x - y == (1 + x) - (1 + y)

Ordinal scale data does not fix the length/distance between integers so a comparable equality:

3 - 2 != 4 - 3

The issue is that the values only represent order (the inverse of rankings).