Datasets and Software for Detecting Algorithmic Discrimination
Student’s SAT® scores in Mathematics with the performance of other test-takers in the total group of 2014 college-bound seniors, as well as with all males and all females in the cohort. Percentile ranks show the percentage of each group who scored lower on the three sections together. Note that composite scores are not the best way to view SAT scores because important differences between the performances on each section are obscured.
Data Types: Multivariate
Default Task: Classification
Attribute Types: Categorical, Integer
Format: pdf, csv
Number of Entries: 1,672,395 (2014) / 1,698,521 (2015)
Number of Attributes: 2
Year: 2014
Source: CollegeBoard