Law School Admissions Survey (1991) with Binarized Outcomes

A survey of students attending law school in the United States in 1991. The dataset was obtained from the R package fairml and has been modified for binary classification tasks. The response variable indicates whether a student's undergraduate GPA was greater than 3, and race information has been binarized into white and non-white categories.

Use Cases

• Train binary classifiers to predict high undergraduate GPA based on survey responses.
• Analyze demographic factors, such as binarized race, in relation to academic performance.
• Benchmark fairness-aware machine learning algorithms using a known dataset from the fairml package.

Strengths

• Data originates from a 1991 survey of U.S. law school students, providing a specific temporal and demographic context.
• Variables have been explicitly prepared for binary classification, with a defined target (UGPA > 3) and a binarized race feature.

Limitations

• Row count and column-level documentation are unknown, which limits suitability assessment and requires manual inspection after download.
• The dataset is from 1991 and may not reflect current admissions trends or demographics.
• Description metadata is limited; actual data quality and completeness require verification.

Provenance

Source

R package fairml

Collection Method

Survey

Time Range

1991

Geography

United States