Law School Admissions Survey with Binarized Outcomes, U.S. 1991

Survey data from students attending U.S. law schools in 1991, obtained from the R package fairml. The response variable indicates whether a student's undergraduate GPA exceeded 3.0, and race information has been binarized into white and non-white categories.

Use Cases

• Train binary classification models to predict high undergraduate GPA based on survey responses.
• Analyze demographic disparities in law school admissions outcomes based on the binarized race variable.
• Benchmark fairness-aware machine learning algorithms using the pre-processed target variable.
• Study the relationship between survey variables and academic performance in graduate education.

Strengths

• Data originates from a specific 1991 survey of U.S. law school students, providing a focused temporal snapshot.
• Key variables like undergraduate GPA and race have been pre-processed for binary classification tasks.
• The dataset is sourced from the established R package fairml, which is used for fairness research.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale modeling.
• Data is from 1991 and may not reflect current admissions trends or demographics.

Provenance

Source

R package fairml

Collection Method

Survey

Time Range

1991

Geography

United States