NotesFAQContact Us
Search Tips
ERIC Number: ED259002
Record Type: Non-Journal
Publication Date: 1984-Apr
Pages: 35
Abstractor: N/A
Reference Count: N/A
An Application of Heckman's Correction for Specification Error.
Nandakumar, Ratna
Heckman's correction for regression in selected samples for predictive validity studies was applied to a large data file on 7,984 law school applicants. Data included ethnic group, sex, socioeconomic status, undergraduate degree, school, scores on the Law School Admission Test (LSAT), writing ability, undergraduate grade point average, and age. The final selection criteria were not known. Data on the 1,845 applicants who were accepted included year of entrance, sex, date of birth, undergraduate grade point average, grade point average for first year of law school, undergraduate college, LSAT score, writing ability, ethnic group, and age. The grade point average for the first year of law school was estimated and compared, with and without using Heckman's correction factor. Three different methods were used to choose the variables for student selection and grade prediction: (1) the same set of variables were used for both selection and prediction; (2) a subset of variables used for selection were used for prediction; and (3) two completely different sets were used. It was found that the relation between the variables used for selection and prediction could affect the accuracy of prediction. (Author/GDC)
Publication Type: Speeches/Meeting Papers; Reports - Research
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A
Identifiers - Assessments and Surveys: Law School Admission Test