Open AccessSTATISTICAL ISSUES RAISED BY TITLE IX REQUIREMENTS ON ADMISSION PROCEDURES 1
Author(s) -
Reed Cheryl Wild
Publication year - 1976
Publication title -
ets research bulletin series
Language(s) - English
Resource type - Journals
eISSN - 2333-8504
pISSN - 0424-6144
DOI - 10.1002/j.2333-8504.1976.tb01111.x
Subject(s) - selection (genetic algorithm) , regression analysis , regression , statistical analysis , psychology , statistics , computer science , mathematics , artificial intelligence
The purpose of this paper is to discuss statistical issues raised by Title IX requirements on selection and admission procedures. Literature is reviewed which indicates that prediction systems used for selecting students into college do frequently differ for males and females. Usually women's criterion scores are underpredicted using regression equations developed from combined groups of males and females. Major field of study, sufficiency of the predictor battery, and systematic differences in male and female samples are discussed as possible reasons for differences in regression equations. It is concluded that admissions procedures that use a single regression equation developed from a combined male and female group may be unfair to female college applicants. Methods for investigating and minimizing unfair selection procedures are discussed.