ERIC Number: ED294915
Record Type: Non-Journal
Publication Date: 1988-Apr
Gender Bias in SAT Items.
Loewen, James W.; And Others
Sex-related bias on the Scholastic Aptitude Test (SAT) was studied in a sample of 1,112 students in SAT coaching classes who took the SAT. Of these, 1,028 answered an additional questionnaire (Appendix A of this report) about high school grade point average, perceived abilities, and background. Almost all of the subjects were 11th graders (97.8%), with 0.7% sophomores and 1.5% seniors. Seven items on the verbal and 10 on the math section showed more than a 10% difference in the percentage of each sex getting them correct. An additional 22 verbal items and 16 math items showed a difference greater than 5% in favoring one sex. Other questions considered were methods of test construction that might reduce bias, the relationships between SAT scores and school performance, the role of test anxiety on SAT performance, and the effects of SAT scores on college choices. As other researchers have found, girls did less well than boys on the SAT, yet had higher school grades. Recommendations for sex-fairer tests include the following: (1) remove items having large response differences between the sexes unless balanced by other items; (2) manipulate mean scores on the verbal test so that males and females score equally; (3) ensure that verbal content on math items favors girls; and (4) make validity studies correlating test scores and first-year college performance widely available both to consumers and researchers. Nineteen tables and two graphs are included. (SLD)
Publication Type: Speeches/Meeting Papers; Reports - Research; Tests/Questionnaires
Education Level: N/A
Sponsor: Women's Educational Equity Act Program (ED), Washington, DC.
Authoring Institution: N/A
Identifiers - Assessments and Surveys: SAT (College Admission Test)