skip to main content skip to footer

Content Factors in Sex Differences on Test Questions SAT

Author(s):
Donlon, Thomas F.
Publication Year:
1973
Report Number:
RM-73-28
Source:
ETS Research Memorandum
Document Type:
Report
Page Count:
20
Subject/Key Words:
Aptitude Tests, Content Analysis, Scholastic Aptitude Test (SAT), Scores, Scoring, Sex Differences

Abstract

SAT-Verbal and SAT-Math items were analyzed individually for content and for differences in performance by males and females. Subjects were 55,717 males and 47,082 females on SAT-M and 55,717 males and 47,083 females on SAT-V who took the test in May of 1964. Analysis was limited to the "large difference" items. In general, SAT-V items classed as dealing with World of Practical Affairs or Science were easier for males; items classed as Human Relations or Aesthetic-Philosophical were easier for females. Sentence completion material was balanced between males and females. Antonyms and analogies tended to favor females and males respectively, by approximately equal amounts, while reading comprehension favored females by the largest amount, .016 per item. The SAT-M item with the greatest advantage for males was +.2l. Only two SAT-M items favored females. One was a verbal problem, possibly more subject matter favorable to females, and the other a difficult algebra problem. SAT-M items were also analyzed by type: subject content; geometric figures; algebraic unknowns; and other. Females did relatively better on algebraic problems. The subject matter items gave males an average advantage of .102 as contrasted to a .042 advantage on the algebra problems. On data-sufficiency problems, males outperformed females by .0624; on multiple choice items, they outperformed females by .0744.

Read More