Files in this item



application/pdf8701567.pdf (5MB)Restricted to U of Illinois
(no description provided)PDF


Title:Item Bias in the Verbal Segment of the Scholastic Aptitude Test (Sat) for High School Students in the United States Virgin Islands
Author(s):Michael, Noreen
Department / Program:Education
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Education, Tests and Measurements
Abstract:This dissertation reports the findings of an item bias study. The study was undertaken using the modified three parameter item response theory (IRT) technique first proposed by Linn and Harnisch (1981) and recently validated by Shepard, Camilli, and Williams (1985). The population of interest was high school juniors and seniors attending public schools in the United States Virgin Islands. There were two components to the study: (a) a quantitative item analysis using the item bias detection technique indicated above, and (b) a qualitative item analysis of the items identified as biased by the modified IRT technique. The qualitative item analysis was based, in part, on an approach discussed by Carroll (1979).
Results of the quantitative item analysis indicated that analogy, antonym, and reading comprehension items were biased for the group of interest. These item types varied with regard to the degree of bias associated with them. In addition, the statistical findings indicated that within this group there exists item x ability level interaction. One sentence completion item was found to be biased in favour of the U.S. Virgin Islands students. It should be noted that several items found to be biased were observed to be too difficult for the students and thus provided little information about the group as a whole.
Results of the qualitative item analysis indicated that items found to be statistically biased differed in type and substance. This analysis also demonstrated that some of the items identified as biased were extremely difficult for the sample of concern and this limited the information that could be obtained from student's responses to such question. The qualitative item analysis also indicated that there might be various probable explanations for the performance of the USVI sample on the test administered. The results point to a need for further research in the area item bias study for the population of interest using the same instrument as used in this study.
Issue Date:1986
Description:158 p.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 1986.
Other Identifier(s):(UMI)AAI8701567
Date Available in IDEALS:2014-12-15
Date Deposited:1986

This item appears in the following Collection(s)

Item Statistics