Statistical methods for the determination of content validity

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Amos Omotayo Olangunju (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site:
Richard M. Jaeger

Abstract: Determination of the content validity of standardized tests is a central problem at all levels of education and in the professions. The problem investigated in this research was that of developing operational rules and statistical guidelines for estimating the content validity of standardized achievement tests. Rating and matching techniques were examined as alternative methods for eliciting judgments about the content validity of test items. These methods of eliciting judgments and newly developed quantitative indices of the content validity of test items and tests were used to validate the Greensboro Public Schools' Mathematics Promotion Standard tests for Grade 4. Rating items on the basis of their "Overall" quality and a matching method were found to be more accurate procedures for eliciting judgments than was a method that required rating items on a given set of dimensions of judgment; however, judges were more consistent in their judgments of test items over domains with the latter method. The assumptions and implications of using the newly developed indices of content validity with tests designed for criterion-referenced interpretations were discussed.

Additional Information

Language: English
Date: 1987
Achievement tests $z North Carolina $z Greensboro
Educational tests and measurements $z North Carolina $z Greensboro
Examinations $x Validity

Email this document to