Open AccessEVIDENCE RELATED TO THE VALIDITY OF THE ENGLISH LANGUAGE PROFICIENCY TEST
Author(s) -
Bridgeman Brent,
Harvey Anne
Publication year - 1999
Publication title -
ets research report series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.235
H-Index - 5
ISSN - 2330-8516
DOI - 10.1002/j.2333-8504.1999.tb01805.x
Subject(s) - generalizability theory , psychology , test (biology) , active listening , language proficiency , construct validity , sample (material) , test validity , reading (process) , mathematics education , set (abstract data type) , achievement test , language assessment , psychometrics , developmental psychology , standardized test , linguistics , computer science , paleontology , chemistry , philosophy , biology , programming language , communication , chromatography
The validity of the English Language Proficiency Test was evaluated by considering six aspects of construct validity identified by Messick: content, substantive, structural, generalizability, external, and consequential. Analyses evaluated both the internal structure of the test and the relationship of the scores to external criteria. In one set of analyses, high school and college language teachers were asked to evaluate their students using the same five proficiency categories that are used in the test, and the teacher ratings were then compared to the test scores. Correlations were averaged over 15 classes in the college sample and over 32 classes in the high school sample. In the college sample, the averaged correlations, corrected for range restriction, were .50 and .57 for reading and listening respectively. In the high school sample the corresponding correlations were .68 and .71. For both the reading and listening proficiency scales, teachers tended to rate their students about one category higher than they were rated by the test.