Premium
ORAL PROFICIENCY TESTING: COMPARATIVE VALIDITIES OF INTERVIEW, IMITATION, AND COMPLETION METHODS
Author(s) -
Henning Grant
Publication year - 1983
Publication title -
language learning
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.882
H-Index - 103
eISSN - 1467-9922
pISSN - 0023-8333
DOI - 10.1111/j.1467-1770.1983.tb00544.x
Subject(s) - psychology , pronunciation , fluency , rasch model , convergent validity , grammar , discriminant validity , raw score , imitation , test validity , psychometrics , statistics , cognitive psychology , developmental psychology , linguistics , social psychology , mathematics , mathematics education , philosophy , raw data , internal consistency
Employing an initial sample of 143 adult Egyptian learners of English as a foreign language, the three oral testing methodologies of imitation, completion, and interview were compared for reliability and validity. Similarly, five components under each method, namely, raw score, fluency, pronunciation, grammar, and combined fluency‐pronunciation‐grammar ratings, were analyzed separately and in tandem. Multicomponent‐multimethod convergent and discriminant validities were determined. Stepwise multiple regression was computed using FSI‐like interview scores as the dependent variable. And Rasch latent trait calibration and tests of fit validity were computed for imitation and completion tests. Results indicated that the pronunciation component of the imitation method exhibited highest overall validity across all indexes. The FSI‐like component of the interview method ranked second and the fluency component of the imitation method ranked third. Comparison of the three oral testing methods across all components for all empirical validity indexes showed (1) imitation, (2) interview, and (3) completion methods to rank in that respective order in terms of available composite validity indexes. Regression analysis showed the FSI‐like interview to be primarily related to grammar skill from among 11 independent predictors examined. Advantages of Rasch measurement were noted for management and analysis of item data.