Telling the rest of the Story

Language ability, especially foreign language proficiency, has become a hot topic in language testing research. Among the models and hypotheses proposed to account for and explain foreign language proficiency, two major hypotheses have been the most widely cited and investigated in the literature. Harris's separate abilities hypothesis claims that foreign language proficiency is composed of different separate abilities, each concerned with a modality of the foreign language skills. Consequently, discrete point tests were proposed as the optimal measurement tools to assess the different skills and modalities of language proficiency. Oller, in his unitary competence hypothesis (UCH), on the contrary, deemed that there is one general factor underlying the various features of the foreign language in their different modalities. Hi proposes integrative of pragmatic tests to be he best choice in assessing language proficiency. Oller and other researchers who empirically supported either hypothesis used subjects with moderately high proficiency in English ranging from 450 to 550 in the TOEFL. This study, however, identified a stage in language proficiency prior to those reported in the literature. The subjects involved came from four different proficiency levels with an average score in the TOEFL ranging from 232 to 388. In addition, two types of data were used in his study: subjects scores on the four different skill tests of English as well as the achievement test in grammar in addition to their scores in the TOEFL. Among the major findings of this study were: although the different proficiency and achievement tests loaded heavily on the general factor, there were also significant loadings on other factors. The achievement reading comprehension test of the preparatory students loaded significantly on a second factor. In addition, the achievement speaking and listening tests of the intermediate students loaded significantly on a third factor. The achievement tests on the different factors were heavier than the loadings of three section of the TOEGL. This disparity could be ascribed to the difficulty level of the TOEGL although it seems that as subjects of this stud become more proficient in English, similar skills tests and TOEGL sections tests will load on the same factors, thus resembling subjects in other studies reported in the literature. In addition, the results of this study explain why the correlations between the different skill tests and the integrative tests (for example, cloze testing procedure, c-test, and multiple- choice cloze test) are almost of similar values in low proficiency level: whereas, the correlations between the integrative tests and the tests of reading comprehension are higher than the correlations between the integrative tests and the other skills tests in high proficiency levels this study concluded by proposing a definition for foreign language proficiency based on a viewpoints which regards it as a continuum with different point on that continuum representing different proficiency levels. The use of both discrete point tests and integrative tests while assessing foreign language proficiency is more profitable than using either approach per se