Since the reliability coefficient of the semantic domain SBCMCIT is .82 in the present study, the low reliability coefficient obtained by Gholami can be attributed to her small sample and the test's being the most difficult among the SBCMCITs as reflected in its mean, i.e., 20.3. The significant correlations, however, disappear when proficiency levels are considered separately, implying that it is the language proficiency level measured by the SBCMCITs which entails the employment of certain cognitive styles rather than vice versa. Since low proficiency learners have little knowledge of the semantic schemata to comprehend the passage, their cognitive styles fails to play any significant role on the SBCMCITs. In other words, the application of both FD and FI cognitive styles by middle proficiency learners of English explains approximately eight percent of their performance on the semantic domain SBCMICT. Since the mean score of the middle proficiency group is 10.12 on the GEFT, it indicates that half of the test takers in this group employ their field dependency to find the keyed response on the SBCMCIT while the other half utilize their field independency to fulfill the same function. The results showed that the SBCMICT is a fair measure of language proficiency because the performance of neither low nor high proficiency test takes' on the SBCMCIT showed any significant relationships with the GEFT as a widely employed measure of field dependency (FD) and independency (FI) cognitive styles. Since choosing the keyed response from among the three syntactically, semantically and discoursely related competitives on the SBCMCIT does require focusing not only on the competitives themselves, i.e., field independence, but also on the context in which the keyed response appears, i.e., field dependency, the performance of low and high proficiency test takers on the SBCMCIT and GEFT do not show any significant relationships with each other.
CITATION STYLE
Khodadady, E., Fatemi, A. H., & Etminan, S. (2012). Cognitive Styles and Performance on Schema-based Cloze Multiple Choice Item Tests: A Fairness Issue. Journal of Language Teaching and Research, 3(4). https://doi.org/10.4304/jltr.3.4.806-813
Mendeley helps you to discover research relevant for your work.