چكيده
The purpose of the present study was to investigate the relationship between teacher- made test and PNU final examination for the aural course of the translation students,which were intended to measure oral ability of the students, together with a detailed analysis of the PUN final written test items in terms of their item facility, choice distribution, item discrimination, and estimating validity and reliability of the total final written test. Sixty English majors, dominantly females, from Payame Noor University of Rasht took part in this study. The result of Pearson product- moment correlation showed that there was a strong positive correlation between oral teacher- made test scores and PNU written test scores (r= 0.709). Moreover, Crobach’s alpha statistic estimated for the written test showed an acceptable reliability index of 0.78. The results of item analysis of the 50 PNU written test multiple choice items revealed that 88% of the items were too easy, 4% were too difficult, and only 8% of the items were normal items. In terms of item discrimination, only 12% of the items had acceptable discriminatory power but 88% of the items were not acceptable. Besides, in terms of choice distribution, 48% of the items had distracters which did not work properly. Item analysis confirmed that the written test definitely needed revision. Comparing content specification and test content revealed that PNU written test lacks content validity specially in terms of the number of the items, unequal distribution of the items from two books and throughout the units, the example items, representative tasks, representative sample, the importance or weight given to certain elements, and learning level of the testees. The finding of the study implied close consideration of test content and test specification in the process of test making by test developers thorough the objectives of the course.
