چکیده:
This study aimed to investigate differential item functioning (DIF) on the English subtest of the Iranian National University Entrance Exam (INUEE). The effect of field of study (math, humanities, and sciences) was taken into account. The study utilized one-parameter IRT model to investigate DIF for a sample of 18000 students randomly selected out of a population of 219579 students who sat for the INUEE in 2005. The results of the study identified thirty three DIF items on different sections of the English subtest. It was also found that all DIF items in the vocabulary, and word order sections worked to the advantage of students with sciences or humanity background, whereas in grammar, language function, doze test, and reading comprehension sections, approximately half of the DIF items were in favor of math students and half in favor of students of sciences and humanities
خلاصه ماشینی:
It was also found that all DIF items in the vocabulary, and word order sections worked to the advantage of students with sciences or humanity background, whereas in grammar, language function, doze test, and reading comprehension sections, approximately half of the DIF items were in favor of math students and half in favor of students of sciences and humanities .
Key Words: , Differential Item Functioning (DIF), Field Of Study, Iranian National University Entrance Exam (INUEE), English Subtest, IRT Introduction Language testing practitioners have always tried to construct a set of items which provides an estimate of a test-taker's ability and is as fair and accurate as possible to all groups of the population.
The results indicated that the grammar, language function and doze test sections were more or less the same for students of different fields of study; that is, some items favored math students and some humanities and sciences students; whereas the vocabulary and word order sections were to the advantage of students of sciences and humanities.
This idea is supported by the observation that the DIF items of the grammar section which worked to the advantage of math students were especially related to nuances of grammar; for example, application of appropriate verb forms in conditional sentences, articles, etc.
It was found that all the items in the vocabulary, and word order sub-tests which showed DIF worked to the advantage of sciences and humanities students whereas almost half of the DIF items in the grammar, language function and doze sections worked to the advantage of math students and half to the advantage of students of sciences and humanities.