A Comparative Study Of Objective Items In The Reading Of NMET

Testing is a means of measuring the outcome of teaching. Learning and testinghave a deep influence on each other. The outcome of teaching and learning could bemeasured much more effectively if the testing structure and content are scientificallydesigned. Language testing is deeply influenced by teaching approaches and methods,such as, structuralism, the grammar-translation approach, and communicativeapproach and so on. So the scientific construction and evaluation of language testingare critical for electing the qualified candidates to further studies in colleges or workin some organizations.The scores of NMET is a kind of scientific statistic for test users to elect seniorhigh school candidates to universities. Being a key element for evaluation in NMET,there are four objective items and two subjective items in reading of NMETs. In thisstudy, the study focuses on researches about the four objective items in the reading ofNMETs. Test paper used in this study is abstracted from nine NMETs of2011, and itcomposes of four objective items of reading. All these four objective items appearingin reading all belong to types of choices selection, such as matching, dialogue filling,paragraph filling, and traditional multiple choices (a key answer in four options). Thecandidates are freshmen of English major in the School of Foreign Languages.The study utilizes the combination of quantitative and qualitative analysis. Then,the collected data have been presented and analyzed by SPSS16.0. The study findsthat the discrimination and difficulty of dialogue filling is the less obvious than the other three objective items through the correlation analysis of participants’ scores onNMET and on each objective item. The discrimination and difficulty of MCs (fouroptions) and paragraph filling are much more obvious. Besides, both the correlation ofcandidates’ scores on NMET and on MCs and the correlation of candidates’ scores onNMET and on paragraph filling are obvious. It signifies that both MCs and paragraphfilling can reflect candidates’ reading capability better. What’s more, the qualityanalysis of these four objective items is based on Bachman’s theories on the quality oflanguage testing. The reliability and construct validity of MCs and paragraph fillingare good, yet the authenticity and interactiveness of matching and dialogue filling aremuch better than MCs and paragraph filling.