A Study On Fuzzy Synthetic Scoring Systemfor Non-English Major Postgraduates Spoken English Test In Southeast University

Owing to the subjectivity of Spoken English Test (SET), the present scoring approaches cannot eliminate the subjective factors as the spoken English ability of examiners, their mood and their understanding of the assessing criteria in its scoring. Therefore, the present study designs a fuzzy synthetic evaluation model for SET of Non-English Major Postgraduates (NEMP) in Southeast University (SEU) to improve the consistency and stability of SET scoring. This empirical study may present a way to reduce the subjectivity of the current scoring methods.This paper first reviews the related theories of SET and classification of SET methods. It also compares the advantages and disadvantages of the current scoring methods for SET. Then it introduces the Fuzzy Synthetic Scoring Method, which has the fuzzy evaluation theory proposed by Zadel as its basis. After that, the benefits and risks of this method and related researches on it are elaborated on.The present study constructs the Fuzzy Synthetic Scoring Model for SET of NEMP in SEU to examine the knowledge, skills and communicative ability of the students, and to maximally avoid the interference of extreme factors. Considering its operability, a series of evaluation standards were set.12 indexes were chosen and the corresponding fuzzy scoring index set was established. Then, the weight of each fuzzy scoring index was determined according to the Experts Consulting Method. Next, the scoring grades such as excellent, good, medium, and poor were set, and so did the corresponding fuzzy scoring grades set. Then, the study determined the membership function between each index level of fuzzy synthetic scoring of SET in SEU and fuzzy scoring index set, and built fuzzy relationship matrix. Finally, the result of examinees was obtained by synthetic calculation of fuzzy matrix. After building the model, the study chooses the score of student X randomly to have a trial application. And then it calculates the scores of all the examinees from the School of Management Science and Engineering. At last, it verifies the reliability and validity of the whole model.The result of the trial application indicates that, being easy to operate, with high reliability and validity, the newly-built model is quite promising in improving the efficiency and fairness ofscoring for SET of NEMP in Southeast University.