Abstract：National Matriculation English Test （NMET） is very important in China.Listening test of it in Sichuan was recovered in 2016.And there is no study to investigate the validity of this test.Hence this paper will analyze the listening test in Sichuan 2017 NMET from the aspect of content validity.

Key words：NMET；listening test；content validity

1 The framework of Content Validity

Content validity means that the correspondence of test specification and test content.Based on the framework of Bachman and Palmer，the research of Li （2016），and the requirements of specifications，the content validity of target test will be elaborated from four dimensions，which can be seen in table 1.

1.1 The Characteristics of Setting

From 2001，the listening test in NMET began，so it has a long history in China.Hence candidates are already familiar with this test.And during NMT，whole city will be required to “keep quite” for candidates.Also，educational department have already thought of many ways to guarantee the fairness in NMT.So，the setting of this test satisfy the requirements of content validity.

1.2 The Characteristics of Rubrics

The instruction of target test is Chinese，which is easy to understand.It contains the ways to write answers，the time allotment and so on.The structure of target test can be seen in table 2.All the question types are multiple choices.Although it will benefit the scoring，the format of listening part may be too simplex.

1.3 Expected input

1.3.1 Length

From table 3，the difference between two standard deviation shows us the tendency of dispersion of these two kinds of conversations.And the difference between the length of conversation 9 and 10 is huge.

1.3.2 Vocabulary range

There are only 6 new words in this test.The proportion of it is 0.61%.So，though there were new words，the proportion is small，which will have little influence for candidates.

1.3.3 Topics

According to the 24 topics mentioned in specification，the correspondent topics of each part will be presented in table 4.From this table，half of topics are daily routines .It failed to satisfy the university of topics in listening test.

1.3.4 Genre

Listening part only has one passage，and it is an expository.

1.3.5 Speed of Listening

Based on Syllabus （2003），the speed of listening test should be about 110 w/m`.For higher goals，the speed should be ranged from 110 to 120 w/m`.

From table 5，questions 6 to 10 are more difficult for the faster speech rate.However，NMET aims to select talents.So，it is reasonable here.At the same time，the standard deviation here shows that the difference of speech rate in each short conversation is larger.And the speech rate of these 9 conversations and 1 monologue is too fast.

1.4 Expected Response

1.4.1 Language Abilities

Based on specification，the language abilities in 2017 NMET should include： main idea，details，inference，intention and so on.From table 7，the language abilities of understanding the main idea did not involve in this listening test.And short conversation did not include the language abilities of acquiring the details.

1.4.2 Types of Response

The types of response mentioned by Palmer and Bachman involves in multiple choices，restricted and productive response.In this listening test，it is obvious that the types of response are all the multiple choices.

2.Conclusion

The target test is basically satisfy the requirements of content validity.However，for the aspects of types of response，length of longer conversation，topics，speech rate of listening part，and the language abilities about understanding the main ideas should be reconsidered.

References

[1]李艷萍.高考英語聽力理解內(nèi)容效度分析——以2015上海卷為例 [碩士學(xué)位論文][D].上海師范大學(xué)，2016.

[2]中華人民共和國教育部.普通高中英語課程標準（實驗）[Z].2003.

作者簡介

熊雪華（1994—），女，漢，四川達州人，碩士在讀，單位：四川師范大學(xué)；研究方向：學(xué)科教學(xué)（英語）。