英语论文网

p;2002).  In  this  case,  linguistic  studies  on  business  discourse  and  international business  communication  are  both  in  its  research  field;  moreover,  its  teaching  goals contain developing language ability, and cultivating international business capability as well.  
..........

2.2 Business Purpose Language Ability and Business Knowledge
Based  on  linguistics  and  psychometrics,  language  testing  is  a  research  field  with  its purpose  of  developing  language  tests  and  evaluating  language  competence.  However like all targets in psychometrics, language ability cannot be measured directly. A proper way to solve this problem is judging it through test performance during examination. In order  to  evaluate  language  competence  correctly,  linguists  have  to  find  connection between  language  performance  and  competence,  in  other  words,  they  need  to distinguish the related and observable performance and integrate it with corresponding levels of language competence. Therefore, competence construct is the prime thing in language assessing.  
...........

CHAPTER THREE   RESEARCH METHOD ......... 25 
3.1 Research Questions ......... 25 
3.2 Research Materials .......... 25 
3.3 Research Tools ...... 26 
3.4 Data Collection ..... 27 
3.5 Analyzing Procedures ..... 29 
CHAPTER FOUR   RESULTS AND DISCUSSION ......... 32 
4.1 Rater Reliability Test ....... 32 
4.2 Descriptive Statistics on Writing Scores and Term Distribution........... 37 
4.3 Correlation Analysis ........ 40 
4.3.1 Independent Tests ............ 40 
4.3.2 Correlation Tests ............. 43
4.4 Discussion ............. 48 
CHAPTER FIVE   CONCLUDING REMARKS ..... 53 
5.1 Summary of Major Findings ..... 53 
5.2 Implications for Teaching ......... 55 
5.3 Limitations of This Study.......... 56 

CHAPTER FOUR  RESULTS AND DISCUSSION  

4.1 Rater Reliability Test 

Bachman (1990) points out that reliability is the premise of validity in language testing, that is to say, test validity should be based on its reliability. In writing scoring, reliability can be more difficult to guarantee, because writing scores mainly depend on subjective judgment  from  raters.  To  eliminate  disturbance  of  raters'  subjectivity  from  writing assessing in this study, a test to ensure reliability of score data before further analysis is necessary.  Multi-Facets  Rasch  Model  (MFRM),  widely  applied  in  the  field  of psychological  and  educational  measurement  for  analyzing  subjective  assessing,  is considered  as  a  good  choice  of  examining  rater  reliability  (Weigle,  1998;  Chalhoub-Deville and Wigglesworth, 2005; Liu, 2007). The i本论文由英语论文网提供整理，提供论文代写，英语论文代写，代写论文，代写英语论文，代写留学生论文，代写英文论文，留学生论文代写相关核心关键词搜索。