testspecification考試規(guī)范

1、A testDefin iti on of test sp ecificati onss sp ecificati ons pro vide the official stateme nt about whatthe test tests and how it tests it. The sp ecificati ons are the blue printto be followed by test and item writers, and they are also esse ntial inthe establishme nt of the tests con struct valid

2、ity.Differe nee betwee n sp ecificati ons and syllabusa testsp ecificati on is a detailed docume nt, while thesyllabus is ofte n muchsimp lified.2.a testsp ecificati on is a internal docume nt for testdevel op ers,sometimescon fide ntial to the test users, while thesyllabus is ap ublic docurae nt wh

3、ich in dicates to can didates andteachers whatthe test will contain.II Who n eeds test sp ecificati ons? And why?Test sp ecificatio ns are n eeded by a range of differe nt people:n eeded by those who p roduce the test itselfTest con structors n eed to have clear stateme nts about who thetest is aime

4、d at, what its purp ose is, what content is to be covered,what methods are to be used, how many papers or secti ons there are,how long the test takes, and so on.In addition,the sp ecificati ons willn eed to be available to those res pon sible for edit ing and moderat ingthe work of in dividual item

5、writers or teams. Such editors mayop erate in a committee or board officials. In smaller in stituti ons.they may simply be fellow teachers. The sp ecificati ons should becon suited and clearly writte n.Needed by those res pon sible for or in terested inestablish ing the testalidity (that , whether t

6、he test tests what itis suppo sed to test).Validators may think it is less imp orta nt to havein formatio n about thelen gth of the test and its sect ions, and moreimp orta nt to know thetheoretical justificati on for the content: whattheories of Ian guage and p roficie ncy make the test, andwhy the

7、 testis the way it is.Needed by test usersTest users n eed descri pti ons of a tests content, and differe nsorts of users may n eed somewhat differe nt descri ptions.Forexa mp le, teachers who will be res pon sible for the lear ners p laced in their classes by a test n eed to know what the test scor

8、es mea n: what the p articular lear ners kn ew, what they can do, what they n eed to learn. Teachers who wish to en ter their stude nts for some p ublicexam in ati on n eed to know which test will be most approp riate for their lear ner in relatio n to the course that they have bee n followi ng.Need

9、ed by p ublishersFin ally, test sp ecificati ons are a valuable source of in formatio n for p ublishers wishi ng to p roduce textbooks related to the test: textbook writers will en sure that the p ractice in books they p roduce are of an approp riate level of difficulty, with approp riate content, t

10、opi cs, tasks and so on.ons.All these users of test sp ecificati ons may have differ! ng n eeds, and writers of sp ecificati ons n eed to bear the audie nee in mind whe n p roduci ng their sp ecificati What is suitable for an audie nee may be quite un suitable for ano ther.Ill Contents of test sp ec

11、ificati onsTest purpose2.Theoretical con struct and op erati onsTest tasks and text types3.Testtakers4.Testcontent5.Testmethod7.Test orga ni zatio ns and weight ingTime durati on9.Scori ng criteri on10.Test adm ini strati onTest devel opment chartIV How to draw up test sp ecificati ons?ons(Test purp

12、o se: tests tend to fall into one of the follow ingbroad categories: p laceme nt, p rogress, achieveme nt, p roficie ncy anddiag no Stic.PI aceme nt tests are desig ned to assess stude nts ability so that they can be pl aced in the approp riate course or class.level of laiP rogress tests are give n

13、at various stages throughout a Ian guage course to see what the stude nts have lear nt.Achieveme nt tests tend to be give n at the end of the course. The content of both p rogress and achieveme nt tests is gen erally based on the course syllabus or the course textbook.course.Proficie ncy tests are d

14、esig ned to test the ability of stude nts with differe nt Ian guage training backgro un ds. Some are inten ded to show whether stude nts have reached a give n level of gen eral la nguage ability; others are desig ned to show whether stude nts have sufficie ntne,tourismability to be able to use a Ian

15、 guage in some sp ecific area such as medici etc.Diag no Stic tests seek to ide ntify stre ngths and weak nesses instude ntsIan guage ability.Test purp ose is the no rraal start ing point for desig ning test sp ecifications.For exa mple:Test A is a pl aceme nt test, desig ned to pl ace stude nts app

16、lying for Ian guage courses at En glish into classes approp riate to theirIan guage level.Test B is inten ded to diag nose stre ngths and weak nesses of fourth year sec on dary school pup ils in En glish grammar.Test con struct and op erati onsEvery test has a theory behi nd it: what Ian guage abili

17、ty con sists of, what la nguage lear ning invo Ives and what la nguage users do withIan guage. Every theory contains con structs> which are its principal components and the relati on shi p betwee n these componen ts. Test sp ecificati ons n eed to make the theoretical framework which un derlies t

18、he test Ian guage skills and relati on shi ps betwee n its con structs, as well as the relati on shi p betwee n the theory and what the test is desig ned for. The test doma in is inten ded not to in elude allthe skills of the con struct but the rep rese ntative contents of the list ing skillsFor exa

19、 raple:Heato n (2000: 105-106), in his book"Writing EnglishLan guage Tests" ,hypo thesizes read ing skills as the ability to:recog nize words and word groups, symbols;associati ng sounds with their corres ponding grap hiededuce the mea ning of wordsun derstand exp licitly stated in formati

20、o n;un derstand relati ons with in the sentence;un derstand relati ons betwee n p arts of a text;p erceivetemporal and sp atial relati onships and also seque nces of ideas:un derstand concep tual mea ning：a ntici pate and p redict what will come next in the text;ide ntify the mai n idea and other sa

21、lie nt features in a text;ge neralize and draw con clusi on：un dersta nd in formati on not exp licitly stated by making refere nces;skim and sea n (look ing for the gen eral mea ning and read ing for sp eeifie in formatio n);read critically;ad opt a flexible app roach and vary read ing strategies ac

22、cord ing to the type of material being read and the purp ose for which it is being read.Test takersWhat sort of lear ner will be tak ing the test.Personal characteristics:age, sex.first la nguage, country of origi n.level and n ature of educatio n, reas on for tak ing the test, pro fessi onal in ter

23、estsTop ical kno wledge: levels of target Ian guage social and cultural un dersta nding.Lan guage ability: levels of target Ian guage kno wledge andIan guage use abilityTest contentTopic is concerned with what the test is ready to measure, grammar or structure kno wledge or writi ng, sp eak ing and

24、read ing abilities.Focus is invo Ived in what fun cti ons test tasks should en able stude nts to use accordi ng to the above topic.Descri pti ons of test content should be detailed and corap rehe nsive: test content should dem on strate that it tests what the test is supp osed to test or test conten

25、t validity; test content should be rep rese ntative for test con struct (con struct validity).Test methodWhat test methods are to be used 一mult iple choice, gap filli ng.match ing, tran sformati on, short an swer questi on, pi cture descri pti on, role p lay with cue cards, essay, structured writi n

26、g, etc?In tak ing test typ es, test writers should think about four asp ects: test validity, reliability, authe nticity and p racticality.Test tasks and text typesWhat sort of tasks are requireddiscrete poin t, i ntegrative.authe ntic, objectively or subjectively assessable? If it is authe ntic.what

27、 text types should be chose nwritte n an d/or sp oke n? Whatshould be the sources of these, the top ics, the degree of authe nticity?embodied in the textsHow difficult or long should they be? What functions should bep ersuasi on, defi niti on, summariz ing, etc?How comp lex should the Ian guage be?o

28、n?Test orga ni zati ons refer to nu mber of items or tasks items are required for each secti on? Weight! ng concerns what is the一how manyrelative weight for each item -equal weight! ng, extra weight! ng for more difficult items.For exa mp le:weight ing for writte n CET4新題型的四級考試主要由聽力理解、閱讀理解、完型填空或改錯、寫

29、作和翻譯四部分組成。其中聽力理解部分分值比例為35%閱讀理解為35%完型填空或改錯部分為10%寫作和翻譯部分分值比例為20%其中寫作部分15%翻譯部分5%從各項所占分值 來看，即聽力249分，閱讀249分，完型填空或改錯70分，作文142分，各單項分相加之和為 總分710分。Time durati onHow many sect ions/papers should the test have, how long should they be and how will they be differe ntiated?For exa mp le:the time differe ntiati o

30、n of CET4PartI Writi ng (30 min utes)PartII Readi ng corap rehe nsion ( skim ming and sea nning) (15mi nu tes)PartIII Liste ning comp rehe nsion (35 minu tesPartIV Read ing comp rehe nsion (Readi ng in dep th) (25 min utes)PartV Cloze (15 mi nutes)Part VI Tran slati on (5 min utes)Scori ng criteri o

31、nWhich criteria will be used for assessme nt by markers? How imp orta nt is accuracy, approp riacy, sp elli ng, le ngth of uttera nee/scri pt, etc?A descri pti on of the criteria which will be used to evaluatep erforma neeCriteria for assessme ntstude ntsThe higher mark allocati on reflects its grea

32、ter orga ni zati onal and in terpretati onal dema nds.An swers of about 300 words will be required. As can didatesiece of writi ng which in terprets the in format! on pro vided, it is clearly imp avoid any atte mpt to describe exte nsively all the in formatio n p rovided.are required to devel op a logical p orta nt that theyCan didates should be able to compare and con trast, show cause and effect relati clusi ons, make hypo theses a