You are on page 1of 6

123 N o.

123
2008 9 CAFLEC Sep. 2008

* *


1 2
,
( 1. , 200083; 2. , 201209)

: ,
,
8, ( CEM ) ( BNC)
, ,

: ; ;
: H 319. 3 : A : 1001-5795( 2008) 05-0010-0006


1

1. 1 ,
,
, ,
( frequency)


( H unston, 2002: 96 )
, ,
( ,
, 2002: 45) :
, ; ;
1996 A ld erson( In T hom as& Shor,t /
254) 1. 2
,
( H unston, 2002: 205, , 2007;
, 2005) ,
, : :
( ETS)
( ,
) , , ,

: : , , , :
: , , :
: 2008-02-10
: 5 - 6 ;

# 10#
, :

:
( UCLES)
, ,
: ( )
: ,
, Rees( in H unston, 1998: 205) 1. 4 8
, 8 ( 2004 ) , 8
: /
) ) ) ;
) ) ) ; 0 ( 2004: 1-
) ) ) ; 2), 8 /
) ) ) 0 ( 2004: 2 ) 8 ( Proo-
f
Rees, reading & Error C orrection) /
, ( 2005) CLEC ( Chinese L earner


Eng lish Co rpus) BNC 0 250
CLEC , 10 10
, / 0/ 0 /
CLEC 0
,
, BNC 1997

1. 3 ,
, / 0,
( test validity) ,
A lderson ( In Thom as ,
and Shor,t 258), /
0 ,
, A lderson ,
, , ? ?
, , ?
?
,
Bachm an( 1990: 244) , :
( conten t relevance) ( con tent
2
coverage) /
/ 2. 1
: ( CEM )
( pre- test) ( pos-
t test) ( BNC) , 8
:
( test specifica tio ns) ( CEM ), ( ) ( typ-
icality); ( BNC ) ,
( A ld erson, 1995: 173-5): ;

# 11#
, :

2. 2 CEM , 2000 2002 30


W o rdSm ith Too ls 12 (
version 4, SPSS13. 0 16 ) , 75% 30
( CEM ) 2400 ( 66 ) [ fm ] [ pc] ( ) [ cn]
( BNC) : CEM ( ) [ cp]
, BNC ( 30) ,
2. 3 , 2:
20002001 2

2002, 30 (, 2005) ( % )

2. 4 pp 20. 0 1

: ( wd 16. 7 2


)? cj 13. 3 3

vp 13. 3 3
?
sn 6. 7 4
?
ad 6. 7 4

3 cc 6. 7 4

aj 3. 3 5
3. 1 np 3. 3 5
CEM ( ar 3. 3 5

, 2008), ( tn 3. 3 5

30 ) , 1: pr 3. 3 5

12 100. 0
1

( % )
,
2000: No. 2;
cj 2001: N o3; 4 13. 3
, 20% ; ;
2002: N o. 1, No. 5 ;
2000: No. 3, N o. 7, No. 8, N o. 9
, ;
pp 2001: N o. 1 6 20. 0

2002: N o. 2 ,
sn
2000: N o. 4
2 6. 7
3. 2 ( CEM )
2001: N o. 7
,
aj 2000: N o. 5 1 3. 3
W ordSm ith T oo ls( versio n 4)
np 2000: No. 10 1 3. 3
( C Concord), CEM 48
ar 2000: N o. 1 1 3. 3
12 2400 66
2001: N o. 6, No. 9
vp 4 13. 3 ,
2002: N o. 3, No. 9
12 3
2001: No. 4, N o. 5, No. 8
wd 5 16. 7
2002: N o. 4, No. 7 : 66 12
tn 2002: N o. 6 1 3. 3 32426
2001: N o. 2 , ( 32426)
ad 2 6. 7
2002: No. 10 , , 4
pr 2002: N o. 8 1 3. 3
cc
2000: N o. 6
2 6. 7
,
2001: No. 10
, 27. 36% ,
12 30 30 100. 0
, 0. 8%

# 12#
, :

3 ( 3) full authority 22;


(% ) ( 4) absolute authority 5;
cj 1372 4. 23
( 5) comp lete auth ority 2;
pp 1640 5. 06
3. 4 SPSS
sn 2367 7. 3
CEM
aj 260 0. 8

np 3903 12. 04
, ,
ar 3063 9. 45 5:
vp 3592 11. 08 5 Co rrelations
wd 8871 27. 36
D esigned E rror C orrela tion C oe ff-
i
tn 4272 13. 17 Spearm anps rho 1. 000 . 117
F requency cient
ad 504 1. 55
S ig. ( 2-tailed) . . 717
pr 1363 4. 2
N 12 12
cc 1219 3. 76
C orpus E rror C orrela tion C oe ff-i
12 32426 100. 0 . 117 1. 000
F requency c ien t

S ig. ( 2-tailed) . 717 .


4
N 12 12
(% ) : 0. 05
wd 27. 36 1

tn 13. 17 2 , ,
np 12. 04 3
vp 11. 08 4

ar 9. 45 5
4
sn 7. 3 6
, :
pp 5. 06 7
,
cj 4. 23 8
8
pr 4. 2 9

cc 3. 76 10
,
ad 1. 55 11 ,
aj 0. 8 12


12 100. 0

,
,
3. 3 ( BNC) CEM , 75%
30 : 2000 , ,
2001,
, , :
: 2000: sing le veh icle sole vehicle; 2001: ,
tota l au th ority full / com plete / abso lu te author ity ,
, ( typicality )?
BNC 2127, / 0
: ( H unston, 2002 / 2006: 42) /
( 1) so le veh icle; 0 ( frequency) ;
( 2) sole 1363:
: pract itioner, m em ber, surv ivor, trader, owner, propr-i / 0
etor, d irecto r, responsib ility, purpose, source, criterio n,
righ,t reason, m ot iv ation, opportunities, un i,t basis; H unston, :

# 13#
, :

,
: ,
, ,
( ) , , ,
, CEM ,
2000~ 2002 12 , :
;
: , ( )
, , , ( )
: ( ), ( )


( ) ,
;
SPSS ( )

,
, ,
, , ,
, (
, ) , 8
H unston( 2002 / 2006: 43) ( 2000- 2002) ,
, / , :
0
,
:
:
, ;
,
: ,
? ,
,
( 1)
) ) ) ( test spec if ica-
, , 30 tions) ,
( + ) /
? ( ),
,
, ;
( BNC) ,
( ) BNC ; ( ,
) BNC ( ) ) ) )
, ,
,
, ( so le veh icle) ; ( 2)
, ( : fu ll ) ) ) ,
autho rity) , ,

# 14#
, :

( 1995.

) [ 2] A lderson, J. C. Do corpora hav e a ro le in languag e assess-


m ent? [ A ]. In Jenny Thom as & M ick Short eds. U sing

Corpora in L anguage R esearch[ C]. :
, 8
/, 1996 /2001.
, ( CEM )
[ 3] Bachm an, L. F. Fundam enta l Cons ide rations in L anguage
( BNC ), T esting[M ]. O x fo rd: O x fo rd U n ive rsity P ress, 1990.
, ( [ 4] H unston, S. Co rpo ra in A pp lied L inguistics[ M ]. :
) , , /, 2002 /2006.
[ 5] .
, , [ M ] . : ;
: , 2000.

,
[ 6] .
( 2004 ) [ M ]. :

,
, 2004.
, t
[ 7] ) ) ) ( CEM ) [M ] . :
, 2008.
[ 8] , . [M ]. :
[ 1] A lderson, J. C. et a.l L anguage T esting Construction and
, 2005.
Eva luation[ M ]. C ambr idg e: Cam bridge U niversity P ress,

U s ing Corpus in D esign ing and V a lidating Testing Procedures


1 2
ZOU Shen , YANG R en-m ing
( 1. School of Eng lish, Shanghai Internat io na l Stud ie s Un iv ersity, Shanghai 200083, Ch in a;
2. Schoo l of Fo reign L anguages, Shanghai Second Po lytechn ic Un iv ersity, Shanghai 201209, Ch in a)
Abstract: Corpus is play in g an ever increasing ly im portant ro le in language research and pedagog ic research. Sim -i
larly, m o re and m ore im portance is be ing attached to th e potential use of corpus in language testing. The paper attem pts
to explore the ro le of corpus in content validation of the Proofread in g Part in T est for Eng lish M a jo rs( TEM 8). Resu lts
have shown that the use of corpus can facilitate our effo rts to im prove item w rit ing and thus conten t validity. Severa l sug-
gest io ns are m ade in the end for possible uses o f corpus in test construct io n.
K ey w ord s: Corpus; C onten t Va lid ity /Va lid atio n; T esting P rocedure


18

( In ternational F ederat io n of ,
T ransla tors) ( T ranslators A ssoc iation of MT I
China) 2008 18 / N ewC lass DL 760
( XV III F IT W orld Congress) 2008 8 4 7 MT I 0, ,
, NewC lass DL760
MT I
(: )

# 15#

You might also like