Persistent problems

and future
directions
 Just as the shift in emphasis from language form to
language use has placed new demands on language
teaching, so too has it placed new demands on
language testing. Evaluation within a communicative
approach must address, for example, new content areas
such as sociolinguistic appropriateness rules, new
Introduction testing formats to permit and encourage creative, open-
ended language use, new test administration procedures
to emphasize interpersonal interaction in authentic
situations, and new scoring procedures of a manual and
judgemental nature.
(Canale 1984: 79)

2
Challenges
As language testing research

As developers and use

language test

3
Authentic language test
Defining Authenticity
Real-Life (RL) approach
defining authenticity essentially
considers the extent to which test
performance replicates some specified
non-test language performance.
Interactional/ Ability (IA)
approach
is in keeping with both the mainstream
approach to measuring language as a
mental ability and the current view of
communicative language use that has
been discussed extensively throughout
this book.
4
 Authenticity as ‘real-life’ language use
RL approach defines language proficiencyas the ability to perform language tasks in
non-test situations, and authenticity as the extent to which test tasks replicate ‘real-
life’ language use tasks

• view of language ability, or proficiency, as pragmatic ascription

1

• The reference to ‘real-life performance’ as a criterion

2
• the belief that ‘face validity’, content relevance, and predictive utility are
3 sufficient bases to justify test use
Proficiency as ‘real-life’ performance
‘real-life performance’ provides the basis for a distinction between
‘direct’ and ‘indirect’ tests of proficiency:

• the testing format and procedure attempts to duplicate as

Direct closely as possible the setting and operation of the real-life
proficiency situations in which the proficiency is normally demonstrated.
(Clark 1975: 10)
testing,

• are not required to reflect authentic language use

contexts and, indeed, they may in many cases bear little
Indirect formal resemblance to linguistic situations that the
measures student would encounter in real life. (Clark 1978b: 26)
Va l i d i t y o f D i r e c t t e s t

Predictive Utility (Face

validity)
Wesche (1985) states that predictive utility
is the major concern of performance tests:
‘performance tests. . , are expected to
predict how successfully the examinee will
be able to communicate using the second
language in certain target situations, rather
than to establish his or her general
proficiency level or to provide diagnostic
feedback for the classroom’
Content Validity
is the extent to which the content of the test
constitutes a representative sample of the
domain to be tested, and the main problems
in demonstrating content representativeness
are related to the adequacy of the sample.