August 2012
Graded / Intergraded KPG Language Tests
(in Greek)
Antonis Tsopanoglou, Aristotle
University of Thessaloniki
in collaboration with Bessie Dendrinos, National
and Kapodistrian University of Athens
[abstract] [html]
[pdf]
This paper explains why it is so complicated and
difficult a task to measure a language user’s
proficiency and to measure it accurately, with
precision. Among the reasons put forth is that the state
of knowing or being able to use a language efficiently
is so complicated that those interested in measuring it
find it difficult to agree on which variables to measure
and what scales to use for this purpose. Actually no
matter which set of variables they choose, it is
unlikely that there will not be among them some
variables that one can measure with the same scale as
the rest. The effort to measure one’s language
proficiency in terms of levels becomes an even more
intricate task because one needs to predetermine the
variables of each level; in other words, in order to
measure them, it is necessary to specify what exactly to
measure at each level of proficiency – to provide
objective descriptors as to what the language user must
be able to do at each level of proficiency and how well.
Those involved in such efforts are aware of how trick
this job is. They also know that while the descriptors
are qualitative in nature, when the project at hand is
to measure these with an instrument such as a test, the
qualitative variables must be measured quantitatively.
This is one of the reasons why discrete level tests are
questioned in this article, which argues in favour of
using one instrument across language levels –an adaptive
electronic test. When it is necessary to use a
pen-and-paper test, an in-between solution is proposed:
to use integrated/graded language tests, following the
three basic levels of the Council of Europe.
-----
[Back
to Previous Articles] |
|
