The Common European Framework in its political and educational context What is the Common European Framework?

9
Assessment
9.1
Introduction
Assessment is used in this chapter in the sense of the assessment of the proficiency of
the language user. All language tests are a form of assessment, but there are also many
forms of assessment (e.g. checklists used in continuous assessment; informal teacher
observation) which would not be described as tests. Evaluation is a term which is again
broader than assessment. All assessment is a form of evaluation, but in a language pro-
gramme a number of things are evaluated other than learner proficiency. These may
include the effectiveness of particular methods or materials, the kind and quality of dis-
course actually produced in the programme, learner/teacher satisfaction, teaching effec-
tiveness, etc. This chapter is concerned with assessment, and not with broader issues of
programme evaluation.
There are three concepts that are traditionally seen as fundamental to any discussion
of assessment: validity, reliability and feasibility. It is useful in relation to the discussion
in this chapter to have an overview of what is meant by these terms, how they relate to
one another, and how they are relevant to the Framework.
Validity is the concept with which the Framework is concerned. A test or assessment
procedure can be said to have validity to the degree that it can be demonstrated that
what is actually assessed (the construct) is what, in the context concerned, should be
assessed, and that the information gained is an accurate representation of the profi-
ciency of the candidates(s) concerned. 
Reliability, on the other hand, is a technical term. It is basically the extent to which the
same rank order of candidates is replicated in two separate (real or simulated) adminis-
trations of the same assessment. 
What is in fact more important than reliability is the accuracy of decisions made in rela-
tion to a standard. If the assessment reports results as pass/fail or Levels A2+/B1/B1+, how
accurate are these decisions? The accuracy of the decisions will depend on the validity
of the particular standard (e.g. Level B1) for the context. It will also depend on the valid-
ity of the criteria used to reach the decision and the validity of the procedures with
which those criteria were developed. 
If two different organisations or regions use criteria related to the same standards in
order to inform their assessment decisions for the same skill, if the standards themselves
are valid and appropriate for the two contexts concerned, and if the standards are inter-
preted consistently in the design of the assessment tasks and the interpretation of the
performances, the results in the two systems will correlate. Traditionally the correlation
between two tests thought to assess the same construct is known as ‘concurrent validity’.
177

---

This concept is obviously related to reliability, since unreliable tests will not correlate.
However, what is more central is the extent of communality between the two tests regard-
ing what is assessed, and how performance is interpreted.
It is with these two questions that the Common European Framework is concerned.
The next section outlines three main ways in which the Framework can be used:
1.
For the specification of the content of tests and 
what is assessed
examinations:
2.
For stating the criteria to determine the attainment 
how performance is interpreted
of a learning objective:
3.
For describing the levels of proficiency in existing 
how comparisons can be made
tests and examinations thus enabling comparisons 
to be made across different systems of qualifications:
These issues relate to different kinds of assessment in different ways. There are many dif-
ferent kinds and traditions of assessment. It is a mistake to assume that one approach
(e.g. a public examination) is necessarily superior in its educational effects to another
approach (e.g. teacher assessment). It is indeed a major advantage of a set of common
standards – such as the Common Reference Levels of the Framework – that they make it
possible to relate different forms of assessment to one another.
The third section of the chapter lays out choices between different types of assessment.
The choices are presented in the form of contrasting pairs. In each case the terms used
are defined and the relative advantages and disadvantages are discussed in relation to
the purpose of the assessment in its educational context. The implications of exercising
one or another of the alternative options are also stated. The relevance of the Framework
to the type of assessment concerned is then pointed out.
An assessment procedure also needs to be practical, to be feasible. Feasibility is partic-
ularly an issue with performance testing. Assessors operate under time pressure. They
are only seeing a limited sample of performance and there are definite limits to the type
and number of categories they can handle as criteria. The Framework seeks to provide a
point of reference, not a practical assessment tool. The Framework must be comprehen-
sive, but all its users must be selective. Selectivity may well involve the use of a simpler
operational scheme, which collapses categories separated in the Framework. For
instance, the categories used in the illustrative scales of descriptors juxtaposed to the
text in Chapters 4 and 5 are frequently considerably simpler than the categories and
exponents discussed in the text itself. The final section of this chapter discusses this
issue, with examples.