"Weighing the pig doesn't fatten it"
by Dimitrios Thanasoulas
the various forms of assessment fall into two major categories
of summative assessment (which shows a level of achievement)
and formative assessment (which guides learning in the future).
Assessment, however, takes a similar form in both cases. In
practice, a particular assessment usually has both functions.
most common forms of assessment involve 'summarising' levels
of achievement. Formal assessments such as exams are of great
importance to the students involved, since they provide the
key to employment or higher education. They are also important
to schools, since they are being used to evaluate the performance
of schools and teachers alike.
Tests are a mainstay of what goes on in schools, but this
can have a backwash effect, in the sense that the content
of tests comes to dominate what is taught. It stands to reason
that a limited test cannot give a realistic assessment of
performance across the whole curriculum. Most formal tests
are (or try to be) selective, focusing on what can easily
be assessed in an examination situation. Everyone is aware
of this but teachers are under constant pressure to deliver
a curriculum that covers only what it can assess, to the detriment
of deeper knowledge and understanding.
There is nothing "innocuous" about assessment; it
constitutes a form of evaluation that pupils often use to
make judgements about their own competence and, by extension,
intelligence. As students go through school, grades and the
results of assessments seem to become increasingly important
in determining their involvement, as well as their self-image,
either positively or negatively.
assessments are used to help direct or 'form' the educational
process for students. One key feature in formative assessment
seems to be the role of feedback to pupils. For feedback to
be effective, it should focus on details of students' work,
giving advice as to which particular areas need remedial work,
rather than merely a comparison with other students' work.
In a study by Butler (1988), 48 students were given feedback
which took one of three different forms. The first type consisted
of detailed comments about the students' performance, in relation
to criteria set for the specific topic of work. The second
type of feedback comprised the students' overall grades, whereas
the third type combined grades with detailed comments. The
students given the detailed comments showed a 30 per cent
improvement, while those who received only grades showed no
improvement whatsoever. Interestingly enough, the same holds
for those who received both grades and comments. As a revealing
gloss on these findings, we could say that, apparently, any
form of evaluative comment tends to distract attention away
from informational content and actually decreases learner
can be carried out for a number of different reasons. MacIntosh
and Hale (1976) have highlighted a number of well-known categories
organised along a formative-summative continuum (taken from
Long, 2000: 52).
Diagnosis: It involves finding out skills, strengths and weaknesses,
implying that teaching should change
Guidance: Test scores can be used to direct students
Selection: Tests can provide the basis for placement in schools
or other forms of education
Prediction: An indication of potential academic progress
Evaluation: Giving a value to a pupil's abilities, which may
be recorded and used for monitoring
Grading and Certification: Test results lead to a qualification
However, Gipps et al. (1983) showed that many tests exist
only for purposes of record keeping. In many cases, even when
teachers themselves use regular testing, they rely on their
own judgements to make assessments about students, disregarding
test scores (Salmon-Cox, 1981).
two main categories of tests are referred to as criterion-referenced
and norm-referenced tests, and have different functions and
rationales. The purpose of a criterion-referenced test is
to compare each individual's abilities with some form of criterion,
whereas the purpose of a norm-referenced test is to discriminate
between individuals or to compare them with one another.
tests assess performance on specific features of ability attainments.
With reading, for example, such a test might look at whether
a student knows some letter sounds, or whether he or she can
read certain words. Inasmuch as they help identify strengths
and weaknesses and may guide future learning, they can be
regarded as formative tests. A criterion-referenced English
test might identify that students have weak comprehension
skills, which would mean that it would be fruitless to go
on to listening comprehension tasks until they have developed
a strong enough basis with these skills.
'norm' we mean a typical value for something. Norm-referenced
tests are designed to allow a student's abilities to be assessed
in relation to all the other students of the same age. In
this light, they are mainly summative tests, although if they
can identify specific skills that can be taught, they constitute
types of formative assessment.
Norm-referenced tests are developed by first constructing
a number of items that are supposed to assess abilities in
a particular domain. With reading, this might involve using
a list of words of increasing complexity and length. Then,
the test is checked for reliability (dependability) and validity
(meaningfulness), and modified until it meets the desired
criteria. After that, the test is standardised by giving it
to a sample of children that cover an appropriate age range,
so that they are representative of the wider population. This
information then feeds into the construction of age-standardised
tables that can be used to compare individual test results.
The standardisation of these tests is very important, but
many such tests were standardised decades ago, which is not
a sound basis on which to assess present-day abilities.
notion of general ability or intelligence has been the most
important way of explaining individual differences. It is
usually assessed by means of measuring performance on a test
of different skills, using tasks that emphasise reasoning
and / or problem solving in different areas. It can be expressed
as an overall IQ (Intelligence Quotient). Early assessments
of IQ were premised upon Alfred Binet's research in 1905,
as part of an attempt to identify students who needed specialist
help. After Binet, many English and American psychologists
and educationists interested themselves in IQ tests, and there
was a general belief that intelligence was largely inherited
and, as a result, quite stable over a student's school career.
There is a wide range of IQ tests purporting to assess verbal
and non-verbal intelligence, but we will not concern ourselves
with them in the present article.
page 3 of 3
to the articles index