Assessment and Language Learning

10.1 Issues in assessing children'slanguage learning

10.1.1 Introduction
The nature of children's foreign language learning might be expected to generate a
range of assessment issues in need of attention. Factors such as the following might
make the business of assessing young learners different from assessment practices in
other FL situations: ·
• Age: children's motor, linguistic, social and conceptual development must be
taken into account in designing and implementing assessment.
• Content of language learning: a focus on oral skills, vocabulary development and
language use at discourse level.
• Methods of teaching: interactive use of games, songs, rhymes, stories to carry
language content and practice.
• Aims: programmes for young learners often cite social and cross- cultural aims,
as well as language learning aims.
• Learning theories: e.g. zone of proximal development; learning through social
interaction, able to do more with helpful other.
A survey of the sparse information that is available on young learner
assessment (but see Rea-Dickins· 2000) suggests that assessment practices, far from
taking careful account of the above factors, may find themselves in conflict with
them. In this first section, macro-level and then more micro-level conflicts· are
identified. Section 10.2 sets out principles for assessment that emerge from our learning-
centred perspective. Key concepts in assessment are explained in section 10.3. In
section 10.4, we consider techniques other than testing· that teachers can use to assess
children's language learning. Section 10.5 discusses self-assessment in the young
learner classroom. Section 10.6 considers how the outcomes of assessment can be
used to support learning through feedback to children. The chapter concludes
with some thoughts on the impact of assessment practices on children's motivation and
future learning.

10.1.2 The Social Realities of Assessment

Social realities, in the form of political, commercial and cultural dynamics
underlie several conflicts around the role of assessment in language teaching and
learning, particularly when we are concerned with assessing children.
It would seem reasonable to require assessment to serve teaching, by providing
feedback on pupils' learning that would make the next teaching event more effective, in
a positive, upwards direction. Teaching and learning needs should dictate the form and
timing of assessment. In practice, the scenario is quite different: assessment seems
to drive teaching by forcing teachers to teach what is going to be assessed. And this
happens around the world, with young learners as well as older students. Three
.examples will illustrate the power of assessment over teaching and over learning.
In England, after decades without any national testing requirement at primary
level, the government introduced a national curriculum and assessment at 'age 7, 11
and 14, with baseline assessment on entry to school at age 5. Initially, assessment
tasks. were designed to involve the children in familiar activities and enable the teacher
to assess each child, but these were soon reduced to paper and pencil tests taken in
exam conditions. Practice books are bought in newsagents by anxious parents, and,
increasingly, classroom work in Year 6 (age 10-11 years) is preparation for the
attainment tests. As I write this; parents and teachers have begun to protest at the
'stress' being felt by seven year old children, and to ask for a review of assessment
procedures.
In Malaysia, the communicative· English language syllabus becomes in
practice a formal, grammar-based syllabus, because the examination at Year 6 is.
grammar-based, and parents and headteachers demand that pupils are prepared for the
examination. From the age of 7, pupils are tested every month, every term and every
year. The marks are used in some schools to place children in different classes and
different groups within a class. The test results do not impact on teaching, because the
next stage of the syllabus ( or course book) will be tested in the next examination.
The testing practices and the syllabus together determine what a child will experience
in lessons, with little room for taking account of his or her individual needs.
On a global scale, a relatively new test for young Iearners developed by UCLES
has· taken off rapidly, mostly through· private language schools. One hundred
and fifty thousand· children world-wide are expected to .take the test in the year
2.000, with particularly large numbers in China and in South America, but also in
Pakistan, Bangladesh and European countries. Although the test assesses a
child's progress rather than awarding a pass / fail, parents often want to know whether
their child has 'passed'. The Examination Syndicate produces a list of words and topics
that will be tested. Books of past papers are being published an will no doubt be
worked through-by anxious children preparing for the tests. It will be increasingly
difficult for a teacher or language school to teach topics or vocabulary that are not on
that list because they may reduce their pupils' chances of success in the test. Changing
an international test is a long process, because new items must be validated against
old ones so that scores are comparable. Inevitably and inexorably, the test,
however well' intentioned and planned, concretises language teaching by
diminishing the opportunity for creativity in the classroom.
In the world of foreign language teaching, assessment, usually in the form of
'testing', has become a multi-million dollar global business, in which the need for
internationally recognised certification of language proficiency works with learners' or
their parents' understandable demands to see proof of the outcomes of their struggle to
learn and the money they have invested in it. In turn, the market-driven development of
quality tests relies on and supports the development of research and theory around
language testing. As a result, the academic field is peopled with intellectuals
who progressively complexify the concepts and techniques to such as extent that
the language teacher or teacher trainer finds it difficult to make sense of what is
written, let alone to ask critical questions. It is an irony that testing experts can
increasingly only talk to and understand each other, when, around the world, on a
daily and weekly basis, teachers who are largely untrained in testing write and
mark tests that determine the learning opportunities of millions of children.
The dynamics of the interacting worlds of learners, academia and
commerce combine to produce effects of testing on language learning that are felt
throughout the system. This "washback' from assessment to learning has an impact on
individual learners, teachers and the wider system:
• stress is placed on children by the demands of assessment;
• individual children's learning needs are downgraded in the push to cover the
syllabus or course book before the next assessment;
• classroom activity is restricted to test preparation;
• educational change is limited by the power of the assessment machinery.
Not all washback effects are negative. We might argue that children will
encounter stress anyway in their educational lives, and that well-designed
assessment may help them learn how to cope with more stressful examinations
later in life. We know, too, that some ideas for educational change may be over-
optimistic or just plain silly; good assessment may prevent damage to children's
learning opportunities from irresponsible change. The power of assessment to
change practice can be used positively and innovative testing can increase attention to
neglected aspects of learning, as happened when oral skills were assessed for the
first time in English schools. Test results provide useful feedback to stake-holders -
governments and parents - about how well children are being taught, and about the
effectiveness of policy and schools. Comparison of school results can highlight where
pupils are underachieving that might otherwise not be noticed, and thus lead to
improvements in learning opportunities.
The various realities of assessment in education create an uncomfortable
scenario for advocates of learning-centred teaching. We have to recognise that these
assessment realities derive from cultural and commercial factors that cannot be
ignored, and that will change only slowly, if at all. Given this background, our
concern must be to understand the impact of assessment on teaching, and its variations in
different cultures, so that we can try to maxime its benefits for learning.
 10.1.3 Classroom realities
When we move from macro-level scenarios to what actually happens in schools
and classrooms, we find evidence of further conflicts in assessing children's language
learning. ·
A survey by Rea-Dickins and Rixon (1999) of 120 teachers and teacher
trainers, mostly from Europe, revealed that the vast majority (93%) of the teachers
do assess children, with the stated purpose of helping their teaching (87%). Most
of this assessment is prepared by classroom teachers, who are also mostly
responsible for marking and record-keeping. However, when they investigated
what was being assessed and how, Rea-Dickins and Rixon found 'a mismatch ...
between curricular aims, pedagogy, and test content' (1999: 96). The focus for
most of the assessment was on children achievements in language learning, rather
than on other curricular aims such as increased language awareness or social
awareness. Moreover, by far the most frequently used method of assessmem was the
paper-and-pencil test, testing single items of vocabulary and grammar through single
entences. The content and method of this type of assessment contrasts vividly with the
classroom experience of children who have learntt language through participation
in discourse-level stories and songs. Rea-Dickins and Rixon note the mismatch in
level of text, but should also note the disjuncture between the interactive
learning environment and the non-interactive, solo experience of doing a test.
A further contrast that emerges from the survey is between the attention
to oral skills in the classroom and in assessment. Very few of the tests that were
reported focused on spontaneous speaking; it seemed that what was assessed was what
was relatively 'easy' to assess. It is much easier to develop written tests than
assessments of spoken language. Since the rise of communicative language
teaching in the 1980s, testers have been struggling with how to assess oral skills, and
spoken tests are now becoming available over the telephone and computer. On
the ground, in schools and classrooms, because it is much more difficult to devise and
mark oral assessments fairly, most assessment is still carried out on paper. It is
possible that a requirement for grades and marks to show progress forces teachers
into using written tests, perhaps with the unfortunate consequence of
concentrating teaching on the written forms before children are able to cope with the
demands that would make.
Elsewhere, Rea-Dickins and Rixon draw attention to the more advanced
assessment practices that can be found in use with English as a Second Language (ESL)
and first language children, and suggest that assessment of children's foreign language
learning can draw on this work to improve methodology (Rea-Dickins and Rixon
1997). We return to the possibilities offered by these neighbouring fields later in the
chapter.
 Before that, the issues that have been identified in this section are used to
generate some guiding principles for the assessment of children's : foreign language
learning.

10.2 Principlesfor assessing children'slanguage learning·

10.2.1 Assessment should be seen from a teaming-centred perspective
We should at this point restate the learning-centred perspective of the book
so that it can be applied to assessment. We have been building up a picture of foreign
language teaching that has children's learning in the centre, trying to understand how
classroom activities and talk will be experienced by children. We have emphasised
children's willingness to participate in social interaction and their drive to make
sense of the activities and talk they engange in. A Vygotskyan perspective on learning
emphasises that learning occurs in social contexts and through interaction with helpful
adults or other children. As we have examined the development of different aspects of
foreign language knowledge and skills, the approach has been to examine how
children's participation and engagement can provide opportunities for changes in
these language resources that we would call 'learning', including the internalisation
(Vygotsky 1962} or appropriation (Wertsch 1998) by the Individual child of language
used first with other people. Vygotsky turned ideas of assessment around by insisting
that we do not get a true assessment of a child's ability by measuring what she or he
can do alone and without help; instead, he suggested that what a child can do with
helpful others both predicts the next stage in learning and gives a better assessment of
learning. This kind of 'scaffolded assessment' (Gipps 1994) is far removed from the child
seated in silent isolation to take a test.

10.2.2 Assessment should support learning and teaching

If learning is our central concern, then, in an ideal world, assessment should
contribute to the learning process, for both an individual child and for the class. From
the three examples in the first section of the impact of assessment, we can see that,
even when a supportive relationship between assessment and teaching / learning is
intended, social realities can rapidly push the relationship into something quite different.
In order to be more in control of the relationship between assessment and
learning, teachers need to have a clear understanding of language learning processes
and of the socio-cultural context in which they operate. They can then predict
the impact of assessment on their teaching and plan accordingly. If the picture of
language learning can be communicated to learners and their parents, then it may
also help parents to understand what assessment can tell them and what its limits are.
Metaphors help us in constructing and communicating ideas and concepts,
and I want to suggest a possible metaphor for language learning. In second
language acquisition research over the last decades, the metaphoric view of language
learning as the building of a higher and higher tower of bricks, through the
accumulation of bits of grammar and vocabulary, has shifted to a more organic
view of language learning as growth and development. At several points in the book a
metaphor of plant growth has been applied to foreign language learning. We extend it
now to assessment, visualising roots of language knowledge spreading out to support
increasing use of the language. The plant develops through the nutrients it absorbs
from its environment and different types of growth occur at different points in its life
cycle. Assessment asks how well the plant is growing. However, growth is not just
about the height of the plant, but concerns the strength of its root system, the quality
of its leaves, the number and richness of the flowers. We would want to avoid the kind
of assessment that involves pulling up the seedling to see how far the roots have grown
and that, in doing so, slows down the very growth we want to happen. This might
be a parallel with grammar-based tests that prevent more communicative approaches
to teaching. Instead, we would want to find some 'non-invasive' methods of
assessment that indicate learning; it is unlikely that a single measure will suffice to
describe overall growth.
Thinking through the metaphor suggests a basic principle that assessment
should not disrupt learning. We can go further than this negative requirement and
ask that it positively support learning. There are various ways in which assessment
can have positive effects:
• the process and outcomes of assessment can motivate learners;
• an assessment activity can provide a helpful model of language use;
• an assessment activity, and feedback from it, can support further learning;
• the outcomes of assessment can help teachers plan more effective
lessons;
• the outcomes of assessment can inform the evaluation and improvement of
courses and programs.
The principle of support will force us to look carefully at what
opportunities an assessment activity offers for learning.

10.2.3 Assessment is more than testing

A gardener continuously assesses how well various plants are growing, noticing
changes and judging what they mean by bringing past experience and knowledge
to what is seen. Similarly, a skilled teacher continuously assesses pupils' learning
through what s/he notices and how s/he interprets these observations in the light
of experience and knowledge. It is not necessary to test children to understand how
much they have learnt - or at least it is not necessary to do it too often. There are other,
supportive, ways to assess language learning that go beyond testing, either the simple
classroom test or the more stringent national or international test. Such 'alternative
assessment' techniques include observation, portfolios, and self-assessment
(O'Malley and Valdez Pierce1996), and we will look at some in more detail later in the
chapter.
Once again though, we should note that such skilled assessment of learning
is not a trivial matter, but needs training and refining over several years.

10.2.4 Assessment should be congruent with learning

By 'congruent' I mean that assessment should fit comfortably with
children's learning experience. On the whole, it is fairer to assess children on the basis
of what they have been taught and how, using assessment activities that are familiar to
children from their classroom experience.
This principle suggests that assessment should, like teaching and learning,
be interactional rather than an isolated, solo experience. McNamara (1996)
discusses the neglect of social interaction within testing of older learners; working
with children highlights the issue even more strongly. We also need to ensure
adequate oral assessment of discourse skills. .

10.2.5 Children and parents should understand assessment issues

Even if individual teachers want to convert their understandings of learning
into new assessment practices, nothing can change without the support and
involvement of key players: the learners themselves and their parents. In current
terminology, they are prime 'stakeholders' in the educational process. As we examine
techniques for assessment, we will emphasise the need for children to understand
the purposes of activities and to play a role in them. Self-assessment can be a part of
learning from the beginning, and can contribute to the development of self-
motivated and self-directed learners at later stages.
Teachers are restricted in the individual decisions they can take on
assessment; often, in fact, they can take no decisions because there are national
regulations that must be followed. In some situations, though, decisions are left to
schools, perhaps about how to prepare children for national examinations or how
often to assess in between national tests. Then teachers can be involved in constructing
school policy that is built on learning-centred principles. School policy on assessment
must of course take account of parents' demands, but also has a responsibility to
inform and educate parents about the theories of learning that underpin a
school's teaching. Parents around the world want the best for their children, but this
often manifests itself as wanting the same as happened in their own schooling.
Parents need to know what teachers are doing and why; the effort it takes to
explain will be repaid by parental support for teachers. Even stronger support
can be had if teachers explain and model what parents can do to help their children.
For example, the Malaysian parents involved in the literacy project mentioned in
Chapter 6 came to understand that reading story books to their children could help
develop literacy skills more effectively than testing spellings at the early stages, and
so started buying story books for them instead of the activity books they used to buy. At
the very least, parents can see how they can most effectively help children cope with
examinations, and teachers can explain how assessment activities other than tests
provide information on children's learning.
Having set up in the first two sections the background for the assessment
of children's language learning, in terms of issues and principles, this section
will clarify some of the important concepts and terms needed for deeper discussion of
assessment.

1.3 Key concepts in assessment

10.3.1 Assessment - Testing - Evaluation
We have already seen that tests are just one technique or method of
assessment. The other term that needs differentiating is evaluation. Evaluation refers
to a broader notion than assessment, and refers to a process of systematically
collecting information in order to make a judgement (Rea-Dickins and Germaine
1982: 22). Evaluation can thus concern a whole range of issues in and beyond
language education: lessons, courses, programs, and skills can all be evaluated. If we
were to evaluate a course, we would need to collect many different types of
information: course documentation, observation of lessons, interviews with pupils
and teachers, course feedback questionnaires, examination results. Analysing and
combining the different types of information would enable a judgement to be made
about the success, or viability, or cost-effectiveness, of the course.
Assessment is concerned with pupils' learning or performance, and thus
provides one type of information that might be used .in evaluation. Testing is a
particular form of assessment, that is concerned with measuring learning through
performance.
10.3.2 Formative and summative assessment
A useful distinction in assessment is made in terms of the purpose and use of
assessment information. Formative assessment aims to inform on-going teaching
and learning by providing immediate feedback. A teacher who assesses pupils'
understanding of a listening text and uses the outcomes to change her plan and give
more practice before moving on to a speaking activity, is carrying out formative
assessment. Ideally, formative assessment should influence both teaching and
learning by giving feedback to both teacher and learners (Gipps 1994).
Summative assessment, on the other hand, aims to assess learning at the end of
a unit, term, year, or course, and does not feed back into the next round of teaching.
,
The formative / summative distinction is hardly visible in the current
 state of child language learning assessment (Rea-Dickins and Rixon 1997).
In order to become better defined they suggest that a wider repertoire of
techniques is needed and attention to closing the gap between pedagogy and
assessment practices is highlighted in section 10.1.3 above.

10.3.3 Diagnostic and achievement assessment

Many assessment activities provide both formative and summative
information, but it is helpful to be clear as to the primary purpose and use of an
assessment because this can affect what kind of information the activity needs to
produce. An assessment of pronunciation skills that is formative will need to tell us
where pupils are having difficulty so that the teacher can decide how to give extra
practice; a test that gives a list of marks will not help the teacher make such decisions,
but an activity that produces a description of each child's performance will. This
example highlights the distinction between assessing achievement, i.e, what a learner
can do, and diagnostic assessment that aims to establish what a child can. and cannot
yet do, so that further learning opportunities can be provided.

10.3.4 Criterion-referenced and nonn-referenced assessment

If we assess learners' achievement, we can produce a ranking of learners which
says that child X has learnt more than child Y and less than child Z; this would be
norm-referenced. Alternatively, we can compare a learner's performance, not to other
learners, but to a set of criteria of expected performance or learning targets. Criterion-
referenced assessment can match the child's performance against an expected
response on an item, or it may make use of a set of descriptors along a scale, on which a
learner is placed. ESL assessment in Australia makes use of a set of 'bandscales' on
which a learner is placed through performance on classroom tasks (McKay 1995).
·
Examples of a statement about expected response and a descriptive scale are
given in respect of speaking skills. In the first example, taken from the Cambridge
Test for Young Learners (UCLES 1999), the learner is asked simple questions by an
examiner. The criterion used to assess the child's speaking skills is the production of
answers in single words or short phrases, and is rated on 'interactive listening ability,
pronunciation and production of words and phrases' (UCLES 1999: 12).
'In the second example, teachers assessing the oral skills of second language
learners are given a six-level scale on which they can place each learner (O'Malley
and Valdez Pierce 1996). The sub-scale for fluency has the following descriptor
statements:
1. repeats words and phrases;
2. speaks in single word utterances and short patterns;
3. speaks hesitantly because of rephrasing and searching for words;
 4. speaks with occasional hesitation;
5. speaks with near-native fluency; any hesitations do not interfere with
communication;
6. speaks fluently.
(O'Malley and Valdez Pierce 1996: 68)

Criterion-referencing of learners' performance relies, of course, on

expected norms, but the distinction is useful and is important for our principle of
wanting assessment to support learning. While norm-referencing may motivate
some children to do better than their peers andmove up the ranking, it may have a
negative effect on a much larger proportion of the children. Giving feedback on
assessment in terms of how well you are doing relative to others does not help you know
how to do better. A band scale like the second example is potentially much more
helpful because it gives a picture of the process of learning that each learner will
move through and thus lets a learner see where they are compared to more expert
performance. However, to be useful in this way means that such scales must be very well
researched and it is likely that, while a very specific description will be more useful, an
increase in precision also makes the scale less reliably applicable to individuals.

10.3.5 Validity
The concepts of reliability (see below) and validity are used to describe the
technical quality of assessment practices. They are more often applied to testing,
although are also important in alternative assessment. Validity is the more important,
particularly in alternative assessment, and concerns how far an assessment assesses
what it claims to. If a test does not measure what it claims to, then there are clearly
dangers in using it (Gipps 1994). In order to evaluate the validity of an assessment, we
must compare the skills or knowledge that we want to assess with what is actually
assessed, and also examine the claims made about pupils' ability from their
performance in the assessment. Consider a simple test that showed children various
pictures of objects to name in the foreign language as a test of vocabulary. Suppose the
pupils being tested had learnt the word 'milk' and linked it to the cartons of milk
that they regularly buy from their supermarket. If the test picture for milk showed a
glass bottle on a doorstep (as milk is experienced by a child in the UK, although
increasingly infrequently!), then the testee may well not be able to answer. Rather
than testing knowledge of the vocabulary item, the test was testing knowledge of
another culture. It would thus not be valid as a test of the word milk.
If an assessment omits some aspects of what is being assessed, its validity
can also be reduced. For example, if we claim to have assessed a child's writing skills
but only give a mark for spelling and neatness, and omit discourse level skills of
organisation and sequencing, this would not be a valid assessment.
 It will be clear that the assessment of young children's, mostly oral, language
learning is not validly done through pencil and paper tests that require written responses
(see section 10.1.3).
To make sure an assessment is as valid as possible, we need to think very
carefully about what exactly we want to assess, what exactly the proposed
assessment will assess, and what can be claimed from the outcomes of the
assessment.
Validity has also been extended to include considerations of the social
consequences of assessment (Messick 1989), in asking whether the uses to which
results of assessment are put are justified by the nature of the assessment. For example,
assessment can be used to select pupils, to stream pupils by grouping together those
with similar results, or to take some pupils out of the mainstream and into special
educational units. Tests or assessments should not be used for purposes they were not
designed for without checking on their validity in the new circumstances.

10.3.6 Reliability
Reliability measures how well a test or assessment assesses what it claims
to: would the assessment produce the same results if it were taken by the same
pupils on different occasions, or if the same test or assessment was scored by different
people? (Gipps and Stobart 199).
When applied to tests that produce numerical marks, reliability can be
checked by statistical comparison of performances on two similar tests a few days
apart. However, this method would not work for criterion-referenced assessments
because individuals may not be widely spaced across the results.
It should be clear that reliability can be affected by the conditions under
which pupils are assessed, and thus by what teachers do when they explain
assessment activities to pupils (Gipps 1994). Reliability is increased by being very
explicit about instructions to pupils and, in scoring, by moderation, i.e. having
markers score the same scripts until they mark consistently in the same way.
Validity and reliability can be conflicting needs for assessment techniques
and procedures. The most reliable assessments will be pencil and paper tests in which
each item measures only a single aspect of a skill and which give each testee a
numerical mark. But the most valid assessments will be those that collect a lot of
information about performance on several aspects of a skill. When validity is
increased, reliability decreases. The validity / reliability conflict echoes that
described in section 10.1.2, and creates a real dilemma for education, that is being
felt across America and Europe, between political movements to use assessment to
increase the quality of education by monitoring national standards via tests, and
new approaches to assessment in which teachers and pupils use a range of techniques
to feed back into setting targets and motivating learners (Gipps 1994). Gipps
suggests that the way through this conflict lies in careful identification of the purposes
and uses of assessment, so that a suitable balance of validity and reliability can be
found for each instance.

10.3.7 Fairness
Fairness, or 'equity' (Gipps 1994), needs to be considered in the design and use
of assessment. In the first section, we saw something of the power of assessment:
• in shaping the curriculum and pedagogy;
• in affecting pupils' motivations and sense of themselves as learners;
• in awarding certificates, setting and streaming;
• and, in some situations, controlling access to the next stage of
education.
(based on Gipps 1994: 144)
Furthermore, assessment in children's language learning, as part of their early
experience, can influence whether or not pupils choose to continue learning the foreign
language or whether they lose interest and motivation. Because assessment
potentially has such a powerful 'washback' effect on children's lives, issues of
fairness must be taken seriously.
Equity principles require that children are given plenty of chances to show
what they can do, and that their language learning is assessed through multiple
methods. It is important in planning and designing assessments that the content is
scrutinised to make sure that culturally unfamiliar pictures or concepts do not
reduce children's chances to demonstrate their language learning. The types of
questions, test items, or assessment tasks should also be familiar to pupils, if they are
to show their ability to best advantage. Children who have not played games in their
classrooms would be at a severe disadvantage if tested through game-like activities
(Dossena 1997). Where assessment requires oral production, we need to be aware of
how children's willingness to talk to adults can vary, and how different children may
come to interview-type assessments with very different previous experience of talking
to adults on a 1-1 basis. Instructions to children also need to be carefully planned and
given, if they are to have equal chances to achieve on the assessment.
Scoring and recording' the outcomes of assessment also need to be carried
out with attention to fairness, so that, for example, observing children. in activities
to assess their oral skills is not biased by their behaviour, or appearance, or gender.

10.3.8 Planning the assessment of children's language learning

The concepts discussed in this section matter for classroom assessment as well
as for large-scale testing. Points of importance from this section can be summarised as a
checklist of questions around assessment that teachers can use when planning
assessment and to help decide what types of assessment are appropriate.
 Questions to guide assessment planning

Purposes and objectives of assessment

Which aspect(s) of language learning do I want to assess?
How does this relate to the learning experience of the children?
What do I want to use the assessment outcomes for?
Who else will use the outcomes? and for what purposes?

Methods of assessment
How will information be gathered to assess the aspect(s) of language?
How will the information that is collected be interpreted?
How will pupils be involved in gathering the information?

Quality managementin assessment

How can I make sure the assessment is valid?
How can I make sure the assessment is reliable?
How can I make sure the assessment is fair?

Feedback
Who will I share the assessment outcomes with?
How will I communicate the outcomes of assessment?

Uses of assessment
How will the outcomes of assessment inform future teaching, planning and
learning opportunities?

Impact of the assessment

What washback effects from assessment to teaching may occur?
What will the impact be on pupils' motivation?

The next section will focus on methods and content of assessment, and on the
kind of inferences that can be made from information collected during assessment.
10.4 Teacher assessment of language learning
At this point in the chapter, we leave testing to one side (see Rea- Dickins
2000) and focus on teacher assessment of language knowledge and skills in
classroom situations. Earlier chapters of the book are reflected upon in assessment
terms.

10.4. 1 Assessing in relation to goals

 In Chapter 2, I emphasised the importance of having clear language learning
goals for classroom activities, and the emphasis was reiterated in connection with
stories and themes as holistic, integrative methods for language teaching. By making
goals explicit, it was argued, we have a check on the potential value of each lesson to
the pupils. These same goals will make assessment a much more straightforward
process because they can act as a target or focus against which we can measure what
was actually learnt. We may decide to assess all goals together, or each separately.
There are some simple, informal ways of assessing learning that will work
with very young children. For example, if goals include learning the names of
animals (as in Chapter 3 ), assessment might be done during teaching using simple
techniques such as:
understanding: The child listens to a word and points to the picture.
The child listens to 3 animal words and chooses 3 pictures in the
same sequence.

production: Teacher points, child says the animal word. Pairs 'test' each other
with picture cards, teacher observes.

For children, informal assessment like this will not feel threatening. For the
teacher, learning activities become assessment activities when there is a clear
assessment focus, i.e, some specific aspect of language learning is attended.
Assessment requires the focused use of skills that are also essential to teaching:
finely tuned observation and systematic, detailed record-keeping. We consider each in
turn.

10.4.2 Selecting an assessment focus

The focus of assessment is the precise aspect of language that is being attended
to and assessed or measured. For example, in attending to how well a child recalls a
word and its meaning, the teacher may also decide to attend to pronunciation. When
the focus is discourse rather than words, as in the task analysed in Chapter 3, the
focus may be on interactional skills, such as turn-taking or answering questions,
on discourse organisation of extended talk or writing, or on the grammatical
complexity or accuracy of the language produced by a child (Skehan 1996).
There may be more than one focus; in the assessment of writing, both word-level
accuracy and discourse organisation may be considered. Knowing that each focus is to
be attended to will help ensure that assessment is fair and valid, because each can be
assessed separately and the effect of one on assessment of the other can be minimised.
When oral language is assessed, however, it can be difficult to observe or attend to
more than one assessment focus. One solution is to make a recording of children's
talk to assess later.
 An important consideration in assessing language learning is whether the
language is to be assessed in or out of the original learning context. For example, a child
will learn here, there, everywhere in the context of 'Old McDonald had a farm' - the
song is the original learning context of those adverbs. But knowing them and being
able to produce them in that context is only a first step in the learning. If we observe
whether or not the child sings them in the song, we have assessed his or her learning of
the words in the original learning context. Eventually we want the child to he able to
use those words out of the context of the song, to talk about things other than animal
noises, and in other structures than here a ... , there a ... (which is actually rather
archaic and hardly heard in everyday spoken English). Different words and phrases
in a learner's repertoire will move beyond their original context and into wider use
at different rates, and depending on the teaching. We cannot therefore generalise
about when to assess. We should note, however, that the teacher needs to be
aware of whether the words should be assessed only in their original learning
context, or whether the child can be expected to use them in other contexts and they
should therefore be assessed in other language contexts.
Possible assessment focuses for children's language learning are set out in
the table below. To be effective in assessment, each focus would need to specify the
actual items: words, discourse units, grammatical forms, learning skills.

Vocabulary Discourse Grammar

Oral Skills understanding precision in talk complexity of clauses and
meaning of words phrases understood and
and chunks fluency in talk produced

recall of words and response or initiation accuracy of morphology

chunks inconversational and syntax - used in
exchanges understanding,
knowledge of in production
thematic word sets understanding of
sentence-level
appropriate choice of discourse e.g,
words and chunks in instructions
discourse
understanding of creative use of whole-
pronunciation of oral texts - stories, learnt chunks
words and chunks songs, rhymes,
chants, dialogues - metalanguage -
overall meanings understanding, use

production of
extended discourse -
retold and original.
 Reading sight vocabulary understanding of working out accurate
Skills stories and other meanings
letter-sound links whole texts

sentence and text

level reading
strategies e.g,
prediction
Writing Spelling organisation of texts accuracy in use of
skills at sentence level and grammar
letter formation above

precision and
accuracy in
conveying ideas
Learning guessing words from context
skilss organising own work on tasks
setting targets for learning \
using targets for learning
work with a partner
work in a group
self-assessment skills
dictionary use

10.4.3 Assessment by observation

Observation is one of the most useful assessment techniques to use with children
because it does not disturb the children and allows them to be assessed in the process of
ordinary classroom activities. Skilful teachers are probably constantly observing their
pupils and adjusting their talk or the· lesson activities to take account of the feedback
they get from their observation: they may pause insinging a song to explain some of the
key words again because they notice that several children do not seem to understand
the meaning of what they are singing, or to practise the pronunciation of a particularly
tricky word on its own. They may repeat instructions for an activity one more time than
planned because they observe that not all the pupils have understood. They may stop as
they walk round the classroom and help one child who is observed to be struggling with
making the shape of one of the alphabet letters in their writing.
In addition to this continuous process of observe - notice - adjust teaching,
observation can be pre-planned for assessment purposes. The teacher selects the focus
of assessment and decides in advance that she or he will observe the children during
the next lesson to assess how well they have learnt that particular aspect of the language.
'Observation', as a metaphor to describe how we collect this type of assessment
information, builds on the idea of having a 'focus' and emphasises that we have to do
much more than just 'look' at what pupils say or do. Rather, we need to look very
carefully at the particular aspect of language that we are concentrating on, and use
our experience and knowledge about language and learning to guide us in what we
look for and how we interpret what we see.
Here is an example. A teacher of a class of seven year olds was eager to show
me, as a visitor, some of the theme-based work that her class had done. The children
performed a short play on the topic and sang some songs. If I had just looked at what
the children did, I would have seen a bunch of happy children having a good time
using English. By observing the children, I could see a more complex picture: some
knew all the words and carried the performance along, but some children clearly did
not know what to say and got by in the noise and excitement by moving their lips or by
just staying quiet. And, as in every class, there were one or two children who took
advantage of such a situation to get on with their own business: kicking the person
next to them or talking in the first language. Of course, it is not valid, reliable or fair to
assess children's learning from a public performance like this, without knowing
individuals and their learning histories. The point I want to make is that the
observation was 'finely tuned' by the focus; it was a very specific and goal-directed
way of looking and seeing. The example can be summarised as follows:
My assessment focus was something like Do all the children understand/the
meaning of what is happening?
The information that would give me some evidence of their understanding
was in their participation, what they actually did and said, and in something
much less tangible about how they participated, trying to assess from their
expressions whether they seemed to understand what they said.
My assessment technique was observation, i.e. close and purposeful noticing
of children's talk and actions.

10.4.4 Creating opportunities for assessment during classroom activities

When observation, or other technique, is used for assessment, there are several
factors to take account of in planning:
Who will be assessed? It may not be realistic to assess each child on the same
occasion, in which case a group of children can be selected for observation. By
focusing on six or seven children during one lesson, the information collected is likely
to be of better quality. Over three or four lessons, the whole class can be assessed.
Larger classes will need other techniques: perhaps observation of groups. Sometimes,
the teacher may choose to observe just one child - perhaps one who seems to be
struggling or one who seems to be bored with lessons. We need to be aware that
children who have picked up the language from home, TV, or their computers, may
find lessons too easy and misbehaviour can result from boredom as well as from
difficulties.
 When in the lesson will I assess? By matching the assessment focus against
the lesson plan, the teacher should be able to spot a point where assessment is
particularly appropriate. If oral language is the focus, the best time may be when
children are working in pairs rather than a whole class. Alternatively, during a
writing or making-and-doing activity, the teacher can talk quietly to individual
children and assess their learning of certain phrases and words through their ability
to use them in conversation.
Task-based assessment is a more formalised version of spotting opportunities
for assessment that has been developed in Australia for use with second language
learners of English (McKay 1995). Teachers and teacher educators worked
together to develop a set of tasks that have a clear assessment stage. Teachers' notes
show exactly what is to be assessed and how. Across all the tasks, the full range of
language skills and knowledge can be assessed, and enable teachers to describe a child's
proficiency by placing him or her on a bandscale. As Rea-Dickins and Rixon (1997)
suggest, this type of assessment methodology may have potential for children's
foreign language learning, not just in providing coverage of assessment through
classroom activities, but in the teacher development that occurs in the preparation of
assessment activities.

10.4.5 Record-keeping
The most common way of recording observations of children's performance is
through a checklist on which the teacher simply ticks when a pupil has achieved a goal.
Figure 10.1 shows an Assessment Chart from Playway to English1(Gerngross and
Puchta 1998). Such a chart could be designed for a unit of work by identifying the
language learning goals and converting them into assessment foci or statements
of performance, as in the left-hand column. I would suggest? however, that 'recognise
and use of language' be separated, since, as we saw in Chapter 3, the very different
demands of understanding and speaking can lead to them appearing in performance at
different times.
Although these checklists are easily managed and convenient, they do limit the
amount of information that can be recorded about each child. An alternative is to
have a loose-leaf record book with a section for each learner. The section can contain
formal records of assessment, such as results of a miscue analysis (Chapter 6) or rating
of oral language assessment. At regular intervals, say, every two weeks, the child can
be observed during a lesson, and notes jotted down about their participation, language
use, confidence, social skills and anything else that is of interest. Over the year, the
formal records and informal notes will give a useful picture of the overall development
of the child as a language learner.
 10.5 Self-assessment and learner autonomy
In this section we look at how far young learners are able to assess their
own language learning. Self-assessment can help learners to understand more
about the language learning process and to become more independent.

10.5.1 Self-assessment
The benefits of self-assessment are easily stated, but need some thought as to
their applicability for young learners of foreign languages.
Through self-assessment:
• learners can understand more about the learning process;
• learners can be motivated towards more involvement in their
learning;
• teachers can understand more about individual pupils;
• learners will be better prepared to carry on learning, beyond the
classroom;
• a more equal relationship is created between teachers and learners.

In Vygotskyan terms, a pupil who learns to assess his or her own work moves
from being 'other-regulated' to being 'self-regulated' or autonomous. Using the
analogy from Chapter 1, of the child learning to feed him or herself, the other-
regulated child depends on the other (parent or other adult) to organise and control the
feeding situation, by presenting the food, holding the spoon, moving food from bowl to
mouth; the self-regulated child will hold the spoon, load the spoon and manipulate the
moving of the spoon into the mouth. Some years later, the other stages of the process
will come under the child's control when she or he learns to shop, prepare and cook
food. The other-regulated language learner depends on the teacher to decide what is
to be learnt, to present the language, to decide on activities, to control activities and
to evaluate how well the language has been learnt through the activities: The self-
regulated learner will take over parts of this process, and be able to adjust them to
suit meaning styles and preferences.
It is commonly recognised in today's world that autonomous and self-
regulated learners will be at an advantage in continuing to learn and adjust
throughout their lives as technology and information develop rapidly and
continuously. Learner autonomy then is 'a good thing' and to be encouraged, but
how realistic is this in classes of five year olds? My own view is that we tend to
underestimate the potential for self-regulation in our children, seeing them too often as
blank sheets to be written on, empty vessels to be filled, or wild and in need of
taming. Young children can, within the limits of their cognitive development, be
helped to organise their resources, both internal and material. It is not unusual
to see classes of five year olds who know where to keep their books and papers, how
to tidy the classroom, how to organise their work and how to decide in what order
they will complete their classroom activities (see Ellis 1991; Brewster, Ellis and
Girard 1992).
In the foreign language classroom, the language used to organise such training
will probably need to be the mother tongue, and so further decisions must be made
about the value of the time required, and balancing between spending the time on
developing learner autonomy or on language learning. In deciding how to develop
learner autonomy, it will be useful to tie increased responsibility to the language
content: for example, the phrases I like I don't like often appear very early in a syllabus.
With these simple phrases, learners can evaluate how much they enjoy different
types of activity or to pick out the good points of a story. Similarly, learners can easily
cope with I learnt and use this to recall what was learnt in order to reflect on it.
Children of seven and eight years of age can begin to understand criteria
for good performance or production, and if these are simply phrased may be able
to use the foreign language. Once, when working with native speaker seven year olds,
I helped them to reflect on their interactional skills by recording them talking and
getting them to listen to the talk; they identified problems such as not listening to each
other, interrupting, or talking too long and off the point, which were then turned
into a list of 'Our rules for talking'. Making the list also involved teaching the children
a new word: relevant.

Our rules for talking

(1) Don't all talk at the same time
(2) Talk one after the other round the circle
(3) Wait until somebody has finished
(4) Don't tell Stories that aren't useful or relevant

After each further discussion session, they gave themselves marks out of 10
for how well they had obeyed their own rules. This is an example of helping children
not only understand someone else's assessment criteria, but also to set their own. In
foreign language classrooms, parallel opportunities may arise when children are
giving a public performance such as a short play or presentation around a theme
(Chapter 7). After listening to each other, the class can be asked what was good and
what could be better, and the ideas they suggest written up as criteria for a good
performance.
When children reach the stage of writing, samples of good work can act as
benchmarks against which they can compare their own writing. Reading other
children's work extends reading skills as well as demonstrating target levels of
achievement. Peer assessment is a good half-way stage towards self-assessment.
Pairs of children can swap pieces writing and be asked to tell each other good
points and things to improve on a next draft. Again, this process will need to be
modelled by the teacher first, and should if possible be done through the foreign
language

10.5.2 Goal setting

Being able to set realistic and useful goals for one's own language learning is one
of the skills of autonomous learners and is part of the cycle that links self-assessment to
learning. Even very young children can be helped to set goals for themselves by, for
example, choosing five words out of eight to learn at home, and then testing each other.
As they get older, so the length of the time for which the goals are set can be increased
from a few days to a term at a time. At the beginning of the term, the teacher may explain
the goals that she or he has set for everyone, and then ask the children to select a further
set of goals for themselves from a list of possibilities; children who need to practise
writing can choose writing goals; those who want to read more may decide they will try
to read four books in the term. At the end of term, or other time period, it is important to
evaluate how much progress has been made towards the goals, and to discuss why some
have been achieved but others have not.

10.5.3 Portfolio essessment

A portfolio is a collection of examples of work that, as a collection, reveal both
the capability and the progress of a learner. Artists, photographers and architects
typically build up portfolios, in which they put together pictures they: feel best represents
their style and skills. They then use the portfolios to demonstrate what they have to
offer to potential customers or employers. Applied to language learning, a portfolio
would include such things as samples of writing and lists of books read (O'Malley and
Valdez Pierce 1996). Portfolios can link assessment with teaching and with
metacognitive development through including pupils in the evaluation of performance,
thereby developing skills in pupils' self-assessment. Pupils are involved in deciding what
to include in their portfolios and in assessing or evaluating the pieces of work according
to clear and explicit criteria.
A limitation of portfolio assessment is its application to oral skills
development. Being paper-based, it is more suited to collecting written texts.
Children can include their self-assessments of oral language activities and progress, but
it is more complicated to include samples of talk. Perhaps as computers with CD ROMs
become more accessible, childrenwill be able to build up computer based portfolios with
recorded samples of speech as well as scanned-in or digitally photographed texts.
Meanwhile, portfolios around literacy skills development seem to offer interesting
possibilities (see O'Malley and Valdez Pierce 1996, Chapter 3 for details on how to build
up portfolios with children). We can also note how portfolio assessment allows for
much greater child involvement in the process of recording progress and
achievement, and try to find such techniques in oral assessment.

10.6 Use of assessment information

Earlier in the chapter we established the principle that assessment should
support learning and teaching. In order to make it do this effectively, we need to
consider how the outcomes of assessment are used to generate various forms of
feedback that can change teaching and learning.

10.6.1 Outcomes of assessment and uses

Summative assessment techniques, such as tests, produce outcomes in the
form of grades, marks and rankings of students, but little more qualitative
information that can be used diagnostically. Summative outcomes can be used to
evaluate the effectiveness of a course or programme, along with other information such
as student and. teacher evaluations, and course documentation (Rea-Dickins and
Germaine 1982). Rankings of students are used for placement purposes: deciding
which schools, classes, sets or groups students will be placed in.
Leaving this type of assessment purpose aside, we can move to
formative assessment that is intended to make a difference to teaching and
learning. Observations, portfolios, checklists, rating scales each offer information
that can be converted into feedback for parents, pupils, and other teachers, and
that influence how the next lesson or unit is planned by the language teacher.

10.6.2 Making feedback helpful to learners

If assessment feedback is to be helpful to learners and improve their
learning, it needs to be specific and detailed enough to make a difference and, equally
importantly, it needs to be related to a target performance or understanding towards
which the learner can move. The target performance, which may be presented as a list of
criteria, a rating scale, or as examples of exemplary performance, offers learners the
opportunity to see what they are aiming at. Feedback should also help learners to
compare their current performance against the target performance, and to close the
gap between them (Gipps 1994). The process in which assessment and feedback can
scaffold the learner to better learning is summarised below:

learner understands the target performance

learner compares target and current performance

learner closes the gap between target and current performance

 The target performance could be modelled by the teacher, or on video or audio
cassette, or by pupils who are at a higher level. Murphey( 2000) suggests that 'near
peer role modelling' will help move children through their zone of proximal
development, not just in relation to assessment but more generally in learning
language, and not just in terms of the language but also in learning strategies and
metacognitive development.
The teacher of young learners can intervene to help the learner 'close the gap' in
several ways:
Corrective feedback aims to help pupils correct their language use towards
the target language. It will explain why correct responses are correct and incorrect
ones are wrong; pupils will be shown a model of the correct responses: In language
learning, corrective feedback will be primarily concerned with accuracy. With
children, and probably older learners too, it is not necessary always to point out
errors; often, repeating what a child says with the correct form stressed will work as
corrective feedback. The pupil should be allowed to repeat the correct form, and
should receive some signal that this is now correct.
Example
Pupil : I come to school at eight o'clock
Teacher : I came to chool at eight o'clock
Pupil : I came to school at eight o'clock
Teacher : Well done!

Evaluative feedback includes a judgement on the pupils' performance.

Examples
1. Quoted from a primary school lesson (LI):
That one was quite easy. Quite a few got that one right.
Now the next one was very hard. You had to think.
2. That was very good - I liked the way you said the
sentences clearly.
Notice that in the first example the teacher did not help the pupils to do better
next time; she just gave an evaluative comment on their performance. In the second
example; the teacher pointed out something specific that was good.
Strategic feedback offers advice on what to do to improve performance.
Examples
1. to help a child say <the> /The/
look at my tongue - put your tongue on your teeth - the
2. to increase fluency in speaking in a dialogue, pupils might be advised to say the
phrases to themselves 'in their heads' during the preparation stage, or just before they
speak. (This is called mental rehearsal).
 In the early stages, some of this feedback will need to be done through the mother
tongue, although a rating scale or strategies for improving performance can be introduced
to the children in the foreign language. If the same phrases are always used, they can
work as feedback.
10.7 Messages from assessment
To conclude this chapter, I want to highlight the idea that assessment practices not
only determine children's futures and how their time is spent, but also carry messages for
children about what, parents and teachers consider important in language learning and in
life.
Feedback practices show pupils directly what is valued in the work that they do. If
spelling and neatness are what the teacher attends to, then the child will get the message
that this is what matters in writing; if a teacher always praises the children, even when
they are not making an effort, they quickly learn that the praise is hollow. If on the other
hand the teacher knows a child's capabilities, recognises when she or he is trying
especially hard and offers praise and supportive feedback, the child learns that his or her
learning matters to the teacher and that it is worth struggling. If all other subjects count
for selection but the foreign language exam does not, then language classes are not likely
to be highly motivated.
Educational research demonstrated long ago that children live up to the
expectations of their teachers, whether those are low 01 high. Expectations are perhaps
more clearly revealed through assessment practices thari anywhere else. The negative
washback from assessment to teaching and learning which I described in the opening
section can be made to work positively by giving weight to what matters in language
learning. For young children, what matters is a solid base in spoken language, confidence
and enjoyment in working with the spoken and written forms of the language, and a
good foundation in learning skills. We should be searching out assessment practices that
will reinforce the value of these to learners and to their parents.