Language Policy Division Division des Politiques linguistiques
Note : This text is included in the series of studies on “EVALUATION AND ASSESSMENT WITHIN
LANGUAGE(S) OF EDUCATION”
The Challenge of Assessment within Language(s) of Education
Mike Fleming, University of Durham, United Kingdom
The purpose of this paper is to provide an overview of some of the key issues and
challenges related to evaluation and assessment within the domain of language(s) of
education. It incorporates elements of the previous paper distributed at the Strasbourg 1conference in 2006 but is also addresses new more practical issues related to (a) the
possibility of situating assessment for language as school subject (LS) within the wider
construct of language(s) of education (LE); (b) the need to reconcile portfolio
approaches with more formal testing. Assessment is important but it is not without
controversy and it can easily lead to polarised views and unhelpful tensions. This is
particularly the case in the context of language as school subject (LS) because of the
diversity and complexity of its aims. Constructive debate around differences of opinion
is always helpful, but too often disagreements about assessment become entrenched
and unproductive. This happens for a number of reasons, including:
? a failure to recognise that assessment needs to fulfil a wide range of legitimate
? an assumption that a single assessment tool will be able to serve all needs;
? a lack of awareness that it is the use made of assessment, not necessarily the
assessment process itself, that will largely determine its impact;
? a tendency to search for universal solutions to assessment issues and neglect the
significance of context.
Teachers of language as school subject are sometimes hostile to the idea of large-scale
or formal testing on the grounds that it diminishes the subject and ignores the
significance of context. This view needs to be considered.
Section One will consider a range of different purposes of assessment based on
different potential audiences. An ideal assessment strategy would meet the needs of all
those interest groups. Section Two will examine different approaches to assessment in
relation to language as school subject (LS), including the value of portfolio assessment.
Section Three will consider the implications of situating the assessment of LS within a
broad strategy of assessment within languages of education (LE). Section Four will
develop further the concept of an integrated approach to assessment which seeks to
reconcile its different purposes within an assessment strategy. It will consider the view
1 Intergovernmental Conference on “Languages of Schooling: towards a Framework for Europe”, Strasbourg,
Council of Europe, 16-18 October 2006. See the Report on: www.coe.int/lang
that a portfolio approach is not necessarily in conflict with more formal testing, as is often assumed.
1. Purposes of assessment
In its simplest formulation, assessment provides information on whether teaching/learning has been successful. However the information it provides has a number of potential different audiences whose precise requirements may vary. Classroom teachers need regular information on how pupils‟ knowledge, skills and
understanding are developing, both to inform how they should adjust their teaching and to determine what kind of feedback is needed to improve pupils‟ learning. On the other hand, school principals and policy makers need additional, broader information on the quality of education in a school or country. The sort of comparative data required for this purpose needs a high level of reliability and uniformity. In the case of language as school subject this requirement is challenging because it is difficult to create tests which are manageable but at the same time faithful to the aims of the subject. Employers and society at large also need reliable information which can help certify achievement and provide a basis for selection. Parents too require information which can help them understand their children‟s achievements and limitations. Learners themselves need to know how they are progressing and how to improve their performance but they may need to be protected from the potentially demotivating effects of negative assessment.
The concept of „accountability‟ when used in relation to assessment usually refers to the imposition of systems of assessment external to the learning process as a form of „policing‟ of standards to ensure that the education system is functioning effectively.
But the term may be employed more broadly and more positively than this, referring to the different obligations that are relevant to all. Teachers have a responsibility to the learner but also to the needs of the wider society. Policy makers clearly have a duty to the public and need to ensure that the education system is delivering results but they also have responsibilities to the individual learners and need to consider consequences of policies in those terms. The concept of accountability interpreted in this way will take people outside of vested interests in order to see the larger context. Accountability needs to be linked with a process of sharing perceptions and fostering understanding. It is important therefore not to exaggerate differences between different potential „stakeholders‟; what all parties have in common is a fundamental concern that assessment should help raise achievement and improve learning. A starting point for resolving tensions related to matters of assessment is to develop understanding of other points of view. A key challenge is to develop a system of assessment that acknowledges the different functions of assessment and it helps to see these as complementary rather than being in opposition to each other. 2. Approaches to assessment
The different purposes of assessment lead to different approaches to assessment. Traditionally assessment of language as school subject took a very simple form: pupils were given a narrow written task which was then awarded a grade or mark. This allowed them (and the teacher) to make a judgement of how they ranked in relation the rest of the group (normative assessment). However the absence of clear criteria meant that the information rarely gave an indication of how they could make progress in their learning. Also the test itself often embodied a very narrow conception of what competence in language entailed; it often centred on knowledge of language form and structures (syntax and grammar) and a narrow range of language uses (often only a written form of essay). The implicit understanding of what reading literacy involved was also very narrow, often involving just decoding and literal understanding. In traditional approaches of that kind there was unlikely to be any attention to oral work, to a range of writing purposes, to a wide range of reading and response to reading. On
2 the other hand, the advantage of a fairly narrow approach to assessment was that is
was easier to provide reliable outcomes; the more complex the system of assessment becomes, the more difficult it is to ensure that the award of grades or marks for particular outcomes are consistent.
In approaches to assessment, two central tendencies emerge which are relevant to language as subject. One places emphasis on the assessment of learning where
reliable, objective measures are a high priority. The focus here is on making summative judgements which in practice is likely to involve more formal examinations and tests with marks schemes to ensure that the process is sound. An alternative approach is to change the emphasis from assessment of learning to assessment for learning, implying
a more formative approach where there is much more emphasis on feedback to improve performance. The approach here might be through course work and portfolio assessment in which diverse information can be gathered which reflects the true broad nature of the subject.
Portfolio assessment has a number of advantages for language as school subject; for
example, it can motivate and empower the learner, it can provide samples of performance collected over time, evidence of use and awareness of process. Portfolio assessment incorporates evidence derived from more realistic tasks in meaningful contexts, rather than relying on artificial, decontextualised tasks undertaken in timed conditions. A further advantage of this approach is that it can embody different forms of self-assessment which can also be helpful ways of motivating learners and having them reflect on their progress. Self-assessment encourages pupils to take
responsibility in the learning process although it is advisable for them to be trained in self-assessment techniques for this to work effectively. The difficulty with portfolio assessment if it is conceived only as the accumulation of evidence produced in informal
settings is that is does not easily satisfy demands for reliability. Work which has been produced over an extended period of time, with formative guidance from the teacher and collaboration with classroom peers is not always convincing evidence of competence.
The broader the approach to assessment (incorporating the judgement of a range of different performances in different contexts), the more it can be said to constitute a meaningful assessment of performance in the subject. However, as suggested, tension emerges because it is sometimes difficult to compare, with any degree of accuracy, the results drawn from broad approaches to assessment. The quest for „objective‟ and reliable methods of assessment driven by narrow ideas of accountability brings with it a number of dangers. So called „teaching to the test‟ may not be a problem if the tests are sophisticated and wide-ranging but there may be practical difficulties in administering those that are too complex. If the tests are too narrow and simplistic then this may have an adverse effect on the teaching. Here then is one source of polarised opinions, one stressing the importance of objectivity, reliability and summative judgement, the other more tolerant of subjectivity in order to ensure that the assessment approach is faithful to the complexity of the aims.
A key concept is embodied in the notion of ‘transparency’, the view that those being
assessed are aware of the criteria which are being used to make judgements about them and how those judgements are made. Knowledge of criteria can help
performance and improve motivation but once again, in the context of language as subject, the issues are more complex than they first seem. A common assumption is that pupils learn best when they know what they are trying to achieve and why. While this view is largely true, there are exceptions. Because the development of language can in some ways be described as a „natural‟ process learners do not always need to be fully focused on specific aspects of their performance in order to improve. In fact too much focal awareness on performance can make them too self-conscious: speakers can appear too groomed and artificial; the writer who has been told to strive for effect by using more adjectives may develop a highly artificial and awkward style. These insights
do not negate the importance of transparency as a principle but highlight the fact that in pedagogical practice the principle needs to be interpreted and implemented with care.
The use of competences to describe outcomes which can be assessed has developed
considerably in recent years. It is worth noting that the term ‘competence’ is used by
writers in different ways which can be a source of confusion. Sometimes it is used in a very general way as a synonym for „ability‟ or capability as in „language competence‟. Other writers use the term to describe broad language modes or domains such as
reading, writing, speaking and listening. More commonly however „competences‟ refers to the specific actions which a learner must perform and which in turn can be assessed to demonstrate achievement in a subject. Advocates of using competence statements for assessment purposes and syllabus design see their value largely in bringing clarity and transparency to the specification of learning outcomes. Critics of a competence approach take the view that performance statements are too narrow and specific, and do not reflect the range and subtlety of what is involved in language development. There are parallels here with differences of opinion over the relative merits of portfolio and formal testing. The balance of advantages and disadvantages needs to be considered.
One of the challenges posed by assessing language as school subject (LS) is that the
2content is so varied and complex (see the paper by Florentina Sâmihaiăn) which is in
turn a reflection of the complexity of the aims. Each mode of „writing‟, „reading‟, „speaking „and „listening‟ can be broken down into further areas. A subject that is so multidimensional raises the question as to whether an assessment task in one area is representative of achievement in the subject as a whole. For example, it is fairly safe to assume that performance in speaking is not necessarily indicative of reading competence. On the other hand it is less clear whether it is necessary to assume that reading ability varies in relation to texts of different types (fiction, non-fiction, media). The assessment of reading can easily remain at a surface level only addressing recall or literal comprehension rather than deeper understanding. Multiple choice questions on a text are easy to mark and may yield high reliability (in the technical sense) but are less equipped to assess the learner‟s deep and individual response to a text. Even with very
young children the reading process is more than simply decoding text and any system of assessment needs to reflect that fact. It is the complexity of the subject that accounts for the hostility teachers of languages as school subject sometimes express towards large-scale and formal testing because only a fairly narrow range of competences can be assessed in a single test.
A similar question arises in relation to writing. Do pupils need to be assessed on a range of different tasks reflecting the fact that writing exists for different purposes and for different audiences? The assessment of writing appears to be more straightforward than that of reading and speaking because at least there is always a product which can be referred to after the event. But there is a major challenge in determining what criteria should dominate in making a judgement (for example the accuracy of the writing as opposed to the impact of the content) and whether the criteria should change in relation to different types of writing. It is also wrong to assume that the only response to pupils‟ writing comes when it is completed. Formative assessment in the form of a dialogue about the work in process is an important way of improving standards. Assessment is always a selection and therefore can unwittingly value some aspects of the subject more than others.
2 Sâmihaiăn F. Content considerations for a framework of reference for Language(s) of School Education,
Speaking and listening is extremely difficult to assess because, even more than other aspects of LS, performance varies with the theme, context and level of motivation. Poor performance in oral work is often to do with the nature of the task which has been devised, the atmosphere of the classroom and the dynamics of the group rather than the competence of the pupils; to provide a valid assessment of speaking and listening, evidence needs to be drawn from a variety of situations. Some might argue that the assessment of speaking and listening is so complex and context specific that it should not be assessed formally. However because assessment so often determines the curriculum and the way it is taught there are arguments to suggest that speaking and listening should be assessed despite the difficulties; the ability to articulate a point of view orally and to argue a case are essential skills for meaningful participation in a democracy.
An ideal assessment system would reflect the full complexity of language as school subject (LS), and would motivate learners by giving useful feedback, while also providing other stake-holders (e.g. policy-makers and employers) with the information they need. An integrated approach to assessment would ensure that the different purposes and approaches are balanced so that no one priority has adverse and undue influence on the system as a whole.
3. Language(s) of Education (LE)
So far consideration has been given to the challenges presented by assessment of „language as school subject‟ (LS). Is there any advantage in considering the assessment of LS within a broader framework of language(s) of education (LE) which incorporates „language across the curriculum‟ (LAC) and foreign language learning (FL)? „Languages(s) of education conceived in that way is not a subject but an umbrella construct; it is fairly clear therefore that the notion of examining or testing language(s) of education is entirely inappropriate. However it is conceivable to profile a pupil‟s competence in
language(s) of education by assembling their competences and achievements in a range of domains. The present Council of Europe‟s European Language Portfolio (ELP) is a useful model to demonstrate the type of instrument which could be considered. This approach has a number of potential advantages. It takes seriously the concept of plurilingualism in the school context: a pupil may be a low achiever in language of schooling taught as school subject but be fluent in two or more other languages. Thus the deficit model which is often applied to assessment of LS is replaced by a more positive recognition of achievement. It would still be important to be able to identify an individual‟s competence in aspects of LS as a component within a broader profile,
but being able to situate that description of competence in a larger context could have a positive impact on motivation and self-esteem.
Could portfolio assessment extend to embrace language competence in other subjects, to include the dimension of language across the curriculum? This is a practical challenge but perhaps not insurmountable. Some uses of language (e.g. giving presentations, writing formal reports, reading for information) are clearly required and demonstrated within different subjects and it is not inconceivable that those subjects should make a contribution to a pupil‟s overall language profile. There is an argument to suggest that if developing language competences across the curriculum is to be taken seriously it must have some impact on how language use is assessed. Underlying questions have to do with subject boundaries and the degree to which competence in language use can be easily separated from the context in which it is used; learners of history could also in some sense be said to be learning the language of history. 4. Integrated assessment
As described earlier a key challenge and source of tension in relation to assessment is to satisfy the different purposes assessment is expected to fulfil and to provide appropriate information for the different interest groups. As suggested, teachers of
language as school subject are sometimes hostile to large-scale, formal testing because
of the perceived narrowing of the subject. There are a number of points to consider in
relation to this view. The testing „industry‟ has become increasingly sophisticated in
recent years and designers of assessment tasks and items are more adept at addressing
issues of validity and reliability. That does not mean that formal tests can address all
language competences but once that fact is recognised then the results of such tests
may make a useful contribution to the overall profile of a pupils‟ competence in language use. Large-scale testing both at national and international level is a fact of
modern life. Policy makers need information on levels of proficiency achieved by
groups in schools, education authorities or countries. It is possible to discharge this
„evaluative‟ function of assessment by a process of sampling rather than by using
summative data derived from an entire population, but it is unlikely that the quest for
information of this kind will diminish. When assessment data is used to compare the
progress of different cohorts of pupils, the use of a „value-added‟ approach which takes
account of the different base-lines from which the pupils are progressing is becoming
more common. All of this information can be informative for teachers as long as the
limitations of the data are recognised.
The concept of an integrated approach to assessment is intended to counter the
tendency towards polarised views. As suggested above, portfolio assessment has many
advantages but without information based on some form of „controlled‟ performance it
will struggle to make convincing claims to reliability. The incorporation of specific
results derived from appropriate tests designed to assess specific (not all) language
competences into a portfolio can help strengthen both the validity and reliability of the
assessment approach. The key may be to think broadly in terms of an assessment
strategy which makes use of a variety of assessment tools, rather than assuming that
one assessment tool will fulfil all purposes.
A sample outline of a Languages of Education Portfolio based on the present Council of
Europe model is given below:
Languages of Education Portfolio
? to provide a record of achievement in all languages of education
? to provide evidence of developing language competences needed for
? to motivate learners to extend their range of language competences
? to ensure that the language needs of all pupils are being addressed
Introductory profile of developing
competence in a range of languages
– linked to other assessment
Languages Language Language as School Subject (LS) (FL etc.) across
A variety of Curriculum Examples: specific test (LAC) reading log incorporating literature and Evidence of results both local non-fiction; evidence of reading for language and national. competence different purposes; writing samples; video Evidence of from other clip of discussion of literary texts; self-achieving subjects: e.g. assessment. threshold video clip of competences presentation in through geography; recognised discursive reporting writing in mechanisms. Language Biography history; Details of technical report certificates, from science. Personal language autobiography diplomas etc. incorporating meta-cognition of language
processes – dialect and accent, language
and identity, etc.