Lessons from around the world: How policies, politics and cultures constrain and afford assessment practices

Abstract

This article outlines the main assessment traditions in four countries – England, France, Germany and the United States – in order to explore the prospects for the integration of summative and formative functions of assessment during compulsory schooling. In England, teachers' judgments do feed into national assessments, at 7, 11, 14 and 16, but concerns for reliability and accountability mean that such judgments are made in a way that has little impact on learning. In France, teachers have no involvement in the formal assessment of students, and, possibly as a result, have been free to concentrate on the use of assessment to serve learning. In Germany, faith in the education system has been considerably undermined by recent unfavourable international comparisons, although faith in the ability of tests both to measure learning accurately and to allocate students to different educational pathways appears to be unshaken. In the United States, multiple demands for accountability at different levels of the system have resulted in multiple assessment systems, but these tend to be focused on measuring the amount of learning that has taken place, providing little insight into how it might be improved. It is concluded that the effective integration of formative and summative functions of assessment will need to take different forms in different countries, and is likely to be extremely difficult.

Full text

Lessons from around the world:
how policies, politics and cultures
constrain and afford assessment practices
Paul Black and Dylan Wiliam*
King’s College London
This article outlines the main assessment traditions in four countries – England, France, Germany
and the United States – in order to explore the prospects for the integration of summative and
formative functions of assessment during compulsory schooling. In England, teachers’ judgments
do feed into national assessments, at 7, 11, 14 and 16, but concerns for reliability and accountability
mean that such judgments are made in a way that has little impact on learning. In France, teachers
have no involvement in the formal assessment of students, and, possibly as a result, have been free
to concentrate on the use of assessment to serve learning. In Germany, faith in the education system
has been considerably undermined by recent unfavourable international comparisons, although
faith in the ability of tests both to measure learning accurately and to allocate students to different
educational pathways appears to be unshaken. In the United States, multiple demands for
accountability at different levels of the system have resulted in multiple assessment systems, but
these tend to be focused on measuring the amount of learning that has taken place, providing little
insight into how it might be improved. It is concluded that the effective integration of formative and
summative functions of assessment will need to take different forms in different countries, and is
likely to be extremely difficult.
Keywords: accountability, comparative education, formative assessment, summative
assessment
Introduction
Public schooling takes very different forms in different countries. In most developed
countries, attendance is compulsory from the age of 5 or 6 to the age of 15 or 16,
although many systems allow alternatives such as home schooling. In addition, most
students attend some form of nursery or pre-school, and most students continue in
education beyond the end of compulsory school. In most countries private schools
*Corresponding author. ETS, Rosedale Road (ms 04-R), Princeton, NJ 08541, USA.
Email: dylanwiliam@mac.com
The Curriculum Journal, Vol. 16, No. 2, June 2005, pp. 249 – 261
ISSN 0958-5176 (print)/ISSN 1469-3704 (online)/05/020249–13
ª2005 British Curriculum Foundation
DOI: 10.1080/09585170500136218

operate entirely outside the state school system, but in others (Australia, for example)
some private schools are supported with public money. In some countries (England,
for example) parochial schools are part of the state system, while in others (such as the
United States), they are outside it.
Attitudes towards selective schooling differ markedly. In most industrialized
countries, the vast majority of students attend schools that are open to all; in
Germany, at the end of four years of primary schooling, however, students are
allocated to different types of schools on the basis of their academic achievement. In
England some local education authorities also retain selection at the end of primary
school, although the proportion of students attending selective secondary schools
ranges from 4 per cent in some districts to over 30 per cent in others. In Japan, the
most prestigious schools also select on ability but for the others there is little, if any,
competition.
In England, France, Japan, New Zealand and Sweden there is a national
curriculum while in Australia and Germany the curriculum is the responsibility of
regional bodies. In the United States, school curricula are nominally the
responsibility of approximately 17,000 directly elected local boards of education,
but 49 of the 50 states (the exception is Iowa) have laid down state curriculum
standards to which local curricula are required to conform.
Curricula invariably include reading and writing the mother tongue, mathematics,
science and the humanities, although there are differences in how these are organized.
For example, the earth sciences are treated as part of geography in England but as
part of science in the United States. Chemistry and physics are separate subjects in
upper secondary schools in England, but are taught together in France. Moral
education is an explicit subject in some countries (e.g. Japan) but subsumed in other
subjects in other countries. Religious education is compulsory in state schools in
some countries (e.g. England) and illegal in others (e.g. the United States). In some
countries (e.g. England, Germany and the United States) the upper secondary school
curriculum allows students considerable choice in terms of what to study, while in
others the choice is between pathways, within which there is little choice (e.g. France,
Sweden).
Policies regarding textbooks also vary. At one extreme, in Japan, textbooks are
required by law, and textbooks must be approved by the government. In Germany,
each of the 16 La¨nder lays down regulations for the adoption, introduction and use of
textbooks. In the United States some states (e.g. California, Florida, Texas) must
approve textbooks before they can be used, while in others, textbook adoption is a
matter for individual school districts. In England, the decision about which textbooks
to use is a matter for the school, and sometimes even the individual teacher.
Given these differences in the fundamental characteristics of education systems, it
is hardly surprising that there are also significant and deep-rooted differences in the
assessment systems of different countries. One tradition in the field of comparative
education is to try to draw out the common trends in different countries, thus
relegating the local differences as being of secondary importance. The aim of this
article is slightly different. It is to try to bring these differences into the foreground in
250 P. Black and D. Wiliam

order to attempt to understand the assessment traditions in each country within their
cultural and political contexts, and then to attempt to reflect on the prospects for the
development of effective formative assessment. To do so, this article focuses on four
countries, England, France, Germany and the United States, and presents a
description of the key features of their assessment systems together with a discussion
of how the assessment systems are influenced by their cultural and social contexts.
England
When the government announced its intention to introduce a national curriculum for
England (and Wales) in 1987, it was made clear that the assessment of students’
achievement at the ages of 7, 11 and 14 would be through a combination of teacher
judgements and externally set tests. Currently, all national curriculum subjects are
assessed by teacher judgement at the ages of 7, 11 and 14, and in addition there are
formal tests for English and mathematics at ages 7, 11 and 14, and for science at 11
and 14.
This attention to the role of teachers’ judgements as part of a summative
assessment has obscured the role that assessments can play in supporting learning at
both the practical and policy levels. In terms of practice, the lack of central guidance
about what would constitute adequate records of student achievement resulted in a
myriad of complex ‘home-grown’ record-keeping systems that were ill suited to
supporting learning but provided comprehensive evidence of students’ achievements.
At the policy level, the failure to appreciate that the ‘teacher assessment’ mandated as
part of the statutory assessment arrangements was entirely summative in nature led to
a situation in which ‘formative assessment’ was discussed only twice by the
government agency responsible for the national curriculum and its assessment
during the first seven years of implementation (Daugherty, 1995). Even recent
pronouncements about the importance of assessment for learning have tended to
reinforce the collection, rather than the use, of data.
Teachers’ judgements also feature in the national assessments for 16- and 18-year-
old pupils. There are two main qualifications systems for students in this age group.
The traditional route, and the one taken by most students aspiring to go on to higher
education, has been to take the General Certificate of Secondary Education or GCSE
at age 16 and the General Certificate of Education Advanced Level (usually
abbreviated to ‘A-level’) at age 18.
The GCSE typically takes the form of a combination of timed written examination
papers (one or two papers for each subject, lasting up to two-and-a-half hours each)
and assessment by teachers of work completed by the student in school and at home.
The weight of the school-based component ranges from 20 per cent (in mathematics)
up to 50 per cent (in technology), and the results are graded on a nine-point scale: U,
G, F, E, D, C, B, A, A*(highest).
For the final two years of secondary education, students following the academic
route specialize in a small number of subjects (typically four or five) and, after a year’s
study, most specialize further by concentrating on just three subjects for the final year
Lessons from around the world 251

of schooling, at the end of which they take the A-level examinations. Alongside these
academic qualifications, and within the same integrated qualifications framework, is a
range of vocational qualifications. When originally developed in the 1990s, the
vocational qualifications made much greater use of teachers’ judgements in
determining final outcomes, and many were designed specifically so that the
assessment evidence generated for summative purposes could also be used
formatively. However, as these vocationally oriented programmes were integrated
into a comprehensive qualifications framework, assessment practices moved much
more towards a model in which test and examination scores are combined with
teachers’ assessments of formal ‘set pieces’ undertaken during the programme.
Thus, in the 5 – 14 general curriculum, and in both academic and vocational routes
for 14 – 19 year olds, the involvement of teachers in the assessment process originally
held out the promise that the assessments undertaken by teachers as part of their
normal teaching could serve both formative and summative functions. However, by
insisting that these teacher assessments should be judged according to the definitions
of reliability that had been developed for traditional tests, combined with a profound
lack of trust in teachers, the potential for teacher assessments to support learning has
been substantially eroded. Some teachers have been able to develop assessment
methods that integrate both formative and summative functions effectively, but, even
here, the emphasis has been on maximizing test scores rather than meeting students’
learning needs, so that, in a parody of Gresham’s law, summative has driven out
formative. As one teacher put it: ‘It is a bit depressing, isn’t it?’ (Black & Wiliam,
2004).
France
In France, there is a national curriculum for all levels of compulsory education that
also extends to pre-school and post-compulsory education. Schools do have a limited
degree of flexibility in the allocation of time between different subjects, and teachers
are expected to use their own creativity to bring out the best in every student. At the
age of 15, students can take the brevet des colle`ges, or for those in vocational
programmes, the certificat d’aptitude professionelle (CAP) and the brevet d’e´tudes
professionelles (BEP). Originally intended as a school-leaving examination, a pass in
the brevet was needed for entry into the lyce´e but is no longer required. However, the
brevet is now seen as good practice for later examinations, and although voluntary, is
taken by 99 per cent of students in the last year of the colle`ge, and 75 per cent achieve a
pass.
The main requirement for entry to higher education is a pass in the baccalaure´at,
taken at the end of the third year of the lyce´e. This has three variants: ge´ne´ral
(academic), technologique (technical) and professionel (vocational). Although originally
intended only for the highest achieving students, the proportion of students taking the
Bac has risen over the last thirty years from 30 per cent to 90 per cent, and about 75
per cent obtain some sort of pass. With the exception of foreign languages, all subjects
are assessed almost exclusively through a series of examination papers, each lasting
252 P. Black and D. Wiliam

from three to three-and-a-half hours, and taken over a period of four days. For most
universities, entry is based solely on the performance on the Bac, but for admission to
the institut universitaire de technologie (polytechnic), school reports are also taken into
account. For students wishing to enter the grandes e´coles – prestigious engineering
schools – students take highly competitive entrance examinations. Students prepare
for these examinations by attending classes preparatoires at institutions that are
administered within the secondary school system. If students fail to gain admission to
one of the grandes e´coles, some, but not all, universities will take the scores on the
entrance examination in lieu of the Bac.
Unlike England, the United States and, more recently, Sweden, in France the
results achieved by students on external high-stakes tests are not seen as a good way of
monitoring standards of achievement in schools (although newspapers do publish
‘league tables’ based on schools’ results in the baccalaure´at examinations). Instead,
the Ministry of Education, through its Office for Assessment and Forecasting,
monitors all aspects of educational provision, including facilities and resources,
classroom practices, students’ achievements and school effectiveness, through
focused surveys. For example, the achievement of students at the end of study in
colle`ges is evaluated by use of sample surveys. As well as measuring achievement on
each of the school subjects, these surveys also assess non-cognitive aspects, including
attitudes and values. In addition, there are cohort studies in which samples of
students are followed through several years of their schooling, so that long-term
trends can be monitored.
As well as these evaluative studies, the Ministry of Education also believed that it
was necessary to provide individual teachers with data on their own students:
This is based on the belief that such stakeholders, among whom teachers are prominent,
will improve their professional practices only if they are shown, as in a mirror, the
consequences of their actions. (Bonnet, 1997, p. 299)
Accordingly, a system was introduced of testing all students in alternate years at the
ages of 8 and 11, and every year in all subjects for students at age 16. The formative
purpose of these assessments was emphasized by having the tests set at the beginning of
the school year, so that they inform each teacher about their new class. Accordingly,
these tests are seen as an aid to teaching, rather than a judgement of teachers.
Although the primary purpose of these tests was to inform teachers about their
students, a sample of the data is analysed by the Ministry of Education in order to
inform pre-service and in-service training programmes. In 1989, for example, it was
discovered that the performance of primary school students in geometry appeared to
be weaker than that in arithmetic, even, though both aspects of mathematics were
meant to be given equal emphasis. The publicity given to this finding was enough to
prompt teachers to place more emphasis on geometry, which subsequently led to
improvement in geometry scores.
Finally, there is another layer of evaluation provided by the system of school
inspectors, which is mainly devoted to evaluating the performance of individual
teachers rather than to schools as such.
Lessons from around the world 253

What is perhaps most remarkable about the French system is the plethora of
different assessment and evaluation systems, each designed to serve a very limited
range of functions. The assessment of students’ attainment is achieved through
examinations such as the brevet and the Bac but the evaluation of the system is
achieved through other, purpose-built, assessments. The quality of the performance
of individual teachers is assessed through the inspection system but schools are
expected, in addition, to analyse their own performance against national and regional
norms, including the use of ‘value-added’ measures to examine the progress of
individual students as they move through the school. But the results on the
summative tests are not seen to provide evidence about the performance of the
schools, for this would ‘most certainly antagonize the teaching profession – it would
be seen as a way to evaluate the teachers themselves – and alter beyond recognition
the formative nature of mass assessment’ (Bonnet, 1997, p. 303).
While it would be going too far to say that the use of traditional, timed
examinations is regarded as unproblematic in France, there certainly seems to be a
strong and widespread belief that such assessments are fair and valid ways of assessing
students. More importantly, the fact that the teacher has no role in assessing the
student summatively leaves the teacher free to concentrate on learning, and it is
therefore probably not a coincidence that the role of assessment in the support of
learning has been such a strong aspect of French teaching for many years. While the
term ‘evaluation’ in French clearly includes what would be termed ‘formative
assessment’ and ‘summative assessment’, the lack of a role for teachers in the latter
appears to have resulted in the former being fully integrated into pedagogy. Indeed, as
the French research literature on this topic makes clear, the term ‘e´valuation formative’
includes all the ways teachers might elicit from students evidence of their
understanding, including discussions with and observations of students (Allal &
Lopez, 2005).
Germany
The responsibility for education policy in Germany rests with each of the 16 La¨ nder,
but through the Conference of Education Ministers (KMK) agreement has been
reached on a wide range of issues in education, including the curriculum, grading
systems used in school, and mutual acceptance of each other’s qualifications.
In grades 1 and 2, parents are given report cards at the end of each year which
detail their child’s work habits, special skills and weaknesses, behaviour, attitude to
learning, and their participation in class, but no grades are given. No student is
required to repeat these grades unless this is requested by the parents (for example, in
the case of a student who has an extended absence due to illness).
In the higher grades, the results of the formal tests and examinations are reported to
parents, who must sign to show that they have seen them. Any student getting one of
the two lowest grades (5 or 6) in two subjects for the year can be required to repeat
the year. Any student who is held back twice is evaluated to see whether they would
benefit from attending a special school. Officially, 4 per cent of students in a cohort
254 P. Black and D. Wiliam

are held back, with the cumulative effect being that by the age of 14 as many of 20 per
cent of students are being taught in classes where the majority of students are at least
a year younger (Beaton et al., 1996a, 1996b).
During the final year of elementary school (grade 4), on the basis of the student’s
performance, teachers recommend the appropriate form of lower secondary school
(in one Land there is a central examination). If parents are not happy with the
recommended allocation they can appeal, but the burden of proof rests on them to
show that the student has the necessary ability to prosper in the chosen school (for
example, through additional tests and examinations).
The teachers’ recommendations are based largely on the scores that students
achieve on a series of tests and examinations administered during the year, the
format, length and frequency of which are determined by the Ministry of
Education. For example, in one Land, the Ministry of Education specifies that
only German and mathematics may be assessed formally, on the following format,
with each subject being assessed according to the following schedule (Nerison-Low,
1999):
2nd grade: 4 tests, no more than 15 minutes each;
3rd grade: 6 examinations, 3 no more than 30 minutes and 3 no more than 15 minutes;
4th grade: 6 examinations, 4 no more than 30 minutes and 2 no more than 45 minutes.
Teachers typically collaborate to ensure that the tests and examinations that they set
make comparable demands on students, to agree how the tests will be scored, and
how the marks on the test will be converted to the six-point scale used throughout the
German education system.
However, the German educational system is currently in a great deal of flux.
German students displayed strong performances in the first and second international
mathematics and science studies (Brown, 1996), but in the third study, which
controlled more effectively for the effects of grade retention, German students were
shown to be considerably weaker than in the previous studies. Moreover, the
correlation of achievement with social class was among the strongest of all the
industrialized countries. As Thu¨ rmann (2004) notes, ‘The media interpreted these
results as nothing short of a national catastrophe’, and a variety of measures have been
put in place to improve the performance of the school system.
The general trend of the reforms is reminiscent of those introduced in England and
Wales in the 1988 Education Reform Act. Schools are to become more accountable
through regular monitoring of performance, but are also to be given greater
autonomy to find their own ways to improve.
The measures to increase accountability have included the adoption, in 2003, of
common curricular standards for the core subjects across all 16 La¨nder, and many
La¨nder have introduced centralized assessment systems to ensure that standards
are being applied consistently, and that schools are using data to monitor students’
learning. For example, in North Rhine–Westphalia recent reforms have intro-
duced:
Lessons from around the world 255

new curricular standards in German, mathematics and foreign languages;
school-based ‘parallel assessment’ for a selection of subjects in grades 3, 7, 10, 11/12;
state-wide mandatory assessment for core subjects in grade 4 and at the beginning of
grade 9;
centrally set examinations at the end of lower secondary school;
centrally set examinations at the end of upper secondary education;
independent school inspection. (Thu¨ rman, 2004)
At the same time, schools are expected to put in place systems of self-evaluation in
order to find their own ways to improve.
Given these radical changes, what is perhaps most surprising to an outside observer
is that, at the moment, faith in the core features of German education seems relatively
unshaken. The research on the effects of selection seem to indicate that the use of
selective systems increases the variance in outcomes, while depressing the mean
(Bursten, 1992), and yet there are few calls for an end to the tripartite system of
secondary schooling. Similarly, while the research on grade retention is controversial
(see Alexander et al., 1995), the use of grade retention adds substantially to the cost of
educational provision, and other ways of using these resources are likely to be more
effective. Finally, and most importantly, there appears to be substantial faith in the
usefulness of regular testing both to motivate students and to provide useful
information to teachers, despite the lack of evidence to support this (Ko¨ ller, 2005).
Nevertheless, it is important to note that education in Germany is undergoing its
most radical upheaval for at least fifty years. As Thu¨ rmann notes, ‘The acceleration of
change and the almost complete change of perspective from input-orientation to
output-control and evaluation is breath-taking’ (2004, p. 14). Even such hallowed
features as selection and grade retention may come under scrutiny as schools come
under greater and greater pressure to improve the achievements of their students,
although the faith in regular summative testing as the best way to achieve this would
appear to remain very strong.
United States
The most important feature of the education system in the United States is that there
isn’t one. It is much more productive to think of the United States as one might think
of Europe, so that there are 51 systems (50 states plus the District of Columbia), and
in many respects there are 17,000, since each school district has considerable
autonomy in determining school organization, curricula, teachers’ pay, and the day-
to-day operation of its schools. The largest school district, New York City Public
Schools, has over 1 million students – more than many countries – while the Lake
Alice School District in Nebraska has just 1 school, 6 teachers and 77 students.
Like France and Germany, the United States operates a grade-based curriculum,
but grade-retention, although increasing, is comparatively rare in grades K–8, so that
while the curriculum is designed as a grade-based system, it is not operated as such.
256 P. Black and D. Wiliam

This, combined with the widespread use of mixed-ability classes in these grades, may
well be why the same material tends to be repeated in the curricula of more different
grades in the US than in other countries (Schmidt et al., 1997) and, in mathematics at
least, more than half of lesson time is spent on reviewing previously taught material
(Hiebert et al., 2003).
Beginning in the third or fourth grade (and continuing through to postgraduate
level!), almost all formally assessed student work is assessed on the same literal grade
scale: A, B, C, D, F (fail), typically corresponding to percentage scores of 90–100,
80 – 89, 70 – 79, 60 – 69 and 0 – 60 respectively. Grades are cumulated by converting
them back to numbers (A = 4, B = 3, C = 2, D = 1, F = 0) and calculating the ‘grade-
point average’ over the year. However, unlike scores or grades given in most
European countries, the grade is usually not a pure measure of attainment, but will
include how much effort the student put into the assignment, attendance, and
sometimes even behaviour in class. Paul Dressel’s definition of a grade was ‘an
inadequate report of an inaccurate judgement by a biased and variable judge of the
extent to which a student has attained an undefined level of mastery of an unknown
proportion of an indefinite material’ (Chickering, 1983), and while this may be a bit
unfair, there can be little doubt that the meaning of a grade varies substantially from
school to school, and even from teacher to teacher.
Nevertheless, great importance is attached to grades, both as an indication of the
progress a student makes at school, and as ‘currency’ for applications to higher
education institutions, and teachers frequently come under pressure to modify their
grades if they are felt by parents to be too low. While only a few districts have gone as
far as establishing ‘grade courts’ where parents can take grievances about grades, and
where teachers have to defend the grades they have awarded, teachers do feel under
considerable pressure to be able to defend the grades they have awarded and grades
are often just based on the outcomes of objective tests.
The other feature of education in the United States that militates against the
introduction of effective formative assessment is the response to increases in testing
for accountability purposes. Most of the money needed to operate schools is provided
by taxes on residential and commercial property within the school district, and most
districts are governed by boards elected by local taxpayers. Over the last forty years or
so, however, state and federal sources have become greater and greater net
contributors (Corbett & Wilson, 1991, p. 25), leading to demands that school
districts become accountable beyond the local community. As noted above, all states
but Iowa have state-wide educational standards, and most have also implemented
some form of state-wide testing programme.
Although some states implemented systems of assessment that threatened
sanctions for individual students for poor performance (e.g. requiring that students
pass minimum-competency tests before they can be awarded high school diplomas),
most state systems were, and remain, low stakes for students but high stakes for the
schools. This principle is continued in the No Child Left Behind Act (NCLB). Under
NCLB, each state must propose a series of staged targets for achieving the overall goal
of all students in grades 3–8 being proficient in reading and mathematics by 2014.
Lessons from around the world 257

Each school is judged to be making ‘adequate yearly progress’ (AYP) towards this
goal if the proportion of students being judged as ‘proficient’ on annual state-
produced tests exceeds the target percentage for the state for that year. Furthermore,
the AYP requirements apply not only to the totality of students in a grade but also to
specific subgroups of students (e.g. ethnic minority groups), so that it is not possible
for good performance by some student subgroups to offset poor performance in
others.
Failure to make AYP has severe consequences for schools and, as a result, many
schools and districts have invested both time and money in setting up systems for
monitoring what the teachers are teaching and what students are learning. In order
to ensure that teachers cover the curriculum, most districts have devised
‘curriculum pacing guides’ that specify which pages of the set texts are to be
covered every week (and sometimes each day). With such rigid pacing, there are
few opportunities for teachers to use information on student performance to
address learning needs.
More recently, there has been a huge upsurge of interest in systems that monitor
student progress through the use of regular formal tests that are designed to predict
performance on the annual state tests. The idea of such regular testing is that students
who are likely to fail the state test, and may therefore prevent the school from reaching
its AYP target, can be identified and given additional support, and for this reason,
these systems are routinely described in the USA as ‘formative assessment’, even
though the results of the assessments rarely impact on learning and, as such, might be
better described as ‘early-warning summative’. The result is that while teachers in the
United States would appear to have considerable freedom to devise assessment
systems that integrate summative and formative uses, the obstacles to doing so are
substantial, and well entrenched.
Conclusion
Assessment methods provide tools that can be used in a variety of ways. The choice
and deployment of these tools, and the interpretation and use of their results, are
subject to a range of educational, public and political influences. The variety and
complexity of these influences may be seen by listing some of them as follows:
.beliefs about what constitutes learning;
.beliefs in the reliability and validity of the results of various tools;
.trust in the objectivity of formal testing;
.a preference for and trust in numerical data, with bias towards a single number;
.trust in the judgements and integrity of one’s children’s teachers;
.trust in the judgements and integrity of the teaching profession as a whole;
.belief in the value of competition between students;
.belief in the value of competition between schools – the market model of
education;
.belief that test results are a meaningful indicator of school effectiveness;
258 P. Black and D. Wiliam

.fear of national economic decline and belief that education is crucial to
improvement;
.belief that the key to schools’ effectiveness is strong top-down management.
All of the above are arenas of contention, and each may reflect beliefs that are neither
based on evidence nor susceptible to change by arguments from evidence. The
various elements interplay in many and complex patterns that are embedded in a
national culture as a whole. Safe generalizations are hard to come by, and any deep
understanding may come only from case-studies of individual countries.
Thus, in the case of France, the baccalaure´at has deep historical roots in the
Napoleonic code (Broadfoot, 1996), and thereby enjoys strong public confidence
with accompanying resistance to radical change. The avenues to change have been
evolutionary, and both regionalization and diversification have changed its character
in response to perceived economic and employment needs. Its shortcomings in terms
of reliability and validity are not explored or even questioned. However, the belief that
improvement in learning depends on the commitment of, and respect for, the
teaching profession has informed national policies, which combine sampling surveys
to inform judgements with provision of tools to help teachers to improve their
assessments.
Germany resembles France in having a tradition, albeit more recent, of a national
test, the Abitur, the status and influence of which are under strain because of the
pressures of mass education. The trust given to teachers varies between different
regions. At the same time, the practices of tracking, of repeating the year, and of
strong differentiation between different types of secondary schools, reflect a
distinctive set of educational and social beliefs, and put unique pressures on the
summative judgements of teachers.
England is different again. Here, the revolution put into effect in the 1990s
reflected deep distrust of teachers combined with fear that education was a cause of
economic decline. The trust in formal tests as an engine of change has led to pupils in
England stealing from the USA the dubious distinction of being the most frequently
tested in the world. However, the test results have shown the typical pattern of initial
improvement followed by saturation, so frustrating politicians’ promises of continued
improvement. Signs of change in national policy are slowly emerging, with trials to
replace some national testing by teachers’ assessments, and initiatives to promote
formative assessment.
In the United States, teachers have been trusted by their local communities (as the
annual Phi Delta Kappan surveys show) but not by policy-makers at the state and
federal level. The result is an uneasy compromise, with districts free to decide what
their students learn, and how student performance is to be assessed while, at the same
time, holding schools accountable for students’ performance on material that they
may not have been taught.
Social trends can also lead to new demands on any testing system. Emphasis on
the need for, and advantages of, mass higher education may reflect, and help
create, demands that cannot be satisfied – so selection for places has to be made
Lessons from around the world 259

more difficult. In France, those seeking entry to the most prestigious university
institutions, the grandes e´coles, have to spend two further years after the baccalaure´at
to prepare for further selection tests – the majority fail and then enrol in
institutions that require no more than a baccalaure´at. In Germany, the Abitur used
to guarantee automatic entry to university courses, but pressure on places in the
popular courses has led universities to introduce their own supplementary tests. In
England, the rise in both numbers attempting, and the proportions succeeding, in
the examinations which are used for university selection (the A-level) are leading
to pressure for a finer grain of reporting for the top grades, while some prestigious
universities also plan to use their own ‘aptitude’ tests to supplement the public test
results.
All of this illustrates several lessons. One is that in each country assessment
practices have impacts on teaching and learning that may be strongly amplified or
attenuated by the national context. Indeed, the overall impact of particular
assessment practices and initiatives is determined at least as much by culture and
politics as it is by educational evidence and values. A further lesson is that it is likely to
be idle to draw up maps for the ideal assessment policy for a country, even if the
principles and the evidence to support such an ideal could be clearly agreed within the
‘expert’ community. The way forward might, rather, lie in those arguments and
initiatives that are least offensive to existing assumptions and beliefs, and which will
nevertheless serve to catalyse a shift in them while at the same time improving some
aspects of present practice.
As far as the integration of formative and summative functions of assessment is
concerned, the analysis of the four national systems described above is something of a
paradox: the better the teacher knows her or his students, through processes of
formative assessment, the less likely it is that the information is used to inform
judgements made about the student. In France, albeit under the banner of pedagogy
rather than assessment, effective classroom assessment practices have been developed
on a widespread basis, but the key decisions about a student’s educational trajectory
are made on the basis of examination results. In Germany, teachers are involved in
making key decisions about student progress but the primary evidence base for this is
provided by formal tests. Similarly, in England, teacher judgements do feed into
national assessments, but the demands of teachers to be accountable for these
judgements has resulted in record-keeping systems that do not serve learning well.
Finally, in the United States, multiple demands for accountability at different levels of
the system have resulted in multiple assessment systems, all geared to serving the
summative function of assessment, so marginalizing, and denying time to, assessment
that supports learning.
Thus not only is there no ‘royal road’ to an assessment system that effectively serves
both formative and summative functions that each country could follow, but it seems
likely that the idiosyncratic road that will need to be taken in each country will also be
very hard going. The final irony is that it is precisely the demand for accountability
which has produced unprecedented pressure to improve education systems that is
likely to be the biggest impediment to achieving that improvement.
260 P. Black and D. Wiliam

Note
Paul Black is Professor Emeritus of Science Education at King’s College London. Dylan Wiliam is
Director of the Learning and Teaching Research Center at ETS, Princeton, New Jersey, but writes
here in a personal capacity.
References
Allal, L. & Lopez, L. M. (2005) Formative assessment of learning: a review of publications in
French, in: J. Looney (Ed.) Improving learning through formative assessment: cases, policies,
research (Paris, France, Organisation for Economic Co-operation and Development).
Alexander, K. L., Entwistle, D. R. & Dauber, S. L. (1995) On the success of failure: a reassessment of
grade retention in the primary grades (Cambridge, Cambridge University Press).
Beaton, A. E., Mullis, I. V. S., Martin, M. O., Gonzalez, E. J., Kelly, D. L. & Smith, T. A. (1996a)
Mathematics achievement in the middle school years: IEA’s third international mathematics and
science study (Chestnut Hill, MA, Boston College).
Beaton, A. E., Martin, M. O., Mullis, I. V. S., Gonzalez, E. J., Smith, T. A. & Kelly, D. L. (1996b)
Science achievement in the middle school years: IEA’s third international mathematics and science
study (Chestnut Hill, MA, Boston College).
Black, P. J. & Wiliam, D. (2004) The formative purpose: assessment must first promote learning,
in: M. Wilson (Ed.) Towards coherence between classroom assessment and accountability: 103rd
Yearbook of the National Society for the Study of Education (part 2) (Chicago, IL, University of
Chicago Press), 20–50.
Bonnet, G. (1997) Country profile from France, Assessment in Education, 4(2), 295–306.
Broadfoot, P. M. (1996) Education, assessment and society: a sociological analysis (Buckingham, Open
University Press).
Brown, M. L. (1996) FIMS and SIMS: the first two IEA international mathematics surveys,
Assessment in Education: Principles, Policy and Practice, 3(2), 193–212.
Bursten, L. (Ed.) (1992) The IEA study of mathematics III: student growth and classroom processes
(Oxford, Pergamon).
Chickering, A. W. (1983) Grades: one more tilt at the windmill, American Association for Higher
Education Bulletin, 35(8), 10–13.
Corbett, H. D. & Wilson, B. L. (1991) Testing, reform and rebellion (Hillsdale, NJ, Ablex).
Daugherty, R. (1995) National curriculum assessment: a review of policy 1987–1994 (London, Falmer).
Hiebert, J., Gallimore, R., Garnier, H., Bogard Givvin, K., Hollingsworth, H., Jacobs, J., Chui,
Miu-Ying A., Wearne, D., Smith, M., Kersting, N., Manaster, A., Tseng, E., Etterbeek, W.,
Manaster, C., Gonzales, P. & Stigler, J. (2003) Teaching mathematics in seven countries: results
from the TIMSS 1999 video study (Washington, DC, National Center for Education Statistics).
Ko¨ ller, O. (2005) Formative assessment in classrooms: a review of the empirical German literature,
in: J. Looney (Ed.) Improving learning through formative assessment: cases, policies, research (Paris,
Organisation for Economic Co-operation and Development).
Nerison-Low, R. (1999) Individual differences and the German education system, in: M. A.
Ashwill (Ed.) The educational system in Germany: case study findings, June 1999 (Washington,
DC, National Institute on Student Achievement, Curriculum, and Assessment, Office of
Educational Research and Improvement, US Department of Education, National Center for
Education Statistics).
Schmidt, W. H., McKnight, C. C. & Raizen, S. A. (1997) A splintered vision: an investigation of US
science and mathematics education (Dordrecht, Netherlands, Kluwer Academic Publishers).
Thu¨ rmann, E. (2004) New curricular formats and classroom development: a change of paradigm
for the German school system?, in: J. Letschert (Ed.) The integrated person: how curriculum
development relates to new competencies (Enschede, Netherlands, CIDREE).
Lessons from around the world 261