Knowledge

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

EXECUTIVE
SUMMARY and Skills for Life
FIRST RESULTS
FROM PISA 2000

EDUCATION AND SKILLS

Programme for International Student Assessment

What is PISA? (Programme for International Student Assessment)

A new three-yearly survey of the A new way of looking at A unique collaboration between
knowledge and skills of 15-year- student performance. countries to monitor educational
olds in the principal industrialised outcomes.
countries.

The survey, conducted first in 2000, PISA assessed young people’s PISA was co-ordinated by
will be repeated every three years. capacity to use their knowledge governments of participating
and skills in order to meet real- countries, through the
265,000 students from 32 countries life challenges, rather than Organisation for Economic
took part. merely looking at how well they Co-operation and Development
had mastered a specific school (OECD).
Students sat pencil and paper curriculum.
assessments in their schools. Leading international experts
PISA assessed literacy in reading, worked to develop an assessment
Students and their principals also mathematics and science. whose results are comparable
answered questionnaires about across different national and
themselves and their schools. This Students had to understand key cultural contexts.
allows PISA to identify what factors concepts, to master certain
are associated with better and worse processes and to apply knowledge PISA improves international
performance. and skills in different situations. information on student outcomes,
giving countries benchmarks and
Information was also collected regular updates on how students
on student attitudes and perform against them.
approaches to learning.

The countries taking part

In 2000, 28 OECD Member countries and four other countries carried out the first PISA survey. A further 13 countries will
conduct the same survey in 2002, and the two OECD countries that did not take part in 2000 will participate in the second
survey in 2003.

Countries reported on here (assessed in 2000):

OECD members: Australia, Austria, Belgium, Canada,
Colour key:

Czech Republic, Denmark, Finland,

France, Germany, Greece, Hungary,
Iceland, Ireland, Italy, Japan, Korea,
Luxembourg, Mexico, Netherlands,
New Zealand, Norway, Poland,
Portugal, Spain, Sweden, Switzerland,
United Kingdom, United States.
Non-OECD members: Brazil, Latvia, Liechtenstein,
Russian Federation.

Countries conducting the same survey in 2002:

Albania, Argentina, Bulgaria, Chile, China, Hong Kong -
Special Administrative Region of China, Indonesia, Israel,
Lithuania, Macedonia, Peru, Romania, Thailand.

OECD countries joining the survey in 2003:

Slovak Republic, Turkey.

2
Results and findings - What PISA 2000 tells us…

This brochure summarises the results of The results for reading literacy are …about the relationship between
PISA 2000 in terms of student summarised on pages 4-9. student performance and family
performance, and wider findings about background…
what lies behind these results. The full The results for mathematical literacy
report, Knowledge and skills for life – are summarised on pages 10-11. Students with more advantaged
First results from PISA 2000, can be family backgrounds tend to perform
obtained from the OECD (available in The results for scientific literacy are better educationally. PISA allows the
English, French and German) and summarised on pages 12-13. strength of this link to be looked at
further information can be found at more closely, showing how it differs
www.pisa.oecd.org. See also back cover. …about what students are like as between school programmes and
learners… between countries.
…about student literacies…
Today’s young people will need to go The findings on family background
PISA 2000 assessed students’ capacities on learning in different ways are summarised on pages 18-19.
to apply knowledge and skills in throughout their lives. PISA looked at
reading, mathematics and science. some aspects of what they are like as …about the relationship between
These capacities are referred to as learners at age 15. It considered their school differences and family
reading, mathematical and scientific motivation and engagement in learning, background…
literacy and widely seen as essential and aspects of their learning strategies.
prerequisites for students to be well Can schools help to moderate the
prepared for adult life. The findings on what students are like effect of family background? The
as learners are summarised on pages answer depends partly on the degree to
PISA does not measure literacy as an 14-15. which students with different
“all or nothing” set of knowledge and characteristics attend different schools,
skills. Rather, each student receives a …about gender differences in and on how much the results of these
score on a continuous scale. The score performance and engagement… schools differ.
obtained by a student indicates the
most difficult type of task that the PISA’s results show differences in The findings on school differences
student is likely to perform correctly. performance between males and are summarised on pages 20-21.
Each PISA task is associated with a females. They also show differences in
scale score. their engagement in school and their …and about the characteristics of
learning strategies. schools where students do well
The PISA scale for each literacy area
was devised so that across OECD The findings on gender differences are Schools differ in terms of resources,
countries, the average score is 500 summarised on pages 16-17. policies and classroom practices. PISA
points, and around two-thirds of 2000 found that students do better, on
students achieve between 400 and 600 average across countries, in schools
points. with certain characteristics.

For reading, five levels of literacy are The findings on school effects
described, Level 5 being the highest. are summarised on pages 22-23.

3
How students perform in reading literacy

Students taking part in PISA were country. Here, countries are ranked by
asked questions based on a variety of the percentage of students who are
written texts, ranging from a short story proficient at Level 3 or above. About
to a letter on the Internet and 60% of 15-year-olds in the combined
information presented in a diagram. OECD area can complete such tasks
They were assessed on their capacity to (see Page 7 for examples). But this
retrieve specified information, on fraction varies from above three-
whether they could interpret what they quarters in Finland and Korea, to
read, and on how well they could reflect below half in Brazil, Latvia,
on and evaluate it, drawing on their Luxembourg, Mexico, Portugal and the
existing knowledge. For each of these Russian Federation.
three aspects of reading literacy,
students were given a score based on the No single indicator can adequately
difficulty of the tasks that they could summarise the performance of students
perform. A combined score shows their in different countries. The figures on
overall reading performance. the following pages show different parts
of the distribution of student
On the basis of these scores, each performance.
student was assigned to one of five
reading levels (see legend on facing In general, within countries similar
page). Examples of tasks at the different percentages of students are proficient
levels are given in the following pages. at each level in the different aspects of
reading literacy – retrieving,
The figure (below) shows the interpreting, and reflection and
percentage of students who are evaluation. However, in some countries
proficient at each level in the significant differences exist between
combined OECD area and in each student performance in the more
“routine” reading tasks of identifying
information and interpreting it, and
tasks requiring reflection and
evaluation (see Tables 2.1b, c and d in
Percentage of students by highest the full report).
level of reading proficiency
1. Combined OECD area

100

80

60
2. Indivdual countries

40

20

0
e
ag
er
av
D
EC
O

Countries are ranked by percentage of students at least at Level 3 * Non - OECD country

4
One in ten students completed PISA’s hardest reading tasks....

For example
727 Students were shown a tree diagram of a country’s working-age population, and
descriptions of the labour force status of individual workers. They had to decide in
which category of the diagram each worker belonged. They had to work out what
criteria to use to classify workers from the structure and content of the diagram,
655
drawing on information in footnotes and therefore not prominent. This task is
associated with a score of 727 points on interpreting scale.

Students were shown a notice from a personnel department about a service that
would help with job mobility. They had to work out the two ways in which this
652
service could help people who lost their jobs – information that was stated indirectly
and had to be distinguished from competing information that could easily be mistaken
for the information required. Associated score: 655 points on the retrieving
information scale.

After reading a three-page story about a woman’s adventure, students were asked to
say whether they thought it had an appropriate ending, explaining why. To obtain
full credit, they had to evaluate the ending in terms of its thematic completeness, by
relating the last sentence to central relationships, issues or metaphors in the story.
Associated score: 652 points on reflection and evaluation scale.

For the full sample items see www.pisa.oecd.org.

Students proficient at Level 5 on the hypotheses, drawing on specialised

combined reading literacy scale are knowledge, and accommodating
capable of completing sophisticated concepts that may be contrary to
reading tasks, such as: expectations.

managing information that is Only 10% of 15-year-olds in the

difficult to find in unfamiliar combined OECD area are proficient
texts; at Level 5 in OECD countries. The
percentage ranges from over 15 per
showing detailed understanding of cent in Australia, Canada, Finland,
such texts and inferring which New Zealand and the United Kingdom,
information in the text is relevant to below 5% in Brazil, Latvia,
to the task; Luxembourg, Mexico, Portugal, Spain
and the Russian Federation (see
evaluating critically and building Table 2.1a in the full report).

Percentage of students at Level 5

100
19 18 18 17 16 14 12 12 11 11 10 9 9 9 9 9 8 8 7 6 6 5 5 5 5 4 4 4 3 2 1 1
Ranked by percentage of students

80

60

40
at Level 5

20

0
e
ag
er
av
D
EC
O

* Non - OECD country

5
...One student in six could at most complete the simplest tasks...

For example
Students were shown a magazine article, written for young people, explaining the
process and purpose of DNA testing. A multiple-choice task required them simply to
recognise that the writer’s main purpose was to inform, rather than to warn, amuse
or convince. This task is associated with a score of 406 points on interpreting scale.
406
Students were asked about a sentence in an article about sports shoes, discussing
aspects of the shoe that can avoid damage to feet. A multiple-choice task required
402 students to recognise the relationship between the two parts of the sentence: that the
408 to 480 points second part provided the solution to the problem stated in the first part (rather than
illustrating, repeating or contradicting it). Associated score: 402 points on the
reflection and evaluation scale

below 335 points 367

After reading a short adventure story, students were asked in a multiple-choice task
what happened next after a brief quoted extract. The answer was stated explicitly in
the narrative and was easy to locate from the information given in the task.
Associated score: 367 points on the retrieving information scale.

For the full sample items see www.pisa.oecd.org.

Reading literacy, as defined in PISA, information, identifying the main 1 tasks. However, this percentage
focuses on the knowledge and skills theme of a text or making a simple varies considerably between countries
required to apply “reading for learning” connection with everyday knowledge. and it is noteworthy that every
rather than on the technical skills country has some students who could
acquired in “learning to read”. In line On average across OECD countries, perform only at this level. Parents,
with most contemporary views about 12% of 15-year-olds are proficient at educators and policy-makers in
reading, PISA focuses on measuring this level but no higher. A further 6% countries where this proportion is high
the extent to which individuals are could not perform these simple tasks. need to recognise that significant
able to construct, expand and reflect This does not mean that these students numbers of students are not benefiting
on the meaning of what they have read cannot read in a technical sense. sufficiently from educational
in a wide range of texts common both Nonetheless, such students have opportunities and may not be
within and outside school. The serious difficulties in using reading acquiring the knowledge and skills to
simplest reading tasks that can still be literacy as an effective tool to advance do so in their further school careers
associated with this notion of reading and extend their knowledge and skills and beyond. In only four countries,
literacy are those at Level 1. Students in other areas (see Table 2.1a in the Canada, Finland, Japan and Korea, do
proficient at this level are capable of full report). 10% or less of 15-year-olds read at
completing only the least complex Level 1 or below. In three OECD
reading tasks developed for PISA, such In total, just over one student in six countries, Luxembourg, Mexico and
as locating a single piece of (18%) does not progress beyond Level Portugal, more than 25% do so.

Percentage of students at Level 1 or below

100
Ranked by percentage of students

80

60
at Level 1 or below

40

20

0
e
ag
er
av
D
EC
O

* Non - OECD country

6
...most students have neither very high nor very low reading skills

For example
Students had to look at two letters posted on the Internet giving conflicting opinions
581 about graffiti. They had to comment on which was written better, analysing writing
style and structure of argument rather than just what the letters said. To do this,
they had to draw on their understanding of what constitutes good writing. This task
is associated with a score of 581 points on the reflection and evaluation scale.

508 Students had to use information in two diagrams, one about historic water levels in
Lake Chad, the other about when various animals appear in cave paintings nearby,
in order to recognise that certain animals disappeared after a period of falling water
levels. Students had to combine information presented in two different ways.
423 Associated score: 508 points on the interpreting scale.

After reading a short extract from a play by Jean Anouilh, students had to work out
what the play is about: one character is playing a trick on another. A multiple-
choice task asked about the purpose of the trick. This required a low level of
inference to work out the main idea in the text. Associated score: 423 points on the
interpreting scale.

For the full sample items see www.pisa.oecd.org.

On average across OECD countries, moderate complexity, such as locating these middle levels of proficiency.
nearly three-quarters of 15-year-olds are multiple pieces of information, Leaving aside those countries with more
at Levels 2, 3 or 4. But in some countries making links between different parts than a quarter of students performing
the proportion falls to just over half and of a text, and relating it to familiar below Level 2 (Brazil, Latvia,
in others it rises to above eight in ten everyday knowledge. Luxembourg, Mexico, Portugal and the
students (see Table 2.1a in the full Russian Federation), this gives some
report). Students proficient at Level 2 are indication of how equally performance
capable of solving basic reading tasks, in each country is distributed. In Korea,
Students proficient at Level 4 are such as locating straightforward as an illustration, 89% of 15-year-olds
capable of solving complex reading information, making low-level are at the middle levels, with few very
tasks, such as locating embedded inferences of various types, working high or very low performers. By contrast,
information, construing meaning out what a well-defined part of a text in New Zealand only two-thirds are in
from nuances of language and means, and using some outside the middle, with the remaining third
critically evaluating a text. knowledge to understand it. split between students at Level 5 (three
times as many as in Korea) and those at
Students proficient at Level 3 are The figure below shows countries ranked Level 1 or below (over twice as many as
capable of solving reading tasks of by the percentage of students who are at in Korea).

Percentage of students at Levels 2, 3 and 4

100
Ranked by percentage of students

80

60
at Levels 2, 3 and 4

40

20

0
e
ag
er
av
D
EC
O

* Non - OECD country

7
Mean reading literacy scores

The figure (below) shows the mean It is evident that because the mean Mean Reading Literacy

score on PISA’s reading literacy scale scores of many countries are quite Range of rank order positions
in each country. Since PISA is a similar, the precise ranking of countries for each country based on
sample (with 95% confidence)
sample survey, the resulting estimates cannot always be determined: there is
are associated with some level of overlap in the bars showing the range Country Rank
uncertainty. The bar around the mean in which each country’s mean score Highest Lowest
possible possible
shows the range within which the can be said with confidence to fall.
Finland 1 1
mean lies with 95% confidence. For The table (right) shows the range of Canada 2 4
details on the comparison of mean rank order positions within which each New Zealand 2 8
Australia 2 9
scores between countries, see Figure 2.4 country’s mean lies, with 95% Ireland 3 9
in the full report. probability. Korea 4 9
United Kingdom 5 9
Japan 3 10
Sweden 9 11
Austria 11 16
Belgium 11 16
Iceland 11 15
Norway 11 16
France 11 16
United States 10 20
Denmark 16 19
Switzerland 16 21
Spain 17 21
Czech Rep. 17 21
Italy 19 24
Germany 21 25
Liechtenstein 20 26
Hungary 21 26
Poland 21 27
Greece 23 28
Portugal 24 28
Russian Fed. 27 29
Latvia 27 29
Luxembourg 30 30
Mexico 31 31
Brazil 32 32

Mean reading literacy scores: country similarities and differences

Points Points
550 550

530 530

510 510
OECD average = 500
490 490

470 470

450 450
The estimate of each country mean is shown by the middle line
430 The range in which the mean can be said to fall, with 95% 430
certainty, is shown by the bar
410 410

390 390
e
ag
er
av
D
EC
O

* Non - OECD country

8
One way to summarise the by considering the range of scores
performance of each country is to achieved by the middle half of the
compute the mean of student scores. population: the gap between the 25th
The figure on the previous page shows and the 75th student in a group of 100,
the mean reading score for each ranked by performance. The greater the
country. gap, the more unequal are the results
within a country.
Some significant differences appear in
the mean performance of students In all countries, the range of
across countries, with 125 score points performances among the middle 50% of
separating the lowest from the highest students exceeds the magnitude of one
performing OECD country. To the proficiency level (73 score points) and
extent that such differences are in Belgium, Germany and New Zealand
predictive of students’ career paths, twice this difference (OECD average
this raises questions about the future 1.8 proficiency levels) (see Table 2.3a
competitiveness of countries with large in the full report).
numbers of low performers (Table 2.3
in the full report). PISA shows that high mean
performance and low disparities can go
The difference in the mean together. The three countries with the
performance between some countries is smallest range in the middle half of the
large, but the variation in student population, Finland, Japan and Korea,
performance within countries is are also among countries with the
generally much larger. This can be seen highest mean performance in reading
literacy. By contrast, one of the three
countries with the highest internal
variation in performance, Germany,
scores below the OECD average.

This shows that wide disparities are

neither inevitable nor a precondition
for countries to attain a high overall
performance. On the contrary, the
results suggest that public policy can
promote equal opportunities and
equitable learning outcomes for all
students.

9
How students perform in mathematical literacy

The figure (below) shows the mean It is evident that because the mean Mean Mathematical Literacy

score for each country on PISA’s scores of many countries are quite Range of rank order positions
mathematical literacy scale. Since similar, the precise ranking of countries for each country based on
sample (with 95% confidence)
PISA is a sample survey, the resulting cannot always be determined: there is
estimates are associated with some overlap in the bars showing the range Country Rank
Highest Lowest
level of uncertainty. The bar around in which each country’s mean score possible possible
the mean shows the range within can be said with confidence to fall. Japan 1 3
which the mean lies with 95% The table (right) shows the range of Korea 2 3
New Zealand 4 8
confidence. For details on the rank order positions within which Finland 4 7
comparison of mean scores between each country’s mean lies, Australia 4 9
Canada 5 8
countries see Figure 3.2 in the full with 95% probability. Switzerland 4 10
report. United Kingdom 6 10
Belgium 9 15
France 10 15
Austria 10 16
Denmark 10 16
Iceland 11 16
Liechtenstein 9 18
Sweden 13 17
Ireland 16 19
Norway 17 20
Czech Rep. 17 20
United States 16 23
Germany 20 22
Hungary 20 23
Russian Fed. 21 25
Spain 23 25
Poland 23 26
Latvia 25 28
Italy 26 28
Portugal 26 29
Greece 27 30
Luxembourg 29 30
Mexico 31 31
Brazil 32 32

Mean mathematical literacy scores: country similarities and differences

Points Points
550 550

530 530

510 510
OECD average = 500
490 490

470 470

450 450

430 430

410 The estimate of each country mean is shown by the middle line 410

390 The range in which the mean can be said to fall, with 95% 390
certainty, is shown by the bar
370 370

350 350

330 330
e
ag
er
av
D
EC
O

* Non - OECD country

10
Literacy in mathematics and science is significance from that in Korea and
important for understanding medical, New Zealand. Other countries that also
economic, environmental and other score above the OECD average are
issues that shape modern societies, Australia, Austria, Belgium, Canada,
which rely heavily on technological Denmark, Finland, France, Iceland,
and scientific advances. PISA 2000 Liechtenstein, Sweden, Switzerland
offers a snapshot of student and the United Kingdom (see
performance in mathematical and Figure 3.2 in the full report).
scientific literacy but assessed them in
less detail than reading literacy. Each Examples:
was assessed on a single scale, without
distinct literacy levels, with an average A difficult PISA mathematical
score of 500 points as with reading. literacy task

PISA looked at mathematics in Students were presented with a diagram

relation to its wider uses in people’s showing the pattern in which different trees
lives. Mathematical literacy in PISA is would have to be planted in an orchard in
measured in terms of students’ capacity order that conifers provide sufficient
to: protection to apple trees. They had to work
out which type of tree would increase faster
recognise and interpret in number as the orchard was enlarged –
mathematical problems encountered in and explain why. This required them to
everyday life; notice that the number of apple trees
increased in proportion to the square of the
translate these problems into a number of conifers. The task required
mathematical context; students to think mathematically and
recognise a general principle. This task is
use mathematical knowledge and associated with a score of 723 points.
procedures to solve problems;
A medium PISA mathematical
interpret the results in terms of the literacy task
original problem;
From a mathematical representation of the
reflect on the methods applied; and dimensions and shape of a pyramidal barn
roof, students had to calculate the area of
formulate and communicate the its base. This required students to identify
outcomes. and perform a straightforward calculation,
understanding the overall concepts of space
PISA mathematical tasks varied in and shape. Associated score: 492 points.
difficulty according to several criteria,
including: An easy PISA mathematical
literacy task
the number and complexity of
computational steps involved; From a graph showing the speed of a
racing car as it travelled round a track,
the need to connect and integrate students had to answer a multiple-choice
material; and task about where on the track the car went
the slowest. This required only a simple
the need to represent and interpret observation and sufficient understanding
material and reflect on situations of the concept of change to realise that the
and methods. slowest speed would be shown at the
lowest point of the plot on the speed graph.
15-year-olds in Japan display the Associated score: 403 points.
highest mean scores in mathematical
literacy, but Japan’s mean performance For the full sample items see
cannot be distinguished with statistical www.pisa.oecd.org.
11
How students perform in scientific literacy

The figure (below) shows the mean It is evident that because the mean Mean Scientific Literacy

score for each country on PISA’s scores of many countries are quite Range of rank order positions
scientific literacy scale. Since PISA is a similar, the precise ranking of countries for each country based on
sample (with 95% confidence)
sample survey, the resulting estimates cannot always be determined: there is
are associated with some level of overlap in the bars showing the range Country Rank
Highest Lowest
uncertainty. The bar around the mean in which each country’s mean score possible possible
shows the range within which the can be said with confidence to fall. Korea 1 2
mean lies with 95% confidence. For The table shows the range of rank Japan 1 2
Finland 3 4
details on the comparison of mean order positions within which United Kingdom 3 7
scores between countries, see Figure 3.5 each country’s mean lies, Canada 4 8
New Zealand 4 8
in the full report. with 95% probability. Australia 4 8
Austria 8 10
Ireland 9 12
Sweden 9 13
Czech Republic 10 13
France 13 18
Norway 13 18
United States 11 21
Hungary 13 21
Iceland 14 20
Belgium 13 21
Switzerland 13 21
Spain 16 22
Germany 19 23
Poland 19 25
Denmark 21 25
Italy 22 25
Liechtenstein 20 26
Greece 25 29
Russian Fed. 26 29
Latvia 25 29
Portugal 26 29
Luxembourg 30 30
Mexico 31 31
Brazil 32 32

Mean scientific literacy scores: country similarities and differences

Points Points
550 550

530 530

510 510
OECD average = 500
490 490

470 470

450 450

430 430

410 The estimate of each country mean is shown by the middle line 410

390 The range in which the mean can be said to fall, with 95% 390
certainty, is shown by the bar
370 370

350 350

330 330
e
ag
er
av
D
EC
O

* Non - OECD country

12
Scientific literacy was scored on a scale Examples
measuring students’ capacity to:
A difficult scientific literacy task
use scientific knowledge;
Students were shown extracts from a 19th
recognise scientific questions; century scientist’s diary, a table with his
observations and a commentary,
identify what is involved in discussing the post-natal death from a
scientific investigations; particular fever of a large proportion of
mothers in two wards of a hospital
relate scientific data to claims and maternity clinic. Students had to indicate
conclusions; and to why the evidence did not support a
contemporary belief that earthquakes
communicate these aspects of caused the fever. This required them to
science. explain the significance of different death
rates in the two wards. This task is
PISA’s scientific tasks varied in associated with a score of 666 points.
difficulty according to several criteria
including: A medium scientific literacy task

the complexity of the concepts used; After reading a text on the risks to the
ozone layer and their implications,
the amount of data provided; students were asked whether each of two
questions could be answered by scientific
the chain of reasoning required; and research. To answer correctly, they
needed, in particular, to recognise the
the precision required in difference between a question that requires
communication. a political choice and one that can be
answered scientifically. Associated score:
Japan and Korea show the highest per- 529 points.
formance on the scientific literacy
scale. Other countries that score statis- An easier scientific literacy task
tically significantly above the OECD
average are Australia, Austria, Canada, Students were asked why washing hospital
the Czech Republic, Finland, Ireland, sheets in high temperatures helps reduce
New Zealand, Sweden and the United the risk that patients will contract a fever.
Kingdom (see Figure 3.5 in the full In their answer they needed to apply their
report). scientific knowledge to this real-world
problem by referring, for example, to the
killing of bacteria. Associated score: 467
points.

For the full sample items see

www.pisa.oecd.org.

13
What are 15-year-olds like as learners?

Students need to leave school not just worth pursuing because they consider it
with sound subject-matter knowledge relevant to their future (see Table 4.2
and skills, but also ready to continue in the full report). Not only does lack
learning throughout life. In order to do of interest tend to be associated with
so, they must be able to manage their poorer student performance (see
own learning, rather than solely relying Tables 4.1 and 4.2 in the full report),
on teachers. This means that students but other research shows that students
must, on the one hand, be interested who are disaffected with learning at
and engaged in learning and, on the school will also be less likely to engage
other hand, have strategies for learning in learning activities, either inside or
effectively. outside school, in later life.

PISA looked at both – motivation and Students who are engaged in

engagement as well as learning reading beyond what is required for
strategies, as reported by students. The school tend to be better readers, but In assimilating new knowledge,
results can be related to students’ the relationship is not straightforward. students use both memorisation and
performance in reading, mathematical In some countries the association is elaboration strategies. (The latter
and scientific literacy in order to provide stronger than in others, and some refers to strategies to process new
some idea of what kinds of learners show countries with fewer keen readers still information, integrating it into a
higher performance at age 15. perform well, on average (see Table 4.3 learner’s prior knowledge base.)
in the full report). The results of PISA 2000 show that
Key Findings students who report placing emphasis
Looking at various aspects of how on memorising information do not
Given substantial investment in students learn, PISA finds that: always achieve better results, while
education and its importance to those who elaborate what they learn
societies’ and students’ future well- The aspect most closely associated tend to do well (see Tables 4.6 and 4.7
being, it is disappointing that a with performance is “controlling the in the full report).
significant minority of 15-year-olds learning process”. Students were asked
display negative attitudes towards to what extent they: Co-operative and competitive
learning and a lack of engagement learning strategies both have their
figure out exactly what they need to
with school. In most countries, more place in effective learning. Students
learn;
than a quarter of all students say that who report adopting competitive
school is a place they do not want to work out as they go what concepts learning strategies tend to do better
go – ranging from less than 20% in they still have not really understood; than those who do not. Likewise,
Denmark, Mexico, Portugal and look for additional information students who report that they learn co-
Sweden to more than 35% of students when they do not understand; operatively tend do better than those
in Belgium, Canada, France, Hungary, who do not. This finding suggests that
Italy and the United States force themselves to check whether co-operative and competitive learning
(see www.pisa.oecd.org). they remember what they have can each help in certain situations:
learned; and they are complementary rather than
Student interest in reading and make sure they have remembered alternative strategies (see Tables 4.8
mathematics also varies widely and is the most important things. and 4.9 in the full report).
closely associated with performance.
About half of 15-year-olds are In every country, the quarter of Without further analysis, it cannot be
generally positive about reading. students who say they adopt such concluded that certain learning
However, this varies across countries, strategies the most have reading strategies cause better performance: it
with fewer than one-third of students scores significantly above the might simply be that students who do
in Belgium and Korea but around two- quarter that adopt them the least. well for other reasons are more likely to
thirds in Denmark, Mexico and In some countries, the gap is as much adopt the most effective strategies.
Portugal agreeing that reading is fun as a full proficiency level – for Nonetheless, the findings do show
and that they would not want to give it example, the difference between the what kind of learning is associated with
up. Interest in mathematics varies even average student with reading literacy at success. Schools and parents need to
more between countries, with in most Level 2 and the average student at consider how to help students to adopt
countries only a small proportion of Level 3 (see Table 4.5 in the full techniques that allow them to regulate
15-year-olds seeing mathematics as report). their own learning effectively.

14
 Reading for enjoyment and reading literacy

Students were asked how much they Are keener readers also better readers?
read for enjoyment. A substantial On average in each country, the
percentage in every country say not at answer is yes. For example, the final
all. The proportion varies from a column of the table shows that
majority of students in Japan, to below students in Australia who read 1-2
20% in Brazil, Latvia, Mexico, Portugal hours a day for enjoyment score 92
and the Russian Federation, as shown points higher on average on the
in the first column of the table. Of reading scale than those who do not
those students who do read for read for enjoyment. This is equivalent
enjoyment, most do so for under an to more than one proficiency level –
hour a day. The second column shows for example, the difference between
how many – 11% on average across being able to perform Level 3 reading
OECD countries - are keener readers, tasks and Level 4 reading tasks – see
spending 1-2 hours a day reading for examples on page 7 above.
enjoyment.
However, it is important to note that:

Country Percent reading for enjoyment Difference between Whereas, within countries, there is a
a) Not at all b) 1-2 hours average reading scores of clear positive association between
a day students in a) and b) reading for enjoyment and
% % More than one proficiency level performance, countries with fewer
keen readers do not necessarily
Australia 33 12 92
Germany
produce fewer good ones. As an
42 9 84
Switzerland 35 8 83 illustration, Japanese students are
Finland 22 18 79 least likely to read for enjoyment, but
Canada 33 10 77 show high average reading
New Zealand 30 10 76 performance. Conversely, in Latvia
Iceland 30 7 73 and Greece, with relatively many
Sweden 36 9 72 keen readers, mean performance
Between half and one proficiency level
remains below-average.

United Kingdom 29 9 70 It is not clear to what extent reading

Latvia 18 20 67
for enjoyment leads to higher reading
France 30 11 67
literacy, or the other way around, or
Norway 35 8 65
Ireland 33 12 65
to what extent some other aspect of
Czech Republic 26 13 63 students’ background contributes to
Austria 41 9 63 both. Nevertheless, the association
Portugal 18 12 62 between engaging in reading and
United States 41 8 59 being good at it is an important one,
Belgium 42 9 59 indicating that it may be productive
Denmark 27 9 57 to encourage both.
Spain 32 9 54
Hungary 26 13 53
Russian Fed. 19 17 49
Poland 24 16 49
Korea 31 12 41
Italy 31 13 41
Netherlands 43 6 40

Less than half of one proficiency level

Japan 55 8 27
Luxembourg 38 12 25
Brazil 19 17 25
Greece 22 20 19
Mexico 14 12 6

15
Different results by gender

Policy-makers have, historically, given

considerable priority to issues of gender A. Reading literacy: females B. Mathematics: males do
equity in education, with particular do better in all countries better in half of the countries
attention to disadvantages faced by (in other countries, no statistically significant difference)

girls and women. PISA’s results point

to the success of many countries’ Females Austria 27
efforts, but also to a growing problem at least half a proficiency level ahead: Brazil 27
for males, particularly in reading 53 Korea 27
Latvia

Average score advantage (points),

19

Average score advantage (points),

literacy. In mathematical literacy, there Finland 51 Portugal
remains a measurable disadvantage for 46 Spain 18
New Zealand
females in about half of the countries 43 Luxembourg 15
Norway
40 Denmark 15
while in scientific literacy gender Iceland

females

males
38 Germany 15
differences tend to average out. Russian Fed.
38 Switzerland 14
Italy
37 France 14
Czech Republic
37 Ireland 13
Greece
37 Liechtenstein 12
Sweden
36 Czech Republic 12
Poland
Norway 11
Canada 10
Females
less than half a proficiency level ahead:
Germany 35
Australia 34
Belgium 33 C. Science: Females
Canada 32 do better in three countries
Hungary 32
Liechtenstein 31

(points), females
Switzerland 30

Average score
23

advantage
Latvia
Average score advantage (points),

Japan 30
Russian Fed. 14
Netherlands 30
New Zealand 12
France 29
Ireland 29
females

United States 29
Luxembourg 27
Austria 26
United Kingdom 26 C. Science: Males do better
Denmark 25
25
in three countries
Portugal
Spain 24
Mexico 20
(points), males
Average score

Brazil 17 19
advantage

Korea
Korea 14 12
Austria
Denmark 12

16
Key findings females in mathematical literacy, much on memorisation strategies, which are
of this is attributable to there being less consistently associated with better
In all countries, females are on more males among the better performance than strategies relating
average better readers than males. performers and the advantage of males new knowledge to existing knowledge,
The most striking gender difference disappears when comparing the which males report that they prefer
revealed by PISA 2000 is that females number of low performers (see (see Tables 4.5, 4.6 and 4.7 in the full
consistently outperform males in Table 5.2b in the full report). These report).
reading literacy. As the table shows, findings suggest that the
this difference is not small. In New underachievement of males in reading
Zealand, for example, the mean score is a significant challenge for education
of males is 507 points, close to the policy, in terms both of closing the
average for all students across OECD gender gap and of reducing the
countries, but the mean score of proportion of students at the lowest
females in New Zealand is 553 points, levels of proficiency.
more than the mean for all students in
any country (see Table 5.1a in the full Some countries show that large
report). Females score, on average gender differences are not inevitable.
across OECD countries, 45 points There is significant variation between
higher than males on the reflection countries in the size of gender
and evaluation scale, compared with differences. Indeed, the results of PISA
29 points on the interpreting texts 2000 suggest that some countries
scale and 24 points on the retrieving provide a learning environment or
information scale. These differences broader context that benefits both
may be influenced by differences in genders equally. By contrast, the
reading interests: males report reading enduring differences in other countries,
more comics, newspapers and web- and the widespread disadvantage now
pages, females more novels (see faced by many young males in reading
Table 5.1b in the full report). literacy, require serious policy
attention.
In mathematical and scientific
literacy, gender differences are Females tend to express greater
smaller than in reading. For interest in reading, and males in
mathematical literacy, males perform mathematics. On average across
on average 11 points higher across OECD countries, approximately one-
OECD countries, but in only half of half of all males but only a quarter of
the PISA countries was the difference females say that they read only when
statistically significant. In the case of they have to (see Figure 5.4 in the full
scientific literacy there is a significant report). By contrast, females tend to
difference in favour of either males or have less interest than males in
females in only six (see Table 5.1a in mathematics. The close
the full report). interrelationship between subject
interest and learning outcomes suggests
Males are more likely to under- that the different habits and interests
perform in reading. But females are of females and males may have far-
not more likely to underperform in reaching consequences for learning
mathematics. An important policy that education policy needs to address
concern emerges from the large gender (see Figure 4.1 in the full report).
differences in reading literacy among
the lowest performing students. In all Males and females tend to adopt
participating countries, males are more different learning strategies. Females
likely than females to be at Level 1 or are more likely to report that they take
below in reading – in the case of control of their own learning, which is
Finland over three times as likely (see a strategy that tends to be associated
Table 5.2a in the full report). While with learning success. However,
males do, on average, better than females tend to be more likely to rely

17
The importance of family background

It is well established that students who

Differences in reading scores by parent's occupation
come from more advantaged family
backgrounds, in terms of factors such as
Average for bottom quarter of students Average for top quarter of students
by parental occupation by parental occupation parental education and occupation,
and resources in the home, perform
Difference in points better at school. PISA was able to look
internationally both at the strength of
this relationship and at its importance
350

400

450

500

550

600
in explaining overall differences in
student performance.

Narrowest differences by student background

Key findings

The association between family

background and student performance
differs greatly from one country to
another. The figure illustrates this
point in the case of average differences
in reading scores according to parental
occupation. PISA ranked occupations
on an internationally recognised scale
based on their measured economic
value. On the figure, the left-hand end
of each bar represents the mean
reading score of the 25% of students in
each country with the lowest-ranking
parental occupations. The right-hand
end shows the mean reading score of
the 25% of students whose parents
have the highest-ranking occupations.
Thus, the length of each bar shows the
mean difference in reading literacy
between the least advantaged students
and the most advantaged ones, based
on their parents’ occupations. In
Korea, at one extreme, there is a
relatively small difference of 33 score
Widest differences by student background

points between the top and bottom

quarters, equivalent to less than half a
proficiency level. In Germany and
Switzerland, on the other hand, the
difference is around 114 score points,
or more than one proficiency level
(see Table 6.1a in the full report).

Not all students from

disadvantaged family backgrounds
perform poorly. In some countries even
the bottom quarter of students show
medium rather than low performance
scores. For example, ranked by parents’
occupational status, students in the
bottom quarter in Canada, Finland,
350

400

450

500

550

600

and Korea score above the average for

all students across the OECD (500)
(see Table 6.1 in the full report).
18
PISA distinguished the effects of many performance is whether students have
specific aspects of family background. items associated with “classical
In addition to the effect of parental culture”, such as literature and works of
occupational status, shown opposite, it art, in their homes (see Table 6.3 in
found that: the full report). While possession of
such “advantages” is related to other
Higher parental education and home background characteristics, its
more social and cultural effects in isolation remain consistently
communication between parents and strong (see Table 8.2 in the full report).
their children are associated with The effects are higher in reading
better student performance. Students literacy than in mathematical and
whose mothers have not completed scientific literacy, emphasising the
upper secondary education have a ways in which educational benefit
particularly strong disadvantage with, accrues from home-based access to
on average across OECD countries, literature and other cultural
reading scores 44 points lower than possessions.
those whose mothers have completed
upper secondary education (see Living with only one parent is, on
Table 6.7 in the full report). The average, associated with lower
impact of mothers’ completion of student performance. On average
tertiary education is weaker and less across OECD countries, students who new language in a home environment
consistent across countries (see live with one parent score 12 points that may not facilitate this learning.
Table 6.7 in the full report). Parental lower in reading literacy than students In either case they may be in need of
education is closely interrelated with who live with two parents, all else special attention. PISA finds a more
other family background factors. being equal (see Table 8.2 in the full mixed picture in the case of students
However, when other family report). In some countries, there is not born in the country but whose parents
background factors are equal, each a significant difference. However, two immigrated. In some countries, they do
additional year of parental education of the countries where there is the not perform significantly differently
still adds almost 4.7 points to the widest difference, the United Kingdom from native born students, but in four
reading scores of students (see Table 8.2 and the United States, are also the two countries – Belgium, Germany,
in the full report). PISA also asked where the proportion of students living Luxembourg and the Netherlands –
students how they interact with their with only one parent is largest (see there is a gap of more than one reading
parents in aspects ranging from Tables 6.9 and 8.2 in the full report). proficiency level (see Table 6.10 in the
discussing books to eating meals full report).
together. Those interacting more have Educational disadvantage among
higher reading scores on average – children born outside the country
cultural communication (e.g., varies greatly. In several countries,
discussing books) playing a greater role non-native born students show much
(on average across OECD countries 59 lower reading literacy scores than
points between the top and bottom students who were born (as were their
quarters, see Table 6.6) than social parents) in the country. They are at
communication (on average 30 score least 71 points behind native-born
points, see Table 6.5). students in 10 out of the 15 countries
with more than 3 per cent of
Students from wealthier families immigrant students, but in some
tend do better, but the relationship countries the gap is much smaller
with possessions relating to classical (see Table 6.10 in the full report). A
culture is stronger. Students were significant difference, 26 points on
asked about a range of possessions in average across OECD countries,
their homes. Students from the remains even when other factors of
wealthiest families typically do better family background are considered
than students from the least wealthy (see Table 8.2 in the full report). These
families, but the differences in students may be academically
performance are modest in many disadvantaged either because they are
countries (see Table 6.2 in the full immigrants entering a new education
report). A stronger predictor of system or because they need to learn a
19
Can schools compensate for socio-economic background?

Variation in student performance in reading School systems aim to provide equality

of opportunity, giving the same
chances to children regardless of the
family circumstances to which they
Between schools Within schools
were born. In practice, children with
different backgrounds do not do

100
80

60

40

20

20

40

60

80
equally well. To what extent is this

0
because of home influences, and to
what extent because of experiences at
school? PISA provides some clues, in
ways that can inform strategies to
improve the performance of less
advantaged students.

Key findings

Differences in the performance

between schools account for much of
the variation in student performance
in some countries. As shown in the
figure, there are wide differences in
reading literacy performance among
schools in countries such as Austria,
Belgium, Germany, Hungary and
Poland where these account for much
of the variation in overall student
performance. Conversely, in countries
such as New Zealand and Norway,
differences are mainly within schools.
The countries with the greatest
differences between schools tend to be
those that send students to different
kinds of secondary school, often on the
basis of prior performance in school
(see Table 2.4 in the full report).
80

60

40

20

20

40

60

80

100

Index of variation (average total variation in OECD countries = 100)

20
 Varying amounts of within-school
and between-school differences in
performance are associated with socio-
economic background. Students in
better-performing schools often come
from more advantaged families. PISA
also suggests that the effects of socio-
economic clustering are larger in school
systems with differentiated school systems
than in systems in which the curriculum
does not vary significantly among schools
(see Table 8.3 in the full report).

The socio-economic composition of

a school’s student population is an
even stronger predictor of student
performance than individual home
background. PISA shows, for example,
that two students with the same family
characteristics going to different
schools - one with a higher and one
with a lower socio-economic profile -
could expect to be further apart in
reading literacy than two students from
different backgrounds going to the same
school. Although this phenomenon
has complex causes, it underlines the
potential link between the socio-
economic segregation of students in
different schools and the polarisation of
students by performance (see Figure 8.4
in the full report).

Schools that have a more Policy implications

favourable climate and are better
resourced tend, to varying degrees, to PISA’s findings have important policy In other countries, there is relatively
have more advantaged students. In implications for education systems. In little socio-economic segregation; that
Belgium, Germany and Luxembourg, some countries, students are highly is, schools tend to be similar in their
for example, schools with a more segregated in terms of socio-economic socio-economic intake. Educational
advantaged socio-economic background variables, in part because of policy in these countries might aim to
tend to employ a greater proportion of residential segregation and economic moderate the impact of home
specialist teachers, often because such factors, but also because of features of background through measures designed
teachers are employed in the more the schooling system. Educational to improve school resources and to
advanced tracks leading to entry into policy in such countries might reduce within-school segregation in
university. Schools with a higher socio- attempt to moderate the impact of accordance with the economic,
economic intake also seem to have a home background on student cultural and social status of students. In
better disciplinary climate, as reported performance by reducing the extent the end, of course, what will matter
by students, particularly in Italy, Japan, of segregation along socio-economic most is how effectively those resources
Spain, the United Kingdom and the lines, or by allocating resources to are used.
United States. Finally, students in schools differentially. In these
schools with low economic, social and countries, it is important to
cultural status also tend not to use understand how the allocation of
school resources as regularly as students school resources within a country is
in better-off schools (see Figures 8.5 related to the socio-economic intake of
and 8.5a in the full report). its school.

21
What can schools do to make a difference?

Home background influences Qualified teachers are among a OECD countries, a positive and
educational success, and schools’ socio- school’s most valuable resources. statistically significant association with
economic status may reinforce its PISA asked school principals to student performance (see Table 8.5 in
effects. Equally important, PISA also indicate the percentage of teachers the full report). These include:
identified several things that schools with a university-level qualification in teacher-related factors affecting
can do that are associated with student their respective subject area. Having school climate, such as teacher
success. In identifying a constellation more of these teachers is associated, on expectations of students;
of factors that interact to influence average across OECD countries, with
performance, this first report does not better student results. For example, in teacher morale and commitment;
claim to provide causal links between reading, a 25 percentage point increase and
what schools do and how their students in the proportion of teachers with a
perform. Nonetheless, the initial university-level qualification in the school autonomy.
findings offer some clues about the relevant subject domain is associated
conditions in schools that are most with an advantage of 9 points on the Some aspects of classroom practice
closely associated with success. The reading literacy scale, on average across are associated with better student
following findings consider the separate OECD countries, the other school performance. Three such factors, as
effect of each factor identified, after factors measured by PISA being equal perceived by students, show a positive
associations with the other observed (see Table 8.5 in the full report). and statistically significant association
school and home background factors with student performance:
have been taken into account. The The ratio of students to teaching teacher-student relations;
findings presented below tend to be staff matters most where it is
disciplinary climate of the
similar for reading, mathematical and relatively high. Among schools where
classroom; and
scientific literacy. the number of students for each
member of the teaching staff exceeds the extent to which teachers
Students’ use of school resources is 25, the mean performance of students is emphasise academic performance
more closely associated with student markedly lower, the higher the ratio. In and place high demands on
performance than is the physical the more typical range, between 10 to students.
infrastructure of schools. Students 25 students per teacher, there is a much
were asked about their use of their weaker association with performance in
school’s library, computers, calculators, reading literacy. Schools with fewer
laboratories and Internet connection. than 10 students per member of the
In schools where usage is relatively teaching staff actually score slightly
high, mean reading scores tend to be below the OECD average which may be
higher, even when other factors are because many such schools serve
discounted (see Table 8.5 in the full students with special needs (see
report). Deficiencies in the quality of Table 8.5 in the full report).
the school’s physical or material
infrastructure, as reported by the Some aspects of school policy and
principal, tend to have a much weaker practice tend to be associated with
impact than students’ use of these better student performance. Three
resources (see Tables 7.10, 7.11 and 8.5 such factors, as perceived by school
in the full report). principals, show, on average across

22
The first two of these factors are Together with home background
stronger than the third (see Tables 7.2, factors, the factors explain 72 per cent
7.3 and 8.5 in the full report). of the variation among schools within
countries and 43 per cent of the
Successful students are more likely variation between countries (see
to do homework. The other school Table 8.5 in the full report).
factor with a close association with
student success is homework. Within These findings give a first indication of
countries, students who do more PISA results. Much further research
homework are, on average across and analysis will be needed to identify
countries, likely to perform better in how each school factor operates,
reading literacy. The quarter who do interacts with home background and
the most homework score on average influences school and student
44 points higher than the quarter performance. Further thematic reports
doing the least. This association is in 2002 and 2003 will seek to
strongest in countries where students understand in more detail why some
do more homework on average (see countries and schools perform better
Table 7.6 in the full report). and achieve more equitable learning
outcomes than others.
Further research is needed

Overall, the combined influence of this

set of school-level factors explains 31
per cent of the variation in reading
literacy performance among schools
within countries, and 21 per cent of
the variation among countries.

23
EDUCATION AND SKILLS

Knowledge and Skills for Life

FIRST RESULTS FROM PISA 2000
Are students well prepared to meet the challenges of the future? Are they able to
analyse, reason and communicate their ideas effectively? Do they have the
capacity to continue learning throughout life? These are questions that parents,
students, the public and those who run education systems continually ask.

Knowledge and Skills for Life, the report summarised in this brochure, provides
some answers. It assesses how far students near the end of compulsory
education have acquired some of the knowledge and skills that are essential for
full participation in society. It presents evidence on student performance in
reading, mathematics and scientific literacy, reveals factors that influence the
development of these skills at home and at school, and examines what the
implications are for policy development.

The report shows considerable variation in levels of knowledge and skills between
students, schools and countries. The extent to which the socio-economic
background of students and schools affects student performance varies. Some
countries have managed to mitigate the influence of social background and some
have done that while achieving a high overall mean performance. This is a
noteworthy achievement. Will other countries take up the challenge?

The OECD Programme for International Student Assessment (PISA) Further information and
is a collaborative process among the 30 Member countries of the OECD and on-line ordering:
some non-OECD countries, bringing together scientific expertise from participating To order the Report
countries and steered jointly by their governments on the basis of shared, policy- (OECD code: 96 2001 14 1 P 1)
driven interests. PISA is an unprecedented attempt to measure student www.oecd.org
achievement across all OECD countries and some non-OECD countries, as is
evident from some of its features: Data underlying the Report
www.pisa.oecd.org
The literacy approach: PISA aims to define each domain (reading, mathematics
and science) not merely in terms of mastery of the schools curriculum, but in All OECD books and periodicals
terms of the knowledge and skills needed for full participation in society. are now available on-line
A long-term commitment: Over the decade to come, it will enable countries www.SourceOECD.org
regularly and predictably to monitor their progress in meeting key learning
objectives.
The age-group covered: By assessing 15-year-olds, i.e. young people near the
end of their compulsory education, PISA provides a significant indication of the
overall performance of school systems.
The relevance to lifelong learning: PISA does not limit itself to assessing
students’ knowledge and skills but also asks them to report on their own, self-
regulated learning, their motivation to learn and their preferences for different
types of learning situation.