Oxford Review of EducationVol. 35, No. 5, October 2009, pp. .docx

Oxford Review of Education
Vol. 35, No. 5, October 2009, pp. 587–600
ISSN 0305-4985 (print)/ISSN 1465-3915 (online)/09/050587–14
© 2009 Taylor & Francis
DOI: 10.1080/03054980903216317
The psychology of education:
achievements and challenges
W. Ray Crozier*
University of East Anglia
Taylor and
FrancisCORE_A_421805.sgm10.1080/03054980903216317Oxfo
rd Review of Education0305-4985 (print)/1465-3915
(online)Original Article2009Taylor &
Francis355000000October 2009Professor W. [email protected]
Psychology has been closely involved with educational
research, policy and practice since its emer-
gence as a scientific discipline in the late 19th century but it has
occupied a less certain place in
recent years. This paper describes aspects of the organisational
framework of psychology as a disci-
pline and draws upon Research Assessment Exercise data to
gain a picture of the research capacity
of the psychology of education. The division of research in
universities between psychology depart-
ments and education departments is identified and its relevance
to the ‘demographic crisis’ in
educational research considered. Three areas of recent
applications of cognitive psychology are
reviewed and promising intervention studies are outlined. The
implications of demographic trends

in educational research together with indications of the limited
capacity of the psychology of educa-
tion pose threats to the potential to develop original education-
based psychological research and to
overcome limitations of much existing research.
Introduction
Psychology is a broad discipline, and its constituent areas—
which might be identified
as social psychology, developmental psychology, biological
psychology, cognitive
psychology, and individual differences – are applied to the
study of education and to
educational policy and practice. How could they not, one might
think, given psychol-
ogists’ extensive studies of processes of learning and memory;
of children’s develop-
ment, their acquisition of skills such as reading and writing,
their understanding of
number, science, or moral principles? Psychology also studies
children of exceptional
abilities, problems of adjustment, parent–child relations, social
interactions and rela-
tionships, friendships, peer influences, inter- and intra-group
processes, bullying,
attitudes, motivation and emotion, and so on. Nevertheless,
despite these endeav-
ours the relation of psychology to education is problematic.
There are debates within
psychology as to how these topics should be conceptualised and
studied and there
*School of Social Work and Psychology, University of East
Anglia, Norwich NR4 7TJ, UK.
Email: [email protected]

588 W. R. Crozier
are sustained criticisms of the nature of the discipline from
within and outside
psychology.
The concerns of this special issue are the relations between
education and its disci-
plines in the context of the ‘demographic crisis’ in education
research (Mills et al.,
2006). I address these concerns in relation to psychology by
offering a brief overview
of psychology’s major contributions to education in the past, a
description of the
organisational framework of psychology and its relation to
education, consideration
of the state of research in the psychology of education, and
reflections on selected
current applications. One theme relates to the implications of
the distinction between
research in the psychology of education that is undertaken in
university psychology
departments and research that is located in departments and
schools of education.
The distinction is relevant for a number of reasons. First, the
demographic profiles
of psychology and education are quite distinct when these are
defined in terms of
Research Assessment Exercise (RAE) units of assessment (Mills
et al., 2006).
Second, the division may contribute to the fragmentation of
research in the psychol-
ogy of education. On the one hand, the psychology of education

plays only a small
part in the research activities of psychology departments, where
theoretical develop-
ments and methodological originality and refinement tend to be
valued more highly
than applications. On the other hand, psychological research is
scattered across many
education departments, few of which have a significant presence
of psychologists
(Crozier, 2007). The demographic crisis ought to be considered
in this context.
Psychology’s major contributions in the past
Psychology has been closely involved with educational
research, policy and practice
since its emergence as a scientific discipline in the final years
of the 19th century. The
first professors of education in British day training colleges and
universities drew
upon psychological principles in their teaching and textbooks,
and their work in
education departments preceded, and in a number of cases led
directly to the estab-
lishment of departments of psychology in the universities
(Thomas, 1996). The
research conducted in education departments contributed to the
development of
psychology as a research based discipline; theoretical and
methodological advances in
psychology were located in both education and psychology
departments. A similar
picture emerges if we focus on the application of psychology to
the education of chil-
dren with special educational needs; the seminal contributions
of Ann and Alan

Clarke, Neil O’Connor, and Jack and Barbara Tizard in the
1950s were made in
psychology and education departments (Hall, 2008);
psychologists would move from
one to the other without detriment to their research.
Thus, psychology played a key role as ‘a foundation discipline’
in the emergence of
educational studies. Psychological theory and research were
applied to a range of
educational matters, from the study of learning processes in
individuals, including
students with learning difficulties, to pedagogy. Furthermore,
unlike other disciplines
contributing to educational research, professional psychologists
established a statu-
tory role within the education system. By the 1920s the role of
the educational
Psychology of education 589
psychologist was established in the British education system,
particularly in the field
of special educational needs. Although educational
psychologists have a statutory role
in assessment and the provision of services for children, they
also regard research as
an important part of their role.
Yet recent commentators on psychology, both within and
beyond the discipline and
profession, see its influence as having waned and the value of
its contributions as less
certain. Thomas (1996, p. 240) points out that doubts about the

application of general
psychological principles to teaching individuals were expressed
as early as 1904.
Norwich (2000) analyses psychology’s difficulties in terms of
uncertainties within
psychology, uncertainties within education, and uncertainties in
the relations between
the two. Anxieties about the psychology of education pre-date
the concerns about the
demographic crisis in educational research and scholarship that
are the focus of this
issue, yet will no doubt contribute to the crisis and the form of
its resolution.
The mechanisms and sites of production
A useful starting point is to consider the role of the British
Psychological Society
(BPS). It accredits undergraduate degree courses, setting out a
curriculum to be
followed and minimum standards of resources. It currently
approves nearly 400
degree courses at over 100 institutions, approximately 11,000
students graduating
each year with an accredited degree. The BPS also accredits
postgraduate profes-
sional training courses including courses leading to the
qualification for educational
psychologists (most recently the Doctorate in Educational and
Child Psychology,
which admitted its first cohort in 2006). The Society’s structure
includes Sections,
which bring together researchers in areas of psychology, for
example the Education
Section, and Divisions, which are organisations of professional
psychologists, includ-

ing a division for educational psychologists. The BPS organises
and facilitates confer-
ences and symposia and publishes major research journals
including the British
Journal of Educational Psychology, founded in 1931. In
summary, research in psychol-
ogy, including the psychology of education, is supported
directly and indirectly by a
national organisation which is outside the university and
research centre systems but
is closely linked to them. There are inevitably tensions within
an organisation that is
the representative body for British psychologists and which has
to accommodate
diverse positions within a broad discipline with conflicting
theoretical and method-
ological approaches. Notwithstanding these tensions, it has
helped to provide coher-
ence to psychology. There is a degree of consensus on what
constitutes the
mainstream and the BPS plays a significant role in this.
The good health of psychology as a discipline is evident in the
numbers studying it
at university (HESA data for 2006–2007 show that it is among
the most popular
subjects in higher education, with 72,475 students including
6415 full-time postgrad-
uate students) and in rapidly growing numbers at GCSE, A-level
and foundation
levels. This helps sustain a sizeable academic community of
lecturers and, indirectly,
researchers. A survey conducted in 2003 showed that
psychology departments had
an average of 5.6 postdoctoral researchers and 8.3 final-year
PhD students (Norgate

590 W. R. Crozier
& Ginsborg, 2006), suggesting that future generations of
researchers are being
nurtured.
Nationally, student numbers on education degree courses are
comparable to
psychology numbers. However, psychology plays only a small
part in the education
degree curriculum and this part has been shrinking since the
1980s following
government-led changes to teacher training which introduced a
greater emphasis on
students’ experience in schools and, within the shorter time
spent in university,
increased focus on curriculum subjects. Consequently,
education students spend
much less time on psychology, which tends to form part of a
general professional
studies component of the university course (Norwich, 2000, p.
165). This trend
augments the problems faced by research in the psychology of
education that are
associated with the demographic trends identified by Mills et al.
(2006); education
departments might not regard it as a priority to replace a vacant
psychology post with
a psychologist.
In summary, the discipline of psychology is currently in a
strong position, attracting
substantial numbers of undergraduate and postgraduate students

and supported by
an organisation that has close links with university departments
and professional
practitioners and which has considerable influence on the
university curriculum. The
position of psychology within education departments is less
secure. How might this
impact upon research in the psychology of education? It is
relevant to note that research
in the psychology of education is located in university
psychology departments as well
as in education departments. As an illustration, a snapshot of
the 83 articles published
in two recent volumes of the British Journal of Educational
Psychology shows that 50%
of the articles had first-authors whose affiliation was in
psychology compared to 32%
in education and 5% in educational psychology (the remainder
had other affiliations
in health, communication and so on). The proportions are
similar for UK and inter-
national authors (68% of authors were based outside the UK). A
different picture
emerges if comparable issues of the British Educational
Research Journal are analysed.
Four per cent of affiliations were from the Psychology and
Human Development
Department at the Institute of Education, London; no other
psychology departments
were represented. Eight per cent of articles had international
authors.
One picture of the capacity of research in the psychology of
education can be
obtained from analysis of RAE information on university
research publications.

Crozier (2007) reported an analysis of the outputs submitted by
university institutions
to the RAE 2001 Education Unit of Assessment (UoA) (HERO,
2002). A systematic
search for authors of publications in psychology journals
identified 235 individuals
who were located in 56 of the 82 institutions making
submissions; this constitutes
some 10% of the 2330 individuals whose work was submitted to
the Education UoA.
The psychologists (defined in this way) were dispersed across
institutions. The mean
number of psychologists per institution where they were located
was 4.2; 17 institu-
tions were represented by a single psychologist while only five
submitted 12 or more.
The psychological research submitted to the Education unit of
assessment was also
dispersed, and psychologists’ outputs were published in a large
number of different
journals (351 articles published in 106 journals). The journals
could be classified into
four groups: general psychology (22% of journal submissions);
psychology of educa-
tion (32%); special educational needs (36%); counselling,
psychotherapy and psychi-
atry (9%). Few of the journals were international and not many
are rated as having
‘high impact’ in the social sciences.
These data omit research in education that was submitted to the

Psychology UoA
and thus this representation of psychology of education research
underestimates its
capacity. Nevertheless, when considering the impact of
demographic changes on
research in education it is important to have some sense of the
capacity of its constit-
uent disciplines. This analysis shows that psychology was
dispersed across university
education departments, few of which had a concentration of
psychologists. This may
diminish the capacity to undertake sustained, large-scale
psychological research. If
research were published in a smaller number of ‘high impact’
journals this could
encourage more stringent criteria for publication, potentially
requiring authors to
produce more convincing evidence. It would facilitate the
cumulative development of
a knowledge base as advocated, among others, by the National
Education Research
Forum (2000). Details of the outputs of the 2008 RAE are not
available at the time
of writing; it would be valuable to conduct a comparable study
to see if capacity has
changed and what the implications of findings might be for the
future of psychology
applied to education.
In summary, psychology faces difficulties in its position in
education degree and
teacher education courses and, as this analysis of RAE data
implies, in educational
research. Nevertheless, high quality applied research is being
conducted and I now
consider issues in the application of psychology to education

before reviewing exam-
ples of such research.
Selected contributions to the psychology of education
Applications of psychology
As indicated at the beginning of this paper psychology is a
broad discipline. It eschews
overarching theory and its explanatory models tend to have
specific areas of focus.
Cognitive psychology is the dominant paradigm within the
mainstream. The cogni-
tive process models that have been constructed are increasingly
augmented by cogni-
tive neuroscience models that aim to relate these processes to
brain structures and
functions. The influence of European-led contributions to
cognitive psychology—the
theories and research of Piaget and Vygotsky for example—has
grown and has
provided a framework for the development of theories and
research that aim to embed
cognitive processes within their cultural and social context.
This trend is evident in
both psychology and the psychology of education.
Theoretical and methodological developments have gone hand
in hand with tech-
nological innovations. The computer metaphor of mind led to
the development of
cognitive psychology and, more recently, to connectionist
models that aim to model
the neural networks that underpin learning. Innovations in
information technology
have facilitated increasingly sophisticated studies of perception

and memory.
592 W. R. Crozier
Developments in brain imaging have encouraged the rapid
growth of cognitive
neuropsychology. Much cognitive research is laboratory based
although these labora-
tories can be located in or are associated with applied settings,
in medicine, for exam-
ple. How is this research applied to education? To begin to
answer this, it is useful to
consider the nature of applied psychology. Smith (2005)
proposed a typology of rela-
tions between developmental psychology and education.
1. Independence. Each field pursues its research agenda without
reference to the
other. Much psychological research is of this nature and much
educational
research does not draw upon psychology, explicitly rejects it, or
seeks a different
kind of psychology to the one that is currently dominant.
Nevertheless, the
education of children must draw upon some kind of theory of
child development,
if only to select age-appropriate materials and teaching
strategies. This theory can
be built up through teachers’ classroom experience to exert a
powerful influence
on practice and in socialising new teachers into ‘what works’.
Policies and guide-
lines produced by the government and agencies also exert an
influence on prac-

tice; these do not necessarily draw upon educational psychology
and may or may
not be consistent with teachers’ understanding of what works.
Theories of prac-
tice are systematised and promoted through instruction and
textbooks in peda-
gogy, again without necessarily drawing upon social science
theory. It can be
bolstered by empirical research that uses social science
methodology. For exam-
ple, research investigates the influence of ability grouping on
attainment
(reviewed by Hallam, 2002), using established research
methodology without
necessarily making reference to psychological theorising on
learning, the role of
social groups in learning or the nature or measurement of
ability.
2. Relevance. Smith discusses this in terms of speculation that
developments in
one field might be useful to the other: relevance is asserted but
not pursued.
Psychologists frequently allude to possible educational benefits
of their work
without spelling these out. Also, there are many claims, often
made by commer-
cial organisations, about the effectiveness of particular forms of
instruction being
based on findings about brain function or learning styles, and
psychologists have
a valuable critical role here, helping education professionals to
understand what
claims can be supported by current psychological and
neuropsychological
evidence (Goswami, 2008b).

3. Implication. This goes beyond relevance in that development
in one field leads
to greater understanding in the other, and can lead to change in
practice.
Implication can be at a very general level, for example, the
change in understand-
ing the nature of learning from reliance on drill, repetition and
‘mental exercise’
to emphasis on children’s thinking, imagination, play and the
social nature of
learning that followed changes in views on the nature of
childhood influenced by
educational thinkers such as Rousseau and Pestalozzi, but also
by psychologists
such as Piaget, Vygotsky and Bruner. Implication can also be
more specific, for
example the emphasis on whole word reading and learning to
read as a ‘guessing
game’ that was influenced by reading research in the 1970s.
4. Application. This tests whether an implication works in
practice and whether it is
conducive to good practice. I consider this in the next section.
5. Inter-dependence. Rather than application from one field to
the other, this involves
both fields working together at their intersection. Smith
considers this to be rare
in the case of developmental psychology and education. A
similar point is made
by Norwich (2000, p. 169) who argues for a psychology

grounded in educational
practice rather than a field of knowledge to be applied to
education from general
psychology.
The next section offers a selective review of recent research
that has aimed to apply
psychology to education and that offers more than ‘relevance’
in Smith’s terms. It has
to be selective because theories and research in all areas of
psychology have been
applied to education; I emphasise cognitive psychology,
focusing on applied research
carried out in Britain even though the theoretical underpinnings
may have been
developed elsewhere. A more detailed review of research into
cognitive development
together with indication of some of its implications for learning
is provided by
Goswami and Bryant (2007).
Working memory
The model of working memory developed by Baddeley and
Hitch (1974) is
distinctive in that it occupies a central place in contemporary
cognitive psychology
and is one of the few theoretical developments to do so that
originated in Britain.
Rather than conceptualise memory as a series of passive stores
the model analyses
it as a structured process that we use to carry out a wide range
of mental tasks.
Baddeley and Hitch derived a number of predictions about
performance, many
flowing from the proposition that working memory has limited

capacity in its exec-
utive processes and in the phonological or visuo-spatial coding
that keeps material
temporarily active. The educational application of the model has
been investigated
in a number of studies, for example, of children’s mental
arithmetic (Adams &
Hitch, 1997) and specific learning difficulties of children with
Down’s syndrome
(Hulme & Mackenzie, 1992). Research conducted by Gathercole
and her associ-
ates (Gathercole et al., 2006 provide a review) has shown that
individual differ-
ences in children’s performance on working-memory dependent
tasks predict their
subsequent reading, vocabulary and mathematical attainments
even when other
sources of variation, for example in ability test scores, are
statistically controlled.
The research shows that the precise relations between working
memory and school
attainment depend on the child’s age and the curriculum—what
is being learnt at
any one time. It found that teachers were frequently unaware of
their students’
difficulties on memory tasks. Working closely with teachers,
the team has devel-
oped materials for assessing students’ memory task performance
and supporting
their learning in the classroom (Gathercole & Alloway, 2008).
Current projects
include classroom based studies, the construction of screening
techniques, inter-
vention studies, and work with children with learning
difficulties.

594 W. R. Crozier
Learning to read
Our second example concerns issues that would be important for
education whether
or not there was any relevant psychological evidence, namely
how children in a liter-
ate society learn to read. Huey (1908, p. 6) wrote that ‘to
completely analyse what we
do when we read would almost be the acme of a psychologist’s
achievements, for it
would be to describe very many of the most intricate workings
of the human mind’.
Considerable progress has been made in this analysis. For
example, Gough (1972)
drew upon eye-movement studies to construct a model that had
the bold aim of spec-
ifying what takes place during one second of reading, from the
eye alighting on the
text to speech. Recent research based on brain imaging
techniques has begun to spec-
ify the activity in the brain that takes place during this process.
For example, Wolf
(2008, pp. 145–155) extends Gough’s work to offer a
neuroscientific account of the
processes that take place in the first half-second of reading.
Research has also made
significant progress in investigating the neural structures
involved in learning to read
(Goswami, 2008a). Huey would surely have been impressed
with the progress made
in analysing the intricate processes involved in reading and
learning to read. Research

also investigates children’s phonological awareness (Bradley &
Bryant, 1983;
Goswami & Bryant, 2007, pp. 16–17) and comprehension,
including the study of
children who have adequate decoding skills but demonstrate
reading comprehension
difficulties (Nation, 2005).
The teaching of reading has been central to debates about
standards in education
and policies designed to raise standards. Psychological research
has the potential to
inform best practice in teaching reading in various ways. First,
its evidence contrib-
utes to reports that influence national policies, for example, the
Rose Review (Rose,
2006), which was commissioned by the government to review
the teaching of early
reading in England.
Second, psychologists engage in debate with educational
researchers and other
experts on literacy and the teaching of reading, writing and
spelling, for example, a
series of articles on the ‘simple model’ of reading in Literacy,
the journal of the United
Kingdom Literacy Association (Goswami, 2008a; Kirby &
Savage, 2008; Stuart,
Stainthorp & Snowling, 2008). Psychologists also offer
critiques of approaches to
teaching reading and related skills, on the basis of theoretical
understanding of the
processes involved and the quality of the research evidence that
is presented in reports
and policy documents, for example, the insufficient weight
given to randomised

control trials (Wyse & Goswami, 2008).
Third, psychologists are involved in large-scale intervention
studies of approaches
to teaching reading. One example is the ambitious project in
West Dunbartonshire
led by educational psychologist MacKay, which aims for ‘the
total eradication of illit-
eracy’ in one local authority. The multiple-component study had
a cross-lagged design
comparing pre- and post-intervention cohorts, lasted 10 years
and involved all pupils
in the pre-school year in the authority, 35 primary schools and
23 nurseries, resulting
in a total intervention sample of 27,244 children and control
sample of 3,659, all
tested individually (MacKay, 2007). Large-scale studies have
also researched the
social factors that influence reading development. An example
is the Effective
Provision of Pre-School Education (EPPE) project, a
government-funded longitudi-
nal study of the effects of different kinds of pre-school
experience upon children’s
cognitive and social/behavioural development (Sammons et al.,
2004).
Teaching thinking skills
Vygotsky’s theorising has a growing influence on psychology of
learning, particu-

larly his emphasis on the social context of learning and
thinking. His influence has
taken various forms, such as the increased attention given to
social interaction
processes in learning (for example, Howe et al.’s (2007) studies
of group activities
in science education). It has provided researchers and teachers
with concepts such
as zone of proximal development, scaffolding, and guided
participation (Slee &
Shuter, 2003) that have heuristic value. The influence is evident
in educational
programmes designed to teach children thinking skills
explicitly, motivated by the
belief that these skills are not adequately developed through the
standard curricu-
lum. CASE (Cognitive Acceleration for Science Education) has
been developed by
Adey and his associates at King’s College, London since 1982
(Adey et al., 2001).
Adey’s background is in chemistry and science education but
the conceptual frame-
work for the project draws explicitly upon Piaget and Vygotsky.
The programme,
which is delivered by science teachers in science lessons set
aside for the purpose, is
based upon the careful arrangement of management of cognitive
conflict and
attaches considerable importance to social interaction in the
development of cogni-
tion and metacognition. CASE has been systematically
evaluated over several years
through quasi-experimental designs with a longitudinal
dimension (see, for exam-
ple, Adey et al., 2002; Adey, 2004). The success of the
programme and the teaching

materials designed to support it (Adey, 2008) has led to its
widespread implementa-
tion and to its extension to additional curriculum subjects.
Other thinking skills
programmes are being researched, such as ACTS (Activating
Children’s Thinking
Skills), currently being developed in Northern Ireland by
McGuinness et al. (2007).
Our selection of case studies has focused on areas that are at the
heart of the cogni-
tive psychology agenda—memorising, thinking, and the
processes involved in reading
and learning to read. It is important to note that the cognitive
paradigm has pene-
trated many areas of psychology including social psychology,
personality psychology,
and clinical psychology. It has been applied inter alia to the
study of motivation for
learning (Chaplain, 1996), the self-concept (Dweck, 2003) and
specific disorders, for
example, autism, where teaching aids are being developed for
parents and teachers of
young children with autism and Asperger’s syndrome (Baron-
Cohen et al., 2007).
Discussion
My examples are taken largely from one area of psychological
research that has been
applied to education. The emphasis has been on studies of
learning and thinking from
a cognitive perspective, to the neglect of important topics such
as social relationships,

596 W. R. Crozier
children’s emotions and classroom behaviour. Other authors
would make a different
selection and no doubt would interpret differently the research I
have presented.
I have been biased towards intervention studies but I do not
intend to imply that
other kinds of research are not valuable. Much teaching and
learning takes place
within groups and social interactions, as illustrated in the CASE
programme, and this
is a key dimension that warrants consideration at greater length.
I have aimed to show that vigorous lines of research are being
pursued by psychol-
ogists working closely with schools. They collect fresh data in
classrooms. Not all
educational research is of this kind. Much is theoretical while
some draws upon
analysis of secondary data or upon interviews with teachers or
pupils outside the
classroom. Of course, psychologists are not the only researchers
working within
classrooms but their work represents a rich source of data that
are not necessarily
collected by anyone else. Researchers are welcome in busy
schools only as long as
schools value their contribution or see the potential for
application. There is
evidence that this does occur, for example in the
implementation of CASE in
primary and secondary schools, the expansion of thinking skills
programmes, and
psychologists’ contributions to teaching reading. There is

increasing recognition
among researchers of the importance of longitudinal designs and
randomised control
trials in educational research and these necessitate long-term
commitment from both
participating schools and researchers. They demand the
formulation of worthwhile
research questions, skills in research design and analysis, and
serious attention to
ethical issues.
My examples come from both psychology and education
departments. If trends
identified by Mills et al. (2006) are maintained, what would be
the consequences for
psychological research if the capacity of research undertaken
within education depart-
ments were reduced? Psychologists within education
departments have a key role to
play. They work alongside colleagues from other disciplines
involved in educational
research; awareness of how problems and issues are approached
by other disciplines
enriches psychological research and increases psychologists’
sensitivity to criticisms
of their research and the assumptions it makes. They are also
close to educational
practice; they frequently have teaching qualifications and
experience and their contri-
butions to teacher training and teachers’ higher qualifications
and professional devel-
opment bring them into close contact with practitioners. This
helps generate and
refine research questions. It also has practical value in forging
links with schools that
are interested in involvement in research. Not least,

psychologists within schools of
education are primarily interested in education. Psychologists
within psychology
departments have other obligations. In a climate that places
increasing emphasis on
competition for external sources of research funding and where
departments’
research income is related to performance in national research
assessment exercises,
whether these are based on peer evaluation, citation metrics or
success in attracting
funding, the pressure is on psychologists to pursue research that
is valued by their
peers and, to be maximally effective, has to meet the criterion
of being regarded as
‘world leading’. The research that gains these plaudits and
funding is not necessarily
applied research or targeted at education.
Education raises many moral, economic and political
questions—the attention that
is paid to ‘educational standards’ by politicians and the
significance of education for
economic development are but two indications of this. The
study of a complex educa-
tion system, where children are taught mostly in large classes in
schools that can have
thousands of pupils, is necessarily difficult, with very many
variables at different levels
of analysis in interaction with one another. It is unlikely that a
single discipline, partic-
ularly one that emphasises the individual learner, albeit within a

social context, can
answer educational questions on its own. Even when research
seems to have potential
for application there are problems of sustaining the features that
have been identified
as crucial for success and of disseminating the research and
embedding it in practice
across an education system that has competing interest groups
and many ‘stakehold-
ers’. A challenge for even the best research is to avoid
programmes losing their impe-
tus or receiving merely lip service, giving rise, perhaps, to a
sense that ‘it’s Thursday
at 3pm, it must be CASE’.
While psychology is only one discipline that can be brought to
bear on the
complexities of education, its emphasis on the study of the
individual learner will,
I believe, continue to have an important place in educational
research. Individuals
are different, and their differences cannot be reduced to
variables such as social class
or socio-economic factors, important as these undoubtedly are.
Children within the
same family are different from one another, and differences
detected in the first
weeks of life presage later development (Rubin et al., 2009).
Research into child
development is neither static nor a simple accumulation of
facts. It is rich in the
generation of theory, but theory that is submitted to empirical
test. Psychology is
regularly criticised for its individualistic bias, positivism and
reductionism, but
I think these criticisms do not pay sufficient attention to the

role of theory in psychol-
ogy, to its capacity for self-criticism, or to the range of
quantitative and qualitative
methods that characterise it (Crozier, 2007). It fails to
acknowledge the value of test-
ing hypotheses, not in the pursuit of some ‘objective truth’, but
to test our under-
standing: ‘If I really understand X, then I expect to find A to
happen in particular
circumstances, or not to find B, or to find A but not B’, and so
on. If we are not
putting hypotheses to the test then our claims to understanding
join all the other
claims about X and add little to them. Of course, what a
particular hypothesis is and
what counts as a test are complex matters, but dealing with that
complexity is the
nature of social science.
What substantive contributions can psychology make in the
future? Within my
area of focus—cognitive psychology of education—one can
expect that significant
advances in cognitive development, in neuropsychology, and in
understanding the
influence of the social and cultural context on development will
continue. The
outcomes of this research will have heuristic value for
practitioners, providing
concepts that aid understanding of learning and development.
The case studies
outlined here show that schools can benefit from high quality
interventions that draw
upon psychological theory and research, and this promises much
for the future. Yet
there are reasons not to be complacent. This research is

effective only when it is
sustained and involves close collaboration between researchers
and practitioners;
598 W. R. Crozier
these factors are vulnerable to demographic changes. Over-
reliance on research
produced within psychology departments may be a problem for
the reasons
discussed above; and there is no guarantee that the strength of
psychology depart-
ments will persist in the future. There is a risk of a downward
spiral in educational
research where smaller capacity generates less influential
research, which feeds back
into further reductions in capacity. Promising developments in
memory research,
large-scale reading interventions, and teaching thinking
programmes have to be anal-
ysed (which elements of a multi-component intervention are key
and which need to
be strengthened?), improved, evaluated and disseminated, all of
which are demand-
ing of resources.
Finally, there is a need, as Norwich (2000) and Smith (2005)
have argued, to
develop a psychology of education that works closely with other
disciplines to gener-
ate original education-based research. Psychologists who have
training and experi-
ence of education can play an important mediating role in such
developments. What

are the obstacles to the interdependence of education and
psychology advocated by
Smith (2005)? One is the unidirectional nature of influence
where there is absence of
feedback from educational research. For example, research into
working memory
can enrich teaching but there is as yet little evidence for the
application enriching
psychological theory. This is surely desirable, as the learning
that is required in
school and in various forms of apprenticeship makes significant
demands on cogni-
tive, motivational and emotional processes and its study ought
to contribute to the
elaboration of theory. Research into CASE offers a promising
example of this
advancement of theory. There are methodological hurdles too.
Psychology relies on
statistical evidence, which may identify meaningful trends and
offer support for
specific theories but which may not be useful to the
practitioner, who works with
individual students and groups of students, not with ‘samples’.
A similar issue is
faced in clinical research. Cognitive psychology has provided
an analysis of, for
example, anxiety, which clinical researchers have drawn upon to
develop therapy,
testing its effectiveness by means of randomised control trials,
statistical analysis, and
systematic review (for example, a review by Cartwright-Hatton
et al., 2004, of cogni-
tive behaviour therapies applied to children’s anxiety).
Nevertheless, the effective-
ness of the therapy depends on the clinical skills and
sensitivities of the individual

therapist and on the quality of the relationship between
therapist and client. There is
a burgeoning research literature on this application that
contributes to the effective-
ness of the practice and the refinement of the cognitive theory
of anxiety, and which
accommodates critical analyses of practice and theory from
within and outside
psychology.
I have aimed to show through selected examples from one area
of the discipline
the contribution that psychology is making and can make to
education. The
changes advocated by Norwich (2000) and Smith (2005) require
a robust psychol-
ogy of education working closely with other disciplines and
with practitioners.
This is less likely to come about if crises of confidence,
inadequate communica-
tion between disciplines, or demographic trends reduce the
capacity for such
endeavours.
Notes on contributor
Ray Crozier, PhD, is currently a Visiting Fellow at the School
of Social Work and
Psychology, University of East Anglia. Previously he was
Professor of the
Psychology of Education at Cardiff University and Professor of
Psychology at the

University of East Anglia. He is a Fellow of the British
Psychological Society.
References
Adams, J. W. & Hitch, G. J. (1997) Working memory and
children’s mental addition, Journal of
Experimental Child Psychology, 67, 21–38.
Adey, P. (2004) Evidence for long-term effects: promises and
pitfalls, Evaluation and Research in
Education, 18, 83–101.
Adey, P. (Ed.) (2008) Let’s Think handbook: cognitive
acceleration in the primary school (London,
NFER-Nelson).
Adey, P., Robertson, A. & Venville, G. (2002) Effects of a
cognitive acceleration programme on
Year 1 pupils, British Journal of Educational Psychology, 72,
1–25.
Adey, P., Shayer, M. & Yates, C. (2001) Thinking science 3rd
edn. (London, Nelson Thornes).
Baddeley, A. D. & Hitch, G. J. (1974) Working memory, in: G.
Bower (Ed.) The psychology of
learning and motivation (New York, Academic Press), 8, 47–90.
Baron-Cohen, S., Golan, O., Chapman, E. & Granader, Y.
(2007) Transported to a world of
emotion, The Psychologist, 20(2), 76–77.
Bradley, L. & Bryant, P. E. (1983) Categorizing sounds and
learning to read: a causal connection,
Nature, 301, 419–421.

Cartwright-Hatton, S., Roberts, C., Chitsabesan, P., Fothergill,
C. & Harrington, R. (2004)
Systematic review of the efficacy of cognitive behaviour
therapies for childhood and adoles-
cent anxiety disorders, British Journal of Clinical Psychology,
43, 421–436.
Chaplain, R. (1996) Making a strategic withdrawal:
disengagement and self-worth protection in
male pupils, in: J. Rudduck, R. Chaplain & G. Wallace (Eds)
School improvement: what can
pupils tell us? (London, David Fulton), 101–115.
Crozier, W. R. (2007) Capacity and methodology in university
educational research: psychology of
education as a case study, Higher Education Review, 39, 25–41.
Dweck, C. S. (2003) Ability conceptions, motivation and
development, in: L. Smith, C. Rogers &
P. Tomlinson (Eds) Development and motivation: joint
perspectives, British Journal of Educational
Psychology Monograph Series II, 2, 13–27.
Gathercole, S. E. & Alloway, T. P. (2008) Working memory and
learning: a practical guide (London,
Sage).
Gathercole, S. E., Lamont, E. & Alloway, T. P. (2006) Working
memory in the classroom, in:
S. Pickering (Ed.) Working memory and education (London,
Academic Press), 219–240.
Goswami, U. (2008a) Reading, complexity and the brain,
Literacy, 42, 67–74.
Goswami, U. (2008b) Neuroscience and education: from

research to practice, Nature Reviews
Neuroscience, 7, 406–413.
Goswami, U. & Bryant, P. (2007) Children’s cognitive
development and learning: Primary Review
Research Survey 2/1a (Cambridge, University of Cambridge
Faculty of Education).
Gough, P. B. (1972) One second of reading, in: J. Kavanaugh &
I. Mattingley (Eds) Language by
ear and by eye (Cambridge, MA, MIT Press), 331–358.
Hall, J. (2008) Mental deficiency – changing the outlook, The
Psychologist, 21(11), 1006–1007.
Hallam, S. (2002) Ability grouping in schools (London,
Institute of Education University of
London).
HERO (2002) RAE 2001. Available online at
http://www.hero.ac.uk/rae/index.htm (accessed 4
February 2009).
600 W. R. Crozier
Howe, C., Tolmie, A., Thurston, A., Topping, K., Christie, D.,
Livingston, K., Jessiman, E. &
Donaldson, S. (2007) Group work in elementary science:
towards organisational principles
for supporting pupil learning, Learning and Instruction, 17,
549–563.
Huey, E. (1908 [1968 edn.]) The psychology and pedagogy of
reading (Cambridge, MA, MIT Press).

Hulme, C. & Mackenzie, S. (1992) Working memory and severe
learning difficulties (Hove, Lawrence
Erlbaum Associates).
Kirby, J. R. & Savage, R. S. (2008) Can the simple view deal
with the complexities of reading?
Literacy, 42, 75–82.
Mackay, T. (2007) Achieving the vision. The final research
report of the West Dunbartonshire Literacy
Initiative (Dumbarton, West Dunbartonshire Council).
McGuinness, C., Eakin, A., Curry, C., Sheehy, N. & Bunting, B.
(2007) Building thinking skills in
thinking classrooms: ACTS in Northern Ireland. Paper presented
at the 13th International
Conference on Thinking, Norrköping, Sweden. Available online
at http://www.ep.liu.se/eco/
021/vol1/ (accessed 1 December 2008).
Mills, D., Jepson, A., Coxon, T., Easterby-Smith, M., Hawkins,
P. & Spencer, J. (2006)
Demographic review of the UK social sciences (Swindon,
ESRC). Available online at http://www.
esrcsocietytoday.ac.uk/ESRCInfoCentre/Images/Demographic_R
eview_tcm6-13872.pdf
(accessed 1 December, 2008).
Nation, K. (2005) The connections between language and
reading in children with poor reading
comprehension, in: H. W. Catts & A. Kahmi (Eds) The
connections between language and read-
ing disabilities (Mahwah, NJ, LEA), 41–53.
National Educational Research Forum (2000) Building research

capacity: sub-group report (London,
National Educational Research Forum).
Norgate, S. & Ginsborg, J. (2006) A survey of postdoctoral
researchers and final-year PhD students in
psychology: research funding, career needs and working
conditions (Leicester, British Psychological
Society).
Norwich, B. (2000) Education and psychology in interaction:
working with uncertainty in interconnect-
ing systems (London, Routledge).
Rose, J. (2006) Independent review of the teaching of early
reading (Nottingham, DfES Publications).
Rubin, K. H., Coplan, R. J. & Bowker, J. C. (2009) Social
withdrawal in childhood, Annual
Review of Psychology, 60, 141–171.
Sammons, P., Sylva, K., Melhuish, E., Siraj-Blatchford, I.,
Taggart, B., Elliot, K. & Marsh, A.
(2004) The continuing effects of pre-school education at age 7
years. Technical Paper 11 (London,
Institute of Education, University of London).
Slee, P. & Shuter, R. (2003) Child development: thinking about
theories (London, Arnold).
Smith, L. (2005) Education, in: B. Hopkins, R. G. Barr, G. F.
Michel & P. Rochat (Eds) The
Cambridge encyclopedia of child development (Cambridge,
Cambridge University Press), 487–490.
Stuart, M., Stainthorp, R. & Snowling, M. (2008) Literacy as a
complex activity: deconstructing

the simple view of reading, Literacy, 42, 59–66.
Thomas, J. B. (1996) The beginnings of educational psychology
in the universities of England and
Wales, Educational Psychology, 16, 229–244.
Wolf, M. (2008) Proust and the squid: the story and science of
the reading brain (Cambridge, Icon
Books).
Wyse, D. & Goswami, U. (2008) Synthetic phonics and the
teaching of reading, British Educational
Research Journal, 34, 691–710.
Database Systems Performance Evaluation
Techniques
Subharthi Paul [email protected] (A project report written under
the guidance of Prof. Raj Jain) Download
Abstract
The last few decades has seen a huge transformation in the way
businesses are conducted. There has been a
paradigm shift from product portfolio based marketing
strategies to customer focused marketing strategies.
The growth and diversity of the market has greatly profited
consumers through higher availability, better
quality and lower prices. The same factors however has made it
more difficult for businesses to maintain their
competitive edge over one another and hence has forced them to

think beyond their product portfolio and
look at other means to gain higher visibility and customer
satisfaction, maintaining all the while their core
advantages on pricing and product through improved and more
efficient methods of manufacturing and
distribution. The advent and spread of computers and
networking has been one of the single largest factors
that has spurred and aided this enormous movement. More
specifically, database management systems now
form the core of almost all enterprise logic and business
intelligence solutions. This survey tries to emphasize
the importance of database systems in enterprise setups and
looks at the methods and metrics that are used to
evaluate the performance of these database systems.
Keywords
Database Systems, Transaction Processing, Performance
Evaluation Techniques, Database Benchmarking
Table of Contents
1. Introduction
2. Database Management Systems: Formal Definition and
Classification
2.1 Classification based on Data Modeling
2.1.1 Hierarchical Model
2.1.2 Relational Model
2.1.3 Network Model
2.1.4 Object-Relational Model
3. A General Approach to Database Performance Evaluation
4. Database Performance Evaluation Techniques for specialized
Databases
5. Database Systems: Performance Evaluation Benchmarks

5.1 TPC Benchmarks
5.1.1 TPC-C Benchmark
5.1.2 TPC-E Benchmark
5.1.3 TPC-H Benchmark
5.2 Other Benchmarks
Database Systems Performance Evaluation Techniques
1 of 12
5.2.1 Bristlecone
5.2.2 Benchmark Factory
5.2.3 CIS Benchmark
5.2.4 SPEC Benchmark
5.2.5 PolePosition
5.2.6 Open Source Development Labs Database Test Suite
6. Summary
References
1. Introduction
The last few decades has seen a huge transformation in the way
businesses are conducted. There has been a
paradigm shift from product portfolio based marketing
strategies to customer focused marketing strategies.
The growth and diversity of the market has greatly profited
consumers through higher availability, better
quality and lower prices. The same factors however has made it
more difficult for businesses to maintain their
competitive edge over one another and hence has forced them to
think beyond their product portfolio and

look at other means to gain higher visibility and customer
satisfaction, maintaining all the while their core
advantages on pricing and product through improved and more
efficient methods of manufacturing and
distribution. The advent and spread of computers and
networking has been one of the single largest factors
that has spurred and aided this enormous movement. More
specifically, database management systems now
form the core of almost all enterprise logic and business
intelligence solutions.
Database Systems are one of the key enabling forces behind
business transformations. Apart from supporting
enterprise logic they also enable business intelligence.
Information is the key to success in today's businesses.
However, maintaining information in logically consistent and
feasibly retrievable format is a daunting task.
More so with the added complications of transaction
consistency management, synchronization across
multiple repositories spread geographically across the globe,
failover management and redundancy
management, today's database systems are truly state-of-the-art
high performance software systems.
Apart from managing a plethora of complicated tasks, database
management systems also need to be efficient
in terms of storage and speed. Businesses have a tendency to
store un-required historical data often as a result
of poor data planning or less frequently owing to federal
obligations or consumer law. Dynamic addition and
deletion of data from the database also pose a challenge to
maintaining an efficient data retrieval mechanism.
Though, limited in speed by some sense due to hardware
limitations, database systems nonetheless need to
achieve full throttle through efficient storage and retrieval
techniques. Another factor that often hurt database

performance is ill-written computationally expensive queries.
Often, such situations are beyond control of the
database system and require external performance tuning by
experts.
As is true for most systems, reliability, availability and fault-
tolerance is a huge concern for database systems.
Reliability of a system is generally improved through
redundancy. Modern businesses cannot afford to loose
data or present wrong data. Modern business activities are
highly centered around and dependant on
electronic data. Modern database systems thus need to build in
high reliability mechanisms in their designs.
Availability is another issue that concerns a lot of businesses.
As an example, "Netflix" an online DVD rental
business receives online requests for 1.9 million DVD's
everyday. An outage for 10 minutes cost huge losses
to these businesses. Similar is the case for a lot of other
businesses. Amazon.com stands to loose
approximately $31000 per minute for a global outage [1].
Clearly, availability is a key metric for measuring
the performance of database systems.
Another metric, which is more qualitative in nature, but
extremely important is security. It is generally
difficult to measure the level of security of any system unless
its security vulnerabilities are exposed.
2 of 12
Database systems often store sensitive enterprise and customer
data that if inappropriately used may cause

huge monetary and business losses for the organization. Hence,
though difficult to quantify through standard
testing procedures, database systems strive for continual
upgrade of their security features.
Performance evaluation of database systems is thus an important
concern. However, easier said than done,
performance evaluation of database system is a non-trivial
activity, made more complicated by the existence
of different flavors of database systems fine tuned for serving
specific requirements. However performance
analysts try to identify certain key aspects generally desired of
all database systems and try to define
benchmarks for them. In the rest of this survey, we shall
provide a formal definition of database systems
followed by a few methods to categorize or classify database
systems. This shall be followed by a look at the
various performance evaluation techniques that are employed to
benchmark database systems, some of the
key benchmarking techniques used in practice in the industry
and some open source benchmarking schemes
available for use in the public domain.
2. Database Management Systems: Formal Definition and
Classification
A Database Management System (DBMS) is a complex set of
software programs that controls the
organization, storage, management, and retrieval of data in a
database. DBMS' are categorized according to
their data structures or types. It is a set of prewritten programs
that are used to store, update and retrieve a
Database [2].
Figure 1 - A generic high level view of a Database Management
System

Figure 1 presents the generic high level view of a database
management system. The "Database Management
System is a collection of software programs that allow multiple
users to access, create, update and retrieve
data from and to the database and the "Databae" is a shared
resource that is at the centre of such a system.
The database's functionality is optimal storage and retrieval of
data, maintain correctness of the data, and
maintaining consistency of the system at all times.
3 of 12
There are two basic criterions on which database systems are
generally classified. One of them is based on
Data Modeling while the other is Functional Categorization.
Here, we briefly discuss each of these
classification techniques. A few functional categorization shall
be discussed in the next section when we
discuss "Database Performance Evaluation Techniques for
specialized Databases" in Section
2.1 Classification based on Data Modeling
2.1.1 Hierarchical Model:
In a hierarchical model data is represented as having a parent-
child relationship among each other and is
organized in a tree-like structure. The organization of data
enforces a structure wherein a parent can have
many children but a child can have only one parent. Thus, this
model inherently forces repetitions of data at

the child levels. The records have 1:N or more generally a "one-
to-many" relationship between them. This
was one of the first data base models to be used and
implemented in IBM's Information Management
System(IMS) [3] . Hierarchical model was the most intuitive
way to represent real-world data but was not the
most optimal one.This model was later replaced by a more
efficient and optimal model called the "Relational
Model" that we shall discuss next.
2.1.2 Relational Model:
The Relational Model [4] was formulated by Edgar Codd and is
one of the most influential model that has
governed the implementation of some of the best known
Database systems. The model is based on "first order
predicate logic". In a representation such as A = {x| P(x) }, P(x)
is the predicate that makes a descriptive
statement about the elements of set A, such as "All positive
integers less than 1000" results in a set A = {
1,..,999}. So, a predicate maps an "entity" to a "truth table". In
the above example, 'x' is an entity and P(x) is
a mapping which determines whether 'x' belongs to set A or not.
Now, suppose that we have a huge domain
and we need to make some statements that apply selectively to
some or all members in the domain. The
formal language that allows us to make such a statement is
called "first order logic". In first order logic
statements, a predicate either takes the role of a defining the
"property" of an entity or the "relationship
between entities". Further and in-depth description on the
mathematical foundations of the relational model is
beyond the scope of the present work.
So, in the relational model, data is represented using a set of
predicates over a finite set of variables that

model the belongingness of certain values to a certain sets and
the constraints that apply on them. The
relational model, owing to its strong foundations in
mathematical formal logic is extremely powerful in
representation and at the same time efficient in terms of storage
requirements (by removing redundancies due
to repeated data) and also speed of retrieval/storage.
2.1.3 Network Model:
The network model [5] can be seen as a generalization of the
hierarchical model. In this model, each data
object can have multiple parents and each parent object can
have multiple children. This the network model
forms a "lattice-type" structure in contrast to the "tree-like
structure" of the hierarchical model. The network
model represents real-world data relationship more naturally
and under less constrained environment than the
hierarchical model.
2.1.4 Object-Relational Model:
The Object-Relational Model [6] is similar to the relational
model except for additions of object-oriented
concepts in modeling. The data modeling is done using
relational concepts while the object-oriented concepts
4 of 12
facilitate complex modeling of the data and its relationship with
the methods of manipulation/retrieval and
storage. This is one of the newer and more powerful models of

all the models discussed in this section.
There are lots of other models classifying databases based on
the object-models such as Semi-structure
Model, Associative Model, Entity-Attribute-Value Model etc.,
but a detailed study of each of these is beyond
the present scope. The aim of the above classification was to
attain a minimal understanding that would help
us appreciate the design of performance analysis methods for
data bases.
3. A General Approach to Database Performance Evaluation
In this section we discuss some of the more "general" methods
that can be used for database performance
evaluation. The word "general" is binding to systems, meaning
that the approaches mentioned here are
generally true for "systems" with a special focus on database
systems.
According to [7], performance analysis of database systems
serve two basic purposes:
1. For the evaluation of the best configuration and operating
environment of a single database system, and
2. Studying two or more database systems and providing a
systematic comparision of the systems.
Accordingly, some of the analytical modeling methods for
evaluating systems that are applicable for database
systems too are:
1. Queuing Models: Queuing models are effective to study the
dynamics of a database system when it is
modeled as a multi-component system with resource allocation

constraints and jobs moving around from one
component to another. Examples of such dynamic studies are
concurrent transaction control algorithms, data
allocation and management in distributed database systems etc.
2. Cost Models: Cost Models are useful in studying the cost in
terms of Physical storage and query processing
time. The cost model gives some real insight into the actual
physical structure and performance of a database
system.
3. Simulation Modeling: A simulation Modeling is more
effective for obtaining better estimates since it not
only analyses the database system in isolation but can
effectively analyze the database system with the
application program running on top of it and the database
system itself operating within the constrained
environment of an operating system on a real physical hardware.
4. Benchmarking: Benchmarking is the best method when
multiple database systems need to be evaluated
against each other but suffer from the inherent setback that it
assumes all systems to be fully installed and
operational. Benchmarking relies on the effectiveness of the
synthetic workloads. Real workloads are non
repeatable and hence not good for effective benchmarking.
[7] presents a 3 step process for benchmark evaluation
a. Benchmark design
b. Benchmark Execution
c. Benchmark analysis
Our study makes a delves deeper into the various database

system evaluation benchmarks and we leave the
5 of 12
more detailed analysis of Database systems benchmark studies
to a later section where we see different
real-life benchmarking techniques.
[8] proposes a set of methods for database performance
evaluation assuming the system to be operating in a
multi-user environment. Accordingly the three factors that
effect the performance of a database system in a
multi-user environment are:
1. Multi-programming level
2. Query Mix
3. Extent of data Sharing
Data sharing is the condition of concurrent access of a data
object by multiple processes. The interesting
factor here is that of the query mix. A proper query mix needs
to test the appropriate levels of CPU and disk
utilization required to serve a particular query. The query mix
needs to properly represent a true multi-user
environment. Also, the query mix may be designed to represent
a certain percentage of data sharing. Once
these have been figured out, the query-mix forms a
representative benchmark program and multiple copies of
the bench-mark program are issued concurrently to simulated
multi-programming effects. Also, different

query-mixes allow diversity in the experimental design
conditions. The response variable studied is system
throughput and response time.
Summarizing, in this section, the focus was primarily to study
the performance evaluation techniques
considering the "general system criterion" of a database system.
In the next section we look at the
performance evaluation techniques more specialized for
particular database system types.
4. Database Performance Evaluation Techniques for specialized
Databases
In the last section we discussed about a few performance
evaluation techniques that are extremely general
and apply to almost all database systems and as such to most
generic systems. In this section, we refine our
discussion to the techniques developed specially for the
performance evaluation of databases of various types
or in special application specific areas.
4.1 Real-time Database Systems performance Evaluation
Simulation studies on real time database systems are limited by
the unavailability of realistic workloads and
hence fail to test the system in real dynamic situations where
they have to operate. Real-time database
systems are generally mission-critical and has high QoS
requirements. Proper evaluation techniques for the
performance real-time systems is thus extremely important and
at the same time extremely challenging. [10]
develops "Chronos", a real-time database test-bed to evaluate
performance of such systems.
Unlike non-real time test-beds, chronos does not evaluate the

average throughput and average response time.
These are not the representative metrics to test real-time
database systems which have added constraints of
"data freshness" , "deadlines" and "time bound". Chronos
provides methods to vary transactions and
workloads applied to a real-time database system and studies
the timeliness and freshness of data metrics
under multiple load conditions.
Chronos also implements a QoS management scheme which
incorporates overload detection, admission
control and adaptive policy update. As mentioned previously,
QoS requirements of real time databases are
6 of 12
generally high and so it is extremely important to be able to
quantify the system state based on such QoS
parameters. A good real-time system should be one that can
correctly evaluate its state and provide QoS
guarantees in terms of timeliness, transaction time bounds and
delays.
4.2 Web- Database Systems performance Evaluation
Web databases serve the back-ends of Web-Servers. Web
databases are the classic examples of high load,
high performance database systems. The most common
performance metrics for susch database systems are
average response time and throughput, fault tolerance,
scalability etc. Availability metrics such as
MTTF(Mean-time to failure), MTTR( Mean time to repair) etc

are also extremely important metrics for such
systems. The challenges in performance evaluation of web
databases is the proper workload selection
representative of the typical load as well as highly loaded
situations. [11] proposes a workload that consists of
a query mix as shown in Table 1. The objective behind the
query design is to provide fundamental queries
that can be elementarily grouped to provide different load and
resource utilization condidtions.
Class Description
Load on DB
Servers
Load Web
Servers Load on
Network
CPU DISK CPU DISK
1 Complex query with small resultsize High Low Low N/A Low
2 Complex query with large resultsize High High High N/A
High
3 Simple query with small result size Low Low Low N/A Low
4 Simple query with large result size Low High High N/A High
5 No query, small file size N/A N/A Low Low Low
6 No query, large file size N/A N/A High High High
Table 1: Workload classification of web database systems [11]

Another architectural component typical of web-database
systems are various input and output queues.
Transactions pile up at both these queues based on the arrival
rate and response time of the database.
Modeling a simulation or test bench for a web-database system
requires considering this architectural element
for proper representation of the system in an operational
environment.
4.3 Enterprise Data Mining System Performance Evaluation
Enterprise database systems form the heart of enterprise
operations as well as enterprise intelligence. They
are huge database systems (often called :data mining systems")
often storing historical and redundant data.
They are the basis of all business intelligence processes and
decision support systems. Efficient storage and
retrieval from suce huge databases ("data-warehouses") require
special techniques of aggregation and
classification of data while storing. The process of enterprise
decision making involves high level canned
queries by managers, mapping of these high level queries into
specific predicate rules and retrieval of the
datasets that match those rules. Also, often the result of a query
involves data retrieval from multiple
distributed databases. The performance evaluation of such
systems entail mostly the performance evaluation
of data mining algorithms. [13] proposes a test bed for the
performance evaluation of frequent pattern
matching algorithms. It incorporates a synthetic data generator
that can generate all kinds of data following a
7 of 12

wide variety of frequency distribution patterns, as set of
classical pattern matching techniques and a
performance database that records the performance of the
algorithm on each generated data sets. The main
idea is to test the data mining system with reference to specific
algorithms under stress conditions of randomly
generated data.
4.4 Object Oriented Systems Performance Evaluation
[14] presents a comprehensive study on the comparison of
object oriented database systems. The standard
benchmarks to evaluate object oriented database systems are
OO1, OO7, HyperModel, and ACOB. We shall
discuss more about these benchmarks in later section. For the
current discussion we try to isolate the issues
specific to OODBMS's performance evaluation and the
techniques employed to study them. The chief factors
governing the performance of such database systems are
evaluated using object operations. The object
oriented database schema consists of different types of
objects(simple and complex) and the
relationships(direct, hierarchical etc) between them. The
evaluation technique for such object oriented
schema involves the development of operations on these objects
such as dense traversal, sparse traversal,
string search, join etc. Also other system level characteristics
such as number of users, load etc are varied just
like techniques for evaluating other database systems. The main
idea is to generate queries on object oriented
schema to test the effectiveness and design optimization of the
object design and then testing for system level
parameters at the same time.

The specific techniques discussed for each type of database
systems in this section warrant a deeper study
into each of these evaluation methods and also just presents a
preliminary survey to the great diversity that is
hidden behind the seemingly simple task of storing and
retrieving data.
5. Database Systems: Performance Evaluation Benchmarks
Having established the importance of database system in the last
few sections, it is needless to say that there
is a plethora of commercial benchmarks to evaluate database
systems. Such benchmarks are extremely
important since a lot of business activities in the real world
dwell around database systems. Also, such
benchmarks are important pointers towards the selection of the
right package and the right configuration for a
particular operating environment. The most famous are the TPC
(Transaction Processing Performance
Council) family of benchmarks[15] . Another set of benchmarks
evaluating security in database systems are
CIS Benchmarks for SQL Server presented by Center for
Internet Security [17].
Benchmarking is generally referred to the process of evaluating
a system against some reference to determine
the relative performance of the system. As already discussed in
the paper, though the basic primitive job of
the database system remains generic, yet database systems
exhibit different flavors and requirements based
on the environment in which they operate. Also, database
systems contribute hugely to the proper and
efficient functioning of organizational and business information
needs. Hence, selecting the right database
with the right features is often a very critical decision. To aid
such decisions, a bunch of bench-marking

techniques have been designed, both commercially and in the
open source platform. In this section we shall
look at some of these benchmarking techniques in more detail.
5.1 TPC Benchmarks
TPC or "Transaction Processing Performance Council" has
defined a suite of benchmarks for transaction
processing and database performance analysis. These
benchmarks do not solely evaluate the database
component of the system but rather evaluates the whole system
of which the Database system is one of the
key differentiating factors. The suite contains a mix of
benchmarks for evaluating different performance
8 of 12
aspects of such systems.
5.1.1 TPC-C Benchmark [19]
The TPC-C benchmark is an On-line Transaction Processing
(OLTP) Benchmark. TPC-C benchmark
contains a mix of different types of concurrent transactions, a
complex database, nine types of tables with
different record and population sizes. It simulates the process of
a multi-user environment making concurrent
queries to a central database. The performance evaluation
involves a thorough monitoring of all system state
parameters for correctness of update as well as performance
parameters such as service time etc. This
benchmark is most suitable for businesses that need database to

support online handling of orders, sell
product and manage inventory.
5.1.2 TPC-E Benchmark [20]
The TPC-E benchmark is another OLTP Benchmark but is
especially designed for evaluating database
systems needed to be installed at brokerage firms. It is quite
similar to the TPC-C benchmark in terms of
setup and components differing only in the design of the
transactions which are more relevant in a brokerage
firm environment such as account inquiries, online trading and
market research etc.
5.1.3 TPC-H Benchmark [21]
The TPC-H benchmark is fine tuned for decision support
systems. The transactions in such environments are
characterized by business intelligence intensive complex data
mining queries and concurrent data
modifications. The performance metric used to evaluate such
systems is generally TPC-H composite query
per hour.
5.2 Other Benchmarks
TPC is the largest and most popular benchmarking authority in
databases and hence was dedicated a separate
sub-section. The other benchmarking tools that are available are
discussed in this sub-section.
5.2.1 Bristlecone [22]
Bristlecone is a java based database performance testing
benchmarking utility. It provides knobs to vary the
system parameters for a single scenario across different set of

rules or environments. The standard run
parameters for synthetic replication include number of users,
number of tables, number of rows per table,
number of rows returned by queries or size and complexity of
queries etc. It is supposed to be lightweight and
easily installable tool that can be used to serve as a load
generation and performance evaluation tool against
any database server.
5.2.2 Benchmark Factory [23]
Benchmark factory is a testing and performance evaluation tools
for databases that can be used
pre-deployment to gauge the effectiveness of the particular
database system in meeting an organizations IT
goals and needs. It simulates multi-user transaction environment
on a database in a production environment or
a synthetic workload in a non-production environment.
Benchmark factory is thus meant to be a tool for cost
effective pre-validation and pre-verification of a database
deployment in a particular operating environment.
5.2.3 CIS Benchmark [17]
9 of 12
The CIS benchmark is a set of security benchmark for the SQL
Server 2005 and SQL Server 2000. These
benchmarks provide a testing tool to evaluate these database
systems against common security vulnerabilities.
Generally while installing databases most administrators focus
on key operating performance issues such as

scalability, load balancing, failovers, availability etc and let
security settings to be default factory settings.
Default security settings are prone to a number of security
hazards and generally not advisable. Also, even if
the security settings are changed from default, they may not
fulfill the security requirement expected. The
CIS benchmark test database system against security
configuration errors in a non-intrusive environment that
does not in any ways jeopardize the normal functioning of the
database.
5.2.4 SPEC Benchmark [16]
SPEC (Standard Performance Evaluation Corporation) has a
benchmark to evaluate Web Servers and is
called SPECweb2005. It simulates multiple users sending
tightly secure (https), moderately secure (http/https)
and unsecure (http) requests to a web server. The dynamic
content is the web server is implemented in PHP
or JSP.
The SPECweb2005 is well representative of actual web-server
scenarios since it simulates multi-user dynamic
page request scenario over a mix of application layer protocols.
Also, it simulates web-browser caching
effects which is one of the chief optimization technique applied
by all such systems.
Though not exactly a database benchmark, SPECweb2005
deserved a special mention because database
systems serve the heart of all web server systems and hence one
of the keys to the performance bottleneck of
such systems.
5.2.5 PolePosition [24]

PolePosition is an open source database benchmark. It consists
of a test-suite that compares the database
engine and the object-relational mapping technology. It has a
java based framework. Keeping up with the
spirit of its name, each database test suite is called a "circuit"
exercising a mix of different tests. Each of these
circuits are intended to test a given database against a pre-
canned set of criteria such as heavy-reads,
frequent-updates etc. An administrator wanting to choose a
database implementation for his requirements
may try and map his requirement to a particular "circuit" and
test databases in that particular circuit to get the
"PolePosition" of databases relative to that "circuit".
5.2.6 Open Source Development Labs Database Test Suite
The OSDL DBT is also an open source database benchmarking
tool. It is one of the most comprehensive
open-source database benchmarking test-suite available in open
source and it is based on the standards set by
TPC. However, they differ in a few areas and hence it is not
advisable to compare results from tests in this
suite with standard TPC tests.
In this section, we have tried to discuss the most commonly
used benchmarks that are actually used in the
real-world. However, there are lots of other benchmarking
suites available for database systems, but a
discussion on all of those are beyond the scope of the current
work An interested reader may follow-up
his/her reading and may be directed to [26] for further pointers.
6. Summary
In this paper, we have tried to highlight the importance of
Database systems in this IT enabled world, the

10 of 12
various flavors of database applications and production
environments, the different requirements and
configurations in these environments and lastly, pointers to be
able to evaluate databases in accordance with
its area of application. This paper is expected to be a
preliminary reading exercise and hopefully an interested
reader would have gain enough understanding from here to be
able to follow up with a more specific are of
interest.
References
[1] Amazon suffers U.S. outage on Friday, CNET News
[2] Wikipedia: Database Management Systems
[3] Rick Long, Mark Harrington, Robert Hain, Geoff Nicholls,
IMS Primer, International Technical Support
Organization, IBM Corporation, 2000
[4] Codd, E.F. , "A Relational Model of Data for Large Shared
Data Banks". Communications of the ACM 13
(6): 377-387, 1970
[5] Charles W. Bachman, The Programmer as Navigator. ACM
Turing Award lecture, Communications of the
ACM, Volume 16, Issue 11, 1973, pp. 653-658
[6] Stonebraker, Michael with Moore, Dorothy. Object-

Relational DBMSs: The Next Great Wave. Morgan
Kaufmann Publishers, 1996.
[7] Yao-S Bing, Alan R. Hevner, A Guide to Performance
Evaluation of Database Systems, Superintendent of
Documents, U.S. Government Printing Office, Washington, DC
20402,1984
[8] Haran Boral, David J DeWitt, A methodology for database
system performance evaluation, ACM
SIGMOD Record, Volume 14, Issue 2, June 1984
[9] S. H. Son and N. Haghighi, Performance evaluation of
multiversion database systems, Sixth International
Conference on Data Engineering, Issue 5-9, Pages 129-136,
February 1990
[10] Kyoung-Don Kang and Phillip H. Sin and Jisu Oh, A Real-
Time Database Testbed and Performance
Evaluation, Proceedings of the 13th IEEE International
Conference on Embedded and Real-Time Computing
Systems and Applications, Pages 319-326 , 2007
[11] Yuanling Zhu and Kevin Lu, Performance analysis of Web
Database systems, LNCS, Volume1873/2000,
pp 805-814, 2000
[12] Peter G. Harrison and Catalina M. Llado, Performance
Evaluation of a Distributed Enterprise Data
Mining System, Vol 1786/2000,pp 117-131, 2000
[13] Jian Pei and Runying Mao and Kan Hu and Hua
Zhu,Towards data mining benchmarking: a test bed for
performance study of frequent pattern mining, SIGMOD '00:
Proceedings of the 2000 ACM SIGMOD
international conference on Management of data, pp 592,2000

[14] Jia-Lang Seng, Comparing Object-Oriented Database
Systems Benchmark Methods, Proceedings of the
Thirty-First Annual Hawaii International Conference on System
Sciences, Volume 6, Page 455, 1998
[15] TPC benchmarks
11 of 12
[16] SPEC Benchmarks for Java Client Server applications
[17] The Centre for Internet Security, CIS Benchmarks for SQL
Server 2005 and SQL Server 2000
Databases, 2005
[18] TPC benchmarks
[19] TPC-C benchmark
[20] TPC-E Benchmark
[21] TPC-H benchmark
[22] Bristlecone Benchmark, 2008
[23] Qwest Software, Benchmark Factory for Databases
[24] PolePosition
[25] Open Source Development Labs Database Test Suite

[26] Database Benchmarking,
http://wiki.oracle.com/page/Database+Benchmarking
Last modified on November 24, 2008
This and other papers on latest advances in performance
analysis are available on line at
http://www.cse.wustl.edu/~jain/cse567-08/index.html
Back to Raj Jain's Home Page
12 of 12
Yang, T.-C., Hwang, G.-J., & Yang, S. J.-H. (2013).
Development of an adaptive learning system with multiple
perspectives
based on students' learning styles and cognitive styles.
Educational Technology & Society, 16 (4), 185–200.
Development of an Adaptive Learning System with Multiple
Perspectives
based on Students’ Learning Styles and Cognitive Styles
Tzu-Chi Yang1, Gwo-Jen Hwang2 and Stephen Jen-Hwa
Yang1*
1Department of Computer Science and Information Engineering,
National Central University, No. 300, Jung-da Road,
Chung-li, Taoyuan 320, Taiwan // 2Graduate Institute of Digital
Learning and Education, National Taiwan University

of Science and Technology, 43, Sec.4, Keelung Rd., Taipei,
106, Taiwan // [email protected] //
[email protected] // [email protected]
*Corresponding author
(Submitted June 10, 2012; Revised November 13, 2012;
Accepted December 11, 2012)
ABSTRACT
In this study, an adaptive learning system is developed by
taking multiple dimensions of personalized features
into account. A personalized presentation module is proposed
for developing adaptive learning systems based on
the field dependent/independent cognitive style model and the
eight dimensions of Felder-Silverman's learning
style. An experiment has been conducted to evaluate the
performance of the proposed approach in a computer
science course. Fifty-four participants were randomly assigned
to an experimental group which learned with an
adaptive learning system developed based on the personalized
presentation module, and a control group which
learned with the conventional learning system without
personalized presentation. The experimental results
showed that the experimental group students revealed
significantly better learning achievements than the control
group students, implying that the proposed approach is able to
assist the students in improving their learning
performance.
Keywords
Adaptive learning, Personalization, Learning style, Cognitive
style

Introduction
The rapid advancement of computer and network technologies
has attracted researchers to develop tools and
strategies for conducting computer-assisted learning activities
(Hwang, Wu, & Chen, 2012; Tsai, 2004). With these
new technologies, learning content becomes rich and diverse
owing to the use of hypermedia and multimedia
presentations. Researchers have indicated that hypermedia
systems are suitable for providing personalized learning
supports or guidance by identifying the personal characteristics
of students and adapting the presentation styles or
learning paths accordingly (Tseng, Chu, Hwang, & Tsai, 2008).
In the past decade, various personalization
techniques have been proposed for developing adaptive
hypermedia learning systems, and have demonstrated the
benefit of such an approach (Mampadi, Chen, Ghinea, & Chen,
2011; Nielsen, Heffernan, Lin, & Yu, 2010; Wells &
McCrory, 2011). For example, Papanikolaou, Grigoriadou,
Magoulas and Kornilakis (2002) developed an adaptive
learning system by taking students’ knowledge levels as the
main factor for adapting the learning content; moreover,
Tseng, Su, Hwang, Hwang, Tsai and Tsai (2008) developed an
adaptive learning system based on an object-oriented
framewok that composes personalized learning content by
considering individuals' knowledge level and the difficulty
level of the learning objects.
Although the knowledge level of the students and the difficulty
level of the learning content are good factors for
adapting presentation layouts and selecting appropriate learning
content for individuals, researchers have indicated
the importance of taking personal preferences and learning
habits into account (Hsu, Hwang, & Chang, 2010; Tseng,
Chu, Hwang, & Tsai, 2008). Among those personal

characteristics, learning styles which represent the way
individuals perceive and process information have been
recognized as being an important factor related to the
presentation of learning materials (Papanikolaou, Grigoriadou,
Magoulas, & Kornilakis, 2002). On the other hand,
cognitive styles have been recognized as being an essential
characteristic of individuals' cognitive process. In the past
decade, researchers have tried to develop adaptive learning
systems based on either learning styles or cognitive
styles; nevertheless, seldom have both of them been taken into
consideration, not to mention the other personalized
factors (Hsieh, Jang, Hwang, & Chen, 2011; Hwang, Tsai, Tsai,
& Tseng, 2008; Mampadi, Chen, Ghinea, & Chen,
2011).
185 ISSN 1436-4522 (online) and 1176-3647 (print). ©
International Forum of Educational Technology & Society
(IFETS). The authors and the forum jointly retain the
copyright of the articles. Permission to make digital or hard
copies of part or all of this work for personal or classroom use
is granted without fee provided that copies
are not made or distributed for profit or commercial advantage
and that copies bear the full citation on the first page.
Copyrights for components of this work owned by
others than IFETS must be honoured. Abstracting with credit is
permitted. To copy otherwise, to republish, to post on servers,
or to redistribute to lists, requires prior
specific permission and/or a fee. Request permissions from the
editors at [email protected]
Researchers have indicated the importance of taking multiple
personalization factors into account in order to deliver
effective learning systems to individual students (Lee, Cheng,

Rai, & Depickere, 2005; Hwang, Yin, Wang, Tseng,
& Hwang, 2008). To cope with this problem, in this study, an
adaptive learning system is developed by taking
students’ preferences and characteristics, including learning
styles and cognitive styles, into consideration.
Moreover, an experiment has been conducted to show the
effectiveness of the proposed approach.
Literature review
Learning styles and cognitive styles
Learning styles have been recognized as being an important
factor for better understanding the model of learning and
the learning dispositions/preferences of students (Filippidis &
Tsoukalas, 2009). Keefe (1987) defined an
individual’s learning style as a consistent way of functioning
that reflects the underlying causes of learning
behaviors. Keefe (1991) pointed out that learning style is a
student characteristic indicating how a student learns and
likes to learn. He also stated that learning style could be an
instructional strategy informing the cognition, context
and content of learning. Reiff (1992) indicated that learning
styles are likely to influence how students learn, how
instructors teach, and how they interact. Coffield, Moseley, Hall
and Ecclestone (2004) further suggested that
teachers and course designers pay attention to students’ learning
styles and design teaching and learning
interventions accordingly.
There have been several learning style theories proposed by
researchers, such as those proposed by Honey and
Mumford (1992), Keefe (1979), Kolb (1984) and Felder and
Silverman (1988). Several previous studies have
demonstrated the use of learning styles as one of the parameters

for providing personalized learning guidance or
contents (Graf, Lin, & Kinshuk, 2007; Papanikolaou, Mabbott,
Bull, & Grigoriadou, 2006; Tseng, Chu, Hwang, &
Tsai, 2008). Among various learning styles, the Felder–
Silverman Learning Style Model (FSLSM) developed by
Felder and Soloman (1997) have been recognized by many
researchers as being a highly suitable model for
developing adaptive learning systems (Huang, Lin, & Huang,
2012; Akbulut & Cardk, 2012). Carver, Howard and
Lane (1999) indicated that FSLSM could be the most
appropriate measurement for developing hypermedia
courseware by taking into personal factors into account. Kuljis
and Lui (2005) further compared several learning
style models, and suggested that FSLSM is the most appropriate
model with respect to the application in e-learning
systems. Consequently, this study adopted FSLSM as one of the
factors for developing the adaptive learning system.
On the other hand, cognitive style has been recognized as being
a significant factor influencing students’ information
seeking and processing (Frias-Martinez, Chen, & Liu, 2008). It
has also been identified as an important factor
impacting the effectiveness of user interfaces and the navigation
strategies of learning systems (Mampadi, Chen,
Ghinea, & Chen, 2011). Several studies have shown the
effectiveness of considering cognitive styles in designing
user interfaces for information seeking (Frias-Martinez, Chen,
& Liu, 2008) and developing adaptive learning
systems for providing personalized learning guidance (Evans, &
Waring, 2011; Lo, Chan, & Yeh, 2012). Among
various proposed cognitive styles, the field dependent (FD) and
field independent (FI) styles proposed by Witkin,
Moore, Goodenough and Cox (1977) are the most frequently
adopted. Several studies have reported the usefulness of
FI/FD cognitive styles in determining the suitability of learning
supports or learning system designs (Gerjets,
Scheiter, Opfermann, Hesse, & Eysink, 2009; Lin, Hwang, &

Kuo, 2009). For example, Weller, Repman and Rooze
(1995) indicated that FI/FD cognitive style is very suitable for
personalized learning design since it reveals how well
a learner is able to restructure information based on the use of
salient cues and field arrangement. Ford and Chen
(2000) further indicated that the FD/FI cognitive style is highly
related to hypermedia navigation and is very suitable
for evaluating the usability of websites to students. Therefore,
in this study, FI/FD cognitive style is adopted as
another factor for developing the adaptive learning system.
Scholars have proposed different aspects to address the
relationships between learning styles and cognitive styles.
For example, some scholars have indicated that learning styles
are applied cognitive styles (Keefe, 1979; Jonassen &
Grabowski, 1993; Papanikolaou, Mabbott, Bull, & Grigoriadou,
2006); some have further concluded that learning
styles could be viewed as a subset of cognitive styles, and could
be classified as activity-centered cognitive styles
(Huang, Lin, & Huang, 2011). However, the common definition
of cognitive style refers to the individual differences
in preferred ways of organizing and processing information and
experience (Chen & Macredie, 2002; Triantafillou,
Pomportsis, & Demetriadis, 2003), while learning style is
defined as a consistent way of functioning that reflects the
underlying causes of learning behaviors (Keefe, 1987).
Moreover, cognitive styles deal with a cognitive activity (i.e.,
186
thinking, perceiving, remembering), while learning styles are
indicators of how learners perceive, interact with and
respond to learning environments, including cognitive, affective
and psychological behaviors (Triantafillou,

Pomportsis, & Demetriadis, 2003).
To deal with the relationship between cognitive and learning
styles, researchers have indicated that cognitive styles
could be classified as cognition centered, personality centered,
or activity centered; moreover, learning style can be
perceived as the activity-centered cognitive style (Sternberg &
Grigorenko, 1997). From this aspect, learning styles
are viewed as a subset of cognitive styles (Riding & Rayner,
1998; Sternberg & Grigorenko, 1997). Accordingly, this
study employs cognitive styles in dealing with the adaption of
the learning environment, such as the navigation
modes, whereas learning styles are used to deal with the
presentation modes of multi-source materials that are
composed of figures, videos and texts.
Accordingly, in this study, learning styles are used to provide
personalized learning materials and presentation
layouts (Liegle & Janicki, 2006), while cognitive styles are used
to develop personalized user interfaces and
navigation strategies (Chen, Fan, & Maredie, 2004; Chen &
Macredie, 2002; Gerjets, Scheiter, Opfermann, Hesse, &
Eysink, 2009).
Cognitive load and multimedia learning
To understand, accommodate and align the interaction between
learners’ cognitive system and the given learning
environment, the cognitive load theory (CLT) has become an
acknowledged and broadly applied theory for
instruction and learning (Van Merriënboer & Sweller, 2005;
Schnotz & Kürschner, 2007). Cognitive load theory is a
framework of instructional design principles based on the
characteristics and relations between the structures that
constitute human cognitive architecture, particularly working

memory and long-term memory (Wong, Leahy,
Marcus, & Sweller, 2012). For a multimedia instructional
design, CLT responses the limited working memory for
holding visual (such as figures) and verbal (such as text)
information as well as the number of operations it can
perform on the information (Van Gerven & Pascal, 2003).
Cognitive load is defined as a multidimensional construct
representing the load that a particular task imposes on the
performer (Paas & van Merrienboer, 1994). It can be assessed
by measuring mental load, mental effort (Sweller, van
Merriënboer, & Paas 1998; Paas, Tuovinen, Tabbers, & Gerven,
2003). Mental effort is related to the strategies used
in the learning activities, whereas mental load refers to the
interactions between the learning tasks, subject
characteristics and subject materials, which are highly related to
the complexity of the learning content that the
students need to face (Hwang & Chang, 2011). To respond to
the reality that most digital learning materials are
developed with multimedia, Mayer (2001) proposed a cognitive
theory of multimedia learning (CLML), which
assumes that human process pictorial and verbal materials via
different sense channels (i.e., sight and hearing).
Consequently, cognitive overloading could occur when learners
receive redundant information, poorly structured
information, or large amount of information in a sense channel.
On the other hand, Paas, Tuovinen, Merriënboer and Darabi
(2005) addressed that learners’ motivation had a
significant relation with cognitive load, especially on mental
effort. They suggested that motivation could be
identified as a dimension that determines learning success,
especially in complex e-learning environments (Paas,
Tuovinen, Merriënboer and Darabi, 2005). The relationship
between cognitive load and motivation is also stated by
Moos (2009).

Adaptive learning systems
An adaptive learning system aims to provide a personalized
learning resource for students, especially learning
content and user-preferred interfaces for processing their
learning (Aroyo et al., 2006). Brusilovsky (2001) has
indicated that two adaptation approaches can be used in
developing web-based adaptive learning systems, that is,
"adaptive presentation" which presents personalized content for
individual students, and "adaptive navigation
support" which guides individuals to find the learning content
by suggesting personalized learning paths. Other
researchers have further indicated the importance of providing
personalized user interfaces to meet the learning
habits of students (Mampadi, Chen, Ghinea, & Chen, 2011).
187
In the past decade, various adaptive learning systems have been
developed based on different parameters that
represent the characteristics or preferences of students as well
as the attributes of learning content (Wang & Wu,
2011). For example, Karampiperis and Sampson (2005)
proposed an adaptive resource selection scheme by
generating all of the candidate learning paths that matched the
learning objectives and then selecting the most fitting
one based on the suitability of the learning resources for
individual students. Hwang, Kuo, Yin and Chuang (2010)
further developed an adaptive learning system to guide
individuals to learn in a real-world environment by
generating the personalized learning paths based on the learning
status of each student and the relationships between

the authentic learning targets. It can be seen that the provision
of personalization or adaptation modules, including
personalized learning materials, navigation paths or user
interfaces, has been recognized as an important issue for
developing effective learning systems (Chiou, Tseng, Hwang, &
Heller, 2010; van Seters, Ossevoort, Tramper, &
Goedhart, 2012).
Several studies have been conducted to develop adaptive
learning systems based on learning styles or cognitive
styles. For example, Tseng, Chu, Hwang and Tsai (2008)
proposed an adaptive learning system for elementary
school mathematics courses by considering students' learning
styles and the difficulty of the learning content.
Mampadi, Chen, Ghinea and Chen (2011) developed a web-
based learning environment by providing different user
interfaces based on students' cognitive styles. Furthermore,
Hsieh, Jang, Hwang and Chen (2011) developed an
adaptive mobile learning system that guided individual students
to learn in a butterfly ecology garden based on
students' learning styles. However, few studies have considered
multiple learning criteria, including learning styles,
cognitive styles, and knowledge levels, for developing adaptive
learning systems.
Research questions
In this study, an adaptive learning system is developed based by
taking both cognitive styles and learning styles into
account. It is expected that the proposed approach can benefit
students in improving their learning achievement,
reducing their cognitive load and promoting their learning
motivation. Accordingly, the following research questions
are investigated:
1. Does the adaptive learning system developed based on both

Oxford Review of EducationVol. 35, No. 5, October 2009, pp. .docx

Oxford Review of EducationVol. 35, No. 5, October 2009, pp. .docx

Recommended

Recommended

More Related Content

Similar to Oxford Review of EducationVol. 35, No. 5, October 2009, pp. .docx

Similar to Oxford Review of EducationVol. 35, No. 5, October 2009, pp. .docx (20)

More from alfred4lewis58146

More from alfred4lewis58146 (20)

Recently uploaded

Recently uploaded (20)

Oxford Review of EducationVol. 35, No. 5, October 2009, pp. .docx