& Patrick Carmichael
Sakai: An Environment
for Virtual Research
The range of electronic
resources and tools available to researchers has increased far beyond
even what early enthusiasts such as Howard Rheingold envisaged when
they described how the internet would put the catalogues and contents
of the world's libraries on one's desktop (Rheingold, 1993; 90-91).
Research projects are increasingly using network technologies to
improve communication between project members, to safeguard data,
and to engage with the ‘users’ of their research. This has led to
the emergence of models of "e-Research" which are perhaps
best developed in the context of international scientific collaborations
in fields such as particle physics and astronomy, and specific projects
such as the Human Genome Project. At the same time, other, domain-specific
versions of "e-Research" are developing, with different
foci and characteristic patterns of collaboration.
In the field in which
we work, educational research, even small-scale publicly-funded
research projects are already expected to publish electronically
their findings and other research outputs and have a responsibility
to archive their original data. But with an eye to the future, there
have been calls for an increased role for electronic networking
for communication, collaboration and dissemination as part of a
commitment to sector-wide capacity building. McIntyre and McIntyre
(1999) and Dyson and Desforges (2002) suggest both that expertise
needs to be both shared between established researchers and that
development opportunities need to be provided for practitioners
and new researchers. Training for individuals needs to be complemented
by strategies which foster institutional and sector-wide capacity
to conduct research, undertake analysis, engage with users and develop
These changes have
been accompanied by the development of thinking about 'networks'
and 'networking' (in some cases, importing models of networks from
the world of internet communication) which has had an impact on
expectations of how research is conducted and disseminated. Networks
are increasingly seen not only as providing access to resources,
but also represent sites for knowledge construction and the development
of new professional practice. Rather than developing 'best practice'
and then attempting to transplant it to a new context, the network
metaphor suggests that knowledge construction and dissemination
requires a shared frame of reference. The question then becomes
how to ensure that this shared frame of reference is preserved and
knowledge embedded across a distributed organisation (see Hakkarainen,
Palonen, Paavola and Lehtinen, 2004; 73+ for a fuller discussion).
In this article we
will describe Sakai, a novel electronic collaboration environment
designed to support e-Research, and will reflect on some of the
issues which have arisen from the first year of our using this platform
in our own work and to support other collaborative and distributed
research projects in the UK.
The SAKAI Virtual
As Wenger states in
his review of ‘community-building’ technologies, “ideal systems
emerge from combinations and convergence” (Wenger, 2001; 5). Sakai
responds to the demand by offering a modular architecture in which
various ‘tools’, services and resources can be combined within a
single, access-controlled framework (Fraser, 2005). The system is
web-based and users require no special software other than an up-to-date
Sakai emerged from
the world of Virtual Learning Environments (VLE) and as such can
be configured to support e-Learning and Distance Learning, with
(for example) schedule, syllabus, assignment and gradebook tools.
Alternatively, it can be set up to work primarily as a personal
information management (PIM) system for secure online access to
a personal file store and other productivity tools. Our interest,
however, has primarily been in its configuration as a Virtual Research
Environment (VRE), in which tools for collaboration within and between
groups of researchers take precedence over other functions. Our
experience has been gained in three arenas: the Applied Educational
Research Scheme (AERS) of Scotland; in our work with a number of
pilot projects at the University of Cambridge; and in an evaluation
of SAKAI being undertaken as part of the Joint Information Services
Committee (JISC) Virtual Research Environment Programme. This programme
involves a range of UK Universities in development and evaluation
activities across different disciplines: Social Sciences; Arts and
Humanities; Medicine; Technology and Science.
The key unit within
Sakai is the ‘worksite’ – a group of tools and resources with a
specific membership. Individual users can be ‘subscribed’ to any
number of worksites, each of which may have different sets of tools
and within which they may play different roles. An individual may
be the ‘maintainer’ of one worksite, meaning that they manage membership
requests, moderate discussions and email lists and make announcements
to the group, while simultaneously being a member or other worksites
in which they are simply contributors to discussions and readers
of others’ work. They can also be an 'accessor' with much more limited
access to tools and resources. However, it is possible to adjust
the permissions of both maintainers and accessors to reflect the
needs and purposes of the group using the site. Looking across the
range of Sakai users with whom we work, we see everything from open-access
groups with hundreds of members to small, temporary teams of two
or three researchers working on specific and private tasks such
as writing or analysis.
When configured as
a VRE, we characteristically see groups of researchers (who can
configure their ‘worksites’ to match their needs) using a a range
of tools offering project planning and management (Schedule) synchronous
and asynchronous communication (Chat, Discussion, Email Archive,
Announcements) to document sharing and storage (File Store, Email
archive, Web content tool) to co-authoring and analysing documents
and data (Wiki). Figure 1 shows a typical Sakai worksite with multiple
1: A Sakai 'Worksite' showing multiple collaboration tools.
Visible are 'panes' with worksite information, recent announcements,
discussions and chat. The left menu contains links to other tools
and shows which members of the worksite are currently logged in.
Provided that the research
team makes the VRE their primary locus of interaction (Wenger 2001),
the environment helps to create a continuum for collaborative work
and communication between face-to-face meetings, and generates a
record of communications for future reference. By providing a distinctive,
common workspace for the team, the VRE can also reinforce the group’s
identity by shared ownership of the worksite and its contents.
In our experience, not all research groups make this qualitative
change to their working practice; for some, the availabilty of a
specific tool within an access-controlled environment is sufficient
reason to use Sakai. In other cases, project members who already
use specific electronic tools are cautious about making the VRE
their sole locus of interaction, and may continue using tools such
as email lists, instant messenger or local file stores alongside
the new environment. At its most prosaic level, the VRE has been
seen by as a convenient way of addressing the requirements of funders
to have a presence on the World Wide Web, a 'communication strategy'
and a means of archiving project data and documentation.
In some cases, individuals
who have limited time have been happy to support deployment of the
VRE but have been only peripherally involved in online activities,
or have delegated others to play more active roles. We have found
that the most pragmatic approach to adopt is to support those individuals
and projects which see the VRE a way of addressing specific needs
and demands, while at the same time encouraging those who might
use the VRE as their main locus of interaction or to develop novel
patterns of work and collaboration.
Sakai as a Virtual
Research Environment. Some Examples
In this section we
will describe how three research groups have configured and used
the SAKAI platform in support of their research activities. These
are drawn from amongst the projects of the UK’s ESRC Teaching and
Learning Research Programme (who participate in the evaluation programme
mentioned above) and the Applied Educational Research Scheme.
Project A: The
first example is a research project investigating the ‘learning
biographies’ of adults in the UK; it involves researchers from four
geographically-distributed universities who are collecting survey
data from a large population and addtionally developing detailed
case studies of a smaller number of respondents. For this project,
it was important that reseachers had opportunities to ‘iterate’
between quantitative and qualitative data in the analysis process,
so a priority was the development of a structured archive of research
data accessible from all the research sites. At the same time, it
was essential that data remained confidential and that access to
data was carefully monitored. What emerged was a configuration of
the VRE in which only a limited set of the tools – those concerned
with data storage and project news - were used to any great extent
and membership was restricted to project researchers.
Project B: Our
second example is a research project based at a single UK university
but involved in a set of related research activities. This project
began using the VRE from the outset, and as a result much of the
early activity involved project management, the development of research
instruments, and the negotiation of access to research sites. As
such, a wider range of VRE tools were used: document storage was
important as research instruments were developed and literature
reviewed; but at the same time synchronous and asynchronous communication
was important, with ‘chat’ playing an important role both as a means
of maintaining contact between project members and producing a record
of decisions taken. This project was quick to see the potential
of the VRE for engaging users with the work of the project, and
set up multiple worksites for public access, the project ‘advisory
group’ and each of the subgroups within the project.
Project C: A final
example is rather different in that it uses worksites specifically
developed to support research communities whose members include
researchers in Higher Education, policymakers, teachers in schools,
and school students. These owe much to the concept of the ‘Community
of Practice’ (Wenger, 1998) in which a community of people engages
in shared activities and practice and have a ‘shared repertoire
of resources’ which develops over time. With their emphasis on developing
knowledge rather than the preservation of practice, they have much
in common with the ‘Innovative Knowledge Communities’ described
by Hakkareinen et al (2004).
These communities use
the VRE in ways designed to strengthen community identity, encourage
discussion, and co-construct and share knowledge. When members come
from different backgrounds and have varying degrees of expertise
in the area of enquiry, they bring new perspectives and themselves
to the group and have to accommodate those of others. A collaborative
process may then evolve in which participants have changing roles
within the work of the group depending on the phase the project
is at. In our experience to date the work of the group has been
governed both by the individuals' expertise that they have been
able to bring to that particular phase of the work and crucially
by other factors which have determined by the capacity and ability
(most importantly time constraints) to participate, which the VRE
has significantly increased.
The VRE worksites are
characterised by a high and sustained use of discussion tools; collaboration
around writing tasks and use of the file store tool to maintain
a record of developing knowledge. Another characteristic is that
members of these worksites have 'permissions' set so as to encourage
the discussion and contribution - rather than having a small number
of 'maintainers' and a larger number of 'accessers' (as explained
earlier), roles are shared and responsibilities distributed across
Sakai as a Developing
Platform for Collaboration
Sakai is not only a
comparatively new software environment, but is also a 'community
development' project involving teams of developers spread across
a large number of institutions. As with many developing pieces of
software, there have ‘teething problems’, exacerbated by the fact
that the projects we describe here were all ‘early adopters’ working
with versions of the platform which lacked the full functionality
of what is now (summer 2006) a better developed and more stable
environment. In some cases, users with experience of other software
(Virtual Learning Environments, discussion tools and digital archives,
for example) found the feature sets and ‘affordances’ of specific
tools disappointing; for others with less experience, it was the
apparent complexity of the web based environment which provided
the greatest challenges.
Other issues, some
of which are now resolved, have been related to the community development
process: there are some differences between tools developed by different
teams (for example, some have integrated search facilities while
others do not) and some combinations of tools 'play together' better
than others. In addition, there have been times when the priorities
of the developer community have seemed not to align with the needs
of specific projects - for example, when developer priorities to
develop the underlying infrastructure of the platform has taken
priority over the development of specific tools. At the same time,
the community development model does allow groups of users to 'lobby'
for the inclusion of new tools and the development of new features
in a way which would be much more difficult if Sakai was a 'closed'
proprietary product. This has led to user suggestions being taken
onboard by the developer teams when possible and are being addressed
in major upgrades of the software, which come out approximately
every six months.
The greatest challenge
for the maintainers of the research sites, then, has proved to the
activation and motivation of users, encouraging them to see past
individual or localised problems and make an informed assessment
as to what Bereiter and Scardamalia (1993:133-152) call the ‘promisingness’
of the VRE as a strategic development. The embedding of project
worksites within a broader Sakai community with many users and groups
allows the identification of opportunities to deploy tools in support
of their research activity. This means that another important role
for administrators is to set up 'sandbox' and demonstration worksites
so that existing and potential users can see what others have done
with the Sakai 'toolkit' and consider how it might impact upon their
A good example both
of the responsiveness of the developer community and of the means
by which new tools are disseminated is the uptake of the 'wiki'
writing tool. The development of this tool was informed by a need
for a collaborative writing environment (in most cases, as a replacement
for project members sending documents with 'tracked changes' to
each other by email). Once the wiki tool was made available and
its existence publicised, individuals and project teams were quick
to identify ways in which they might employ it; not just for collaborative
writing of abstracts, papers and reports, but also in collaborative
analysis, in biographical research and for the compilation of glossaries,
bibliographies and literature reviews.
2: The Wiki Tool within Sakai. A wiki allows members of a
worksite to work together on a document, editing and elaborating
it through a standard web browser; any user can see the ‘history’
of the document including which edits have been made by different
We asked researchers
in the projects described above to reflect on their characteristic
and shared repertoire (Wenger, 1998), ‘ways of thinking and practising’
(Entwistle et al. 2002; Meyer and Land, 2003) and barriers to collaboration.
We also encouraged researchers to identify in what ways they were
or were not supported by existing electronic tools and platforms.
This process frequently brought to light issues which were not necessarily
spelt out in project designs and publications, but emerged as a
result of collaborative and collective elicitation of participant
‘tacit knowledge’ within a structured activity, the value of which
is highlighted by Nonaka and Takeuchi (1995) and Engeström (1999).
Even at this early stage in our work, the potentially transformative
impact of the VRE was becoming apparent.
The examples we have
drawn on in our descriptions are all from educational research projects,
and so reflect some of the specific concerns and characteristic
approaches of that domain. However, what McAteer, Crook, Macleod,
Tolmie & Musselbrook (2002) call the 'issues to manage' in the
context of online communities transcend disciplinary boundaries.
The key issues and associated questions and decisions with which
the TLRP and AERS projects have engaged will face researchers in
many contexts. Indeed, in the deployment of SAKAI to diverse groups
at the University of Cambridge, we have found that the following
issues have meaning and relevance across disciplinary boundaries.
The first issue was
what we came to refer to as the 'focus' of collaboration,
the key activities or points in the 'workflow' of the project where
collaboration was most evident, or was an important or essential
element of a broader process. In some of the projects much of the
collaboration was focussed on elaborating project designs, developing
research instruments and reviewing literature in order to develop
research questions and working hypotheses. But we also saw collaboration
once projects began to collect and analyse data, and we found that
this collaboration manifested itself differently in different contexts.
Some projects had a clear commitment to expose their entire data
set to a wide audience while others restricted access to some data,
citing reasons which ranged from issues of respondent anonymity
to purely pragmatic questions of workload and lack of time. We
found it useful (given the educational context of our study) to
relate this back to Stenhouse's (1978) distinction between case
data, the case record, case studies, and analysis.
While in some projects this focus was indeed the 'raw' case data,
in others, collaboration was focussed on data selected by an individual
or group within the project, or even on cross-case analyses, with
researchers not revisiting original data at all.
The second issue is
that of participant roles and responsibilities and the expectations
that participants have of each other. We found a range of organisational
and management structures within projects and widely varying roles
for research participants. Several of the projects with whom we
are working are now considering how the VRE can support distinctive
elements of their research designs including extended relationships
with respondents in longitudinal studies, participants who themselves
are generating reflective accounts or 'action research' projects,
and those which are concerned with the expression of 'student voice'.
A third issue is how
the group relates to larger groups and particularly to those
to which they report or have other responsibilities. Altrichter
(2005; 22) describes how much educational research takes place in
'small collegial groups' protected by 'special conditions of confidence'
and in which it is possible to test and develop arguments and prepare
for a 'public' that is one step 'bigger'. Several of the groups
who are using the VRE have responded to this need to address ‘graded
publics’ by developing multiple worksites with different memberships,
together with workflow processes by which resources are transferred
from one area to another. This, of course, presents another dilemma
- whether to engage with graded publics through a process
of inviting them 'in' to the VRE or to use the VRE as a base from
which to address them - what McQuail (2000; 129-132) characterises
as 'consulation' and 'conversation' as opposed to 'allocution' or
A fourth significant
issue relates to the nature of the research group itself.
The educational programmes to which we have deployed the SAKAI VRE
are large and complex organisations. While they are both involved
in the coordinate research activities, they are organised in slightly
different ways; the TLRP is a 'coordinated research programme' within
which there are projects of varying sizes, thematic groups and seminar
series and a small number of research fellowships held by individuals.
AERS is organised into 'thematic networks' within which are projects
and individual research fellows. Both the TLRP and AERS are also
keen to support the development of individual and institutional
capacity across the wider educational research community beyond
the networks they facilitate and the projects they fund. When we
came to design VRE worksites as the 'virtual' manifestations of
these various groupings, however, we became aware that the notion
of the 'project' in particular conceals a very wide range of organisational
and collaborative configurations. Many of the projects represent
temporary coalitions of individuals based in different institutions.
In some cases these individuals have a previous history of working
together, but in others ‘the project’ represents a first attempt
at collaborative activity. Even then, ‘project’ organisation varies
widely. Some projects only convene meetings attended by all members
once or twice each year, or arrange these to coincide with other
events such as conferences. Others invest considerable time (and
money) in maintaining a regular ‘cycle’ of meetings every month,
or even more frequently. Another area in which there is a wide
variation is the extent to which the project is centrally managed;
some have an established ‘management group’ which oversees activity
in participating institutions and research sites, while others have
looser ‘federated’ structures with minimal central coordination.
Any successful deployment
of an application as the VRE, then, needs to consider the organisational
form of the group to be supported. While we have talked about ‘communities’
in the broad sense, most of the projects we currently support are
in fact similar to what Swaak, Verwijs, & Mulder (2000) describe
as ‘task groups’, with external funding and reporting responsibilities
and (to a greater or lesser extent) an externally defined research
agenda to address. As the VRE platform has becomes more established
and users more confident, we have noted that there has been a tendency
for groups to establish worksites for specified purposes rather
than to provide an online ‘home’ for an entire project. Small groups
set up worksites to analyse data, engage with specific users and
to write documents, apparently without any expectation that these
will continue to exist beyond the life of the activities concerned.
These self-directed, temporary groupings seem to correspond more
to the ‘knotworks’ described by Engeström, Engeström & Vähäaho
(1999). Individuals and groups may need to work together to identify
what organisational and network forms are best ‘fit for purpose’
for their intended research activities; deployment of the VRE may
represent an opportunity to ‘leverage’ discussions to this end.
A challenge at programme and institutional level then, as Swaak,
Verwijs, & Mulder (2000) suggest, is how to embed knowledge
and useful practices, introduced developed within these task-oriented
groups within a broader, long-lived and self-regulating community.
We have been promoting
and supporting the use of Virtual Research Environments for some
time now. Looking back over the past year (2005-2006) what we now
find is that individuals and groups do not simply identify those
tools and services which address specific and predefined project
'needs'. Increasingly, we also find them discussing the potential
of new tools to qualitatively change their ways of working; their
relationships with research participants; and role of the VRE in
ensuring the sustainability of their research activities. Our longer-term
interest is in exploring to what extent use of appropriate technologies
can not only support established 'ways of thinking and practising',
but how they can support different kinds of research activity and
new relationships between researchers and research participants.
The authors themselves
have made progressively more use of the VRE. This article was written
collaboratively in a specially-configured VRE worksite using the
'wiki' tool, with one author based in Cambridge and the other in
Strathclyde. For this reason amongst many we would like to acknowledge
the efforts of the Sakai developer team at CARET in Cambridge who
were responsible for building that particular tool. We would also
like to thank members of the SAKAI community and the participants
in the TLRP and AERS research programmes who have participated in
the development and evaluation of the Sakai VRE as a whole. Of these,
we would especially like to acknowledge the contribution of the
Learners, Learning and Teaching Network of AERS in the creation
of innovative ways of using the Sakai VRE.
If you are interested
in seeing what the Sakai software looks like in real life, you can
register to test drive the environment at http://www.sakaitestdrive.com/.
Once you have registered please wait for an email confirming your
registration and follow the link provided.
(MA), Research Officer at the Applied Educational Research Centre,
the University of Strathclyde, is a researcher and an administrator
of the Virtual Research Environment for the Applied Educational
Research Scheme (AERS) of Scotland. Sanna is a member of the Learners,
Learning and Teaching Network of AERS. She graduated from the University
of Turku in 2000.
Dr. Patrick Carmichael
is Head of Evaluation at the Centre for Research in Education Technologies,
University of Cambridge. Since 2001 he has been a member
of the ESRC project 'Learning how to Learn: in Classrooms, Schools
and Networks' and he currently manages a JISC-funded project on
online research collaboration using 'Virtual Research Environments'.
Altrichter, H. (2005)
'The Role of the Professional Community in Action Research' Education
Action Research 13(1) 11-26.
Bereiter, C. and Scardamalia,
M. (1993) Surpassing Ourselves: an Enquiry into the Nature and
Implications of Expertise. Chicago: Open Court.
Dyson, A. and Desforges,
C. (2002) 'Building Research Capacity: Some Possible Lines of Action'
Engeström, Y. (1999)
‘Innovative Learning in Work Teams: Analyzing Cycles of Knowledge
Creation in Practice’ in: Engeström, Y., Miettinen, R. and Punamäki,
R-L. (Eds.) Perspectives on Activity Theory Cambridge: Cambridge
University Press; 377-404.
Engeström, Y., Engeström,
R. & Vähäaho, T. (1999) When the Center Does not Hold: the Importance
of Knotworking. In Chaiklin, S. Hedegaard, M. & Jensen, U.J.
(Eds.) Activity Theory and Social Practice: Cultural-Historical
Approaches. Aarhus University Press, 345-374.
Entwistle, N., McCune,
V. & Hounsell, J. (2002) 'Approaches to Studying and Perceptions
of University Teaching-Learning Environments: Concepts, Measures
and Preliminary Findings' Edinburgh, Coventry and Durham: ETL Project
Occasional Report 1. Online at: http://www.tlrp.org/dspace/handle/123456789/97
Fraser, Michael (30
July 2005). Virtual Research Environments: Overview and Activity.
Ariadne 44. Online at: http://www.ariadne.ac.uk/issue44/fraser/intro.html
Hakkarainen, K., Palonen,
T., Paavola, S. & Lehtinen, E. (2004) Communities of networked
expertise: Professional and educational perspectives Amsterdam:
McAteer, E., Crook,
C., Macleod, H., Tolmie, A. & Musselbrook, K. (2002) 'Learning
Networks and the issue of Communication Skills' in C. Steeples &
C. Jones (Eds.) Networked Learning: Perspectives and Issues
London: Springer-Verlag; 309-321.
McIntyre, D. &
McIntyre, A. (1999) 'Capacity for Research into Teaching and Learning'
Cambridge: TLRP. Online at: http://www.tlrp.org/dspace/handle/123456789/330
McQuail, D. (2000)
McQuail's Mass Communication Theory London: Sage.
Meyer, J. & Land,
R. (2003) 'Threshold Concepts and Troublesome Knowledge: Linkages
to Ways of Thinking and Practising within the Disciplines' Edinburgh,
Coventry and Durham: ETL Project Occasional Report 4. Online at:
Nonaka, I. & Takeuchi,
H. (1995) The Knowledge-Creating Company: How Japanese Companies
Create the Dynamics of Innovation New York: Oxford University
Rheingold, H. (1993)
The Virtual Community: Homesteading on the Electronic Frontier
New York: Addison-Wesley
Swaak, J., Verwijs,
C. & Mulder, I. (2000) Task Groups and Communities Compared:
Can results from task groups be transferred to communities? Enschede:
Telematica Insitute. Online at: https://doc.telin.nl/dscgi/ds.py/Get/File-10580/Task_groups_and_
groups_be_transferred_to_communities_.doc (accessed 01-03-06)
Stenhouse, L. (1978)
'Case Study and Case Records: towards a contemporary history of
education British Education Research Journal 4(2) 21-39
Wenger, E. (1998) Communities
of Practice: Learning, Meaning and Identity Cambridge: Cambridge
Wenger, E. (2001) 'Supporting
Communities of Practice: a survey of community-oriented technologies'
Etienne Wenger Research and Consulting. Online at: http://ewenger.com/tech/index.htm