Learning virtual project work
Pentti Marttiin
1,3
, Göte Nyman
2
, Jari Takatalo
2
, Jari A. Lehto
3
1
Department of Information Systems, Helsinki School of Economics, P.O. Box 11, Finland
2
Department of Psychology, University of Helsinki, P.O. Box 9, Finland
3
Nokia Technology Platforms, P.O. Box 45, 00045 Nokia Group, Finland
Abstract. There are a variety of challenges in organizing productive virtual
collaboration projects. Problems of commitment and co-ordination are all
common. Organized use of collaboration technologies with appropriate
organization and management of activities can help avoid these obstacles. We
have developed and tested a model that is intended to support learning of
project management and virtual teamwork. It can be applied for introducing
collaboration models and practices to global firms, or as we have done, to
collaboration between university seminars. The model includes a multi-layered
organization structure according to which virtual or semi-virtual teams are built.
Facilitation of the exercise is supported with students operating in two layers:
coordination and reflection layers. The model has evolved and we report
experiences from tests with two seminars during 2003. The first seminar was
arranged between courses at three universities, and two universities was
participating the second one. We describe the basic constructs of the model, our
experiences and preliminary results of studies.
1 Introduction
Virtual and especially semi-virtual communication, collaboration, and learning have
become an essential part of organizational life. Unpredictable and dramatic factors
such as September 11th and the recent sars epidemic have boosted the need for distant
collaboration. Global organizations, subcontractor networks, distributed teams, and
many other areas of production and services meet the challenge.
While it is customary to discuss virtualness mainly as an entity of its own [6, 8, 15,
16], the de facto challenge of present organizations is how to best manage the mix of
their virtual and other activities. Virtual collaboration can become expensive in terms
of human resources and time needed for communication and coordination. The
adoption of new tools may introduce unpredicted organizational friction due to
insufficient understanding of roles and networked performance. A strong
organizational culture can even prevent novel interaction through the network [18],
and the lack of shared contextual experiences between members can cause inefficient
communication and motivation [13].
We can distinguish between two learning strategies in utilizing information and
communication technology (ICT). In the open, self-organizing, approach individuals
and organizational units are supported in their activities by offering access to
databases, links, services and applications through numerous channels [12]. In doing
Marttiin P., Nyman G., Takatalo J. and A. Lehto J. (2004).
Learning virtual project work.
In Proceedings of the 1st International Workshop on Computer Supported Activity Coordination, pages 91-102
DOI: 10.5220/0002681500910102
Copyright
c
SciTePress
so organizations and individuals are trusted to be capable of using new tools and
materials for their benefit. The focus is in finding and developing optimal tools and
environments. The open strategy is not uncommon to most business organizations,
perhaps because of the richness of technologies and business processes that exist.
In the closed or integrated approach, certain application environments or working
models are provided to support projects (e.g. http://www.knowledgeforum.com
).
Typically, a software framework, application, or model of individual or organizational
work is applied to provide the structure and function for the work. The selected
collaboration model and groupware are often used to support work processes.
In practice, there is a need to apply a mixed strategy. Reasons for this include fast
changing technology, a risk of incompatibility in adopting new tools, and an ongoing
change that concerns both information technology and organizational structure and
processes. Whatever the particular situation is, organizations must rely on virtual or
semi-virtual processes that are built according to an explicit or implicit model of
collaboration. Such models are not abundant and, hence, people in organizations must
continuously learn new ways of virtual collaboration and project work.
In this paper we present a model for learning project management and virtual
teamwork. We believe that a mixed strategy with a planned co-ordination and
reflection is beneficial especially in cases where the members of the virtual or
distributed team do not share a common organizational history or apply a predefined
process model. This is typical e.g. for new distributed teams, collaboration networks
consisting of variable organizational cultures, and for new members of organizations.
The model is called Virtual Project Model (VPM) and it aims to constitute a
structure for a “learning by doing” exercise. VPM contains layers of co-ordination
and reflection. Tasks are partly emergent and self-organized as in open model. The
exercise reported in this paper was arranged twice as a part of higher education
courses: the first exercise between three Finnish Universities (Spring 2003) and the
second one (Fall 2003) between two Universities involved also in the earlier exercise.
The coordination layer details differed between exercises. Based on our experiences
we believe participants can become committed if they are guaranteed a well-defined
structure of collaboration and a certain amount of power and responsibility.
The paper is organized as follows. Section 2 describes the basic concepts of the
VPM. We use coordination theories to discuss how co-ordination is arranged. Section
3 goes through two seminars in which the VPM has been applied and tested. We also
discuss our experiences and show preliminary results of studies done during the
exercises. Section 4 presents concluding remarks and future work.
2 Virtual Project Model
Originally, the development of Virtual Project Model (VPM) had the goal to support
reflection of distributed collaboration experiences. When consulting distributed
projects at Nokia [10], we built an approach to increase management’s knowledge of
various individual, team, and community related aspects. We concluded that teaching
these aspects and practices require an organized interplay between individuals sharing
a common task. In order to develop and test VPM we arranged a joint exercise
between University seminars. It is based on the following principles:
92
1. Learning by doing,
2. Individual commitment,
3. A model of collaboration, and
4. Technology independence.
Firstly, the purpose of VPM exercise has been to give students possibility to learn
and enhance their skills in the areas of project management and teamwork, especially
in a distributed environment. It is necessary to give students means to conceptualize
and experiment [11, 14]. Before starting the exercise we provided relevant practices
and tips [4, 6], process models [7] and frameworks [8, 10].
Secondly, project groups and also university seminars that do not share
organizational or other contextual motivation for collaboration may suffer from the
lack of commitment and isolation of member activities. In order to achieve a better
motivational basis for collaboration the exercise takes place as interplay between
teams, participants have a responsibility to act and interact, and a sufficient feedback
is guaranteed. A time window is reserved for teams to build their team identities, e.g.
imaginary company, logos, and virtual places (see [7]).
Thirdly, VPM is based on a model of collaboration: it embodies structures enabling
communication, co-ordination and reflective feedback. Goals are given on a coarse
level but details are left emergent. Communication and co-ordination processes are
discussed in Sections 2.1 and 2.2, and reflection processes in Section 2.3.
Finally, the exercise is carried out with the help of various ICT tools including
phone, e-mail, videoconferences, discussion areas, and shared repositories.
Technology is not the main focus on the exercise. Instead, people choose among the
tools they have. Our goal is to make VPM work for virtual teams. However, teams in
our exercises are semi-virtual, that is, only some participant worked virtually.
2.1 Teams and responsibilities in VPM
We applied a multi-layered structure in which various parties are dependent on each
other. The parties are: Facilitators (Lecturers and Assistants), Co-ordination team
(Students), Sub-contracting teams (Students), and Research team (Students). Figure 1
shows the responsibilities and dependencies of each team.
Cross et al. [2] presented two important viewpoints in their discussion of myths in
networks. Firstly, when a network grows large no single actor knows what takes place
in the network. Secondly, when one key actor is becoming a bottleneck her tasks need
to be delegated. Several emergent situations that require quick reaction may occur.
Because of that we have arranged communication between facilitators and virtual
teams through the co-ordination team that is more aware of the whole process and
situation.
93
V
IRTUAL TEAMS
Sub-goal and
collaborative tasks
Role of project manager
Self-organized
RESE
A
RCH TEAM
Study the exercise and
teams
Utilization of frameworks
Reflective, results as
feedback
CO-ORDINATION TEAM
Goal decomposition
Co-ordination over teams
Tasks delegated from
facilitators
Self-organized
FACILITATORS
Learning initiative and goals
Tasks and instructions
(a customer)
Network environment
Collaboration model
communicate
interfaces
analyse,
reflect
coordinate specify
research
specify
coordination
analyse,
reflect
Fig. 1. Teams and responsibilities in VPM
Each virtual team has its own goal, e.g., a sub-goal of the main goal given to co-
ordination team. The VPM collaboration between virtual teams can be planned in
advance (e.g., joint tasks) or during the process by co-ordination team. It may also
occur emergently as a need arises.
Facilitators and the research team have a close relationship and aim at making
sense of work processes and team behavior in the exercise and to develop means for
data analysis. Research team studies what takes place in the exercise and how people
experience the work. It analyses each team and gives the results as feedback to them.
2. inter-team
3. intra-team
1. facilitation
Co-ordination team
and manager
(Virtual)
teams with
managers
facilitators
Fig. 2. Communication paths in VPM
Figure 2 shows the main communication layers and participants in each layer.
Communication in the first layer, that is facilitation, takes place between facilitators
and the manager of the co-ordination team. The second layer tackles inter-team
communication and takes between co-ordination team (nominated members) and
virtual projects (managers). The third layer is inter-team communication that takes
place among team members. All communication can be made transparent using a
team discussion forum.
94
We claim that without such a structure the roles of each party remain unclear and
can even harm the progress of the exercise. We believe that leaving the appropriate
co-ordination and decision-making responsibility for each layer has a facilitatory
effect on the team commitment.
2.2 Co-ordination processes and mechanisms
We use the co-ordination theory by Malone and Crowston [9] and the concept of “co-
ordination mechanism” by Schmidt and Simone [17] to discuss how co-ordination in
VPM is planned.
Malone and Crowston define co-ordination as “management of dependencies
between activities”. This includes an assumption that if there is no interdependence
there is nothing to co-ordinate. It is clear that actors performing interdependent
activities may have conflicting interests. The goal in VPM has been to delegate tasks
and responsibilities to students. The dependencies of VPM can be pre-defined by
facilitators, left as a responsibility of the co-ordination team, or to take place
emergently between teams. Teams themselves decide practices for their intra-team
level collaboration, which often leads to an agreement by discussion (see e.g. [19]).
Malone and Crowston [9] identified four following dependencies:
Task/subtask relationships.
When a set of activities are all subtasks for achieving
an overall goal they need to be integrated either through top-down goal
decomposition or through bottom-up goal identification. In VPM each team has its
own predefined goal. Facilitators and co-ordination team discuss the goal
decomposition. The main responsibility of task accomplishment is given to the co-
ordination team. The clarification of sub-goals requires also negotiation and
agreement between co-ordination team and virtual teams. At any milestone, co-
ordination team reviews the deliverables of each virtual team and checks
dependencies (e.g. conflicts in interfaces).
Relationships between producers and consumers.
This dependency occurs when
one activity produces something that is used by another activity. It requires the
sequencing and transfer of products, and ensuring their accessibility from the
perspective of the receiving activities. This dependency is an essential part of VPM.
The learning stresses negotiation skills for arriving at an agreement between teams.
The number of interfaces between teams determines the degree of difficulty.
Simultaneity constraints. Activities need to occur at the same time. This type of
dependency requires synchronizing of activities. In VPM the synchronization of
activities is arranged using a set of milestones. The synchronization need is due to
producer-customer dependencies discussed above.
Management of shared resources.
Whenever multiple activities share a limited
resource (e.g. person, deliverable, or tools) a resource allocation process is needed to
manage the interdependencies among these activities. VPM focuses on a fluent
communication instead of creating bottleneck roles (communication layers). Most
tools and data are available for all (all the time) videoconferencing time as an
exception.
A certain coordination mechanism can be embedded in dependencies between
teams or as a part of team co-working. Schmidt and Simone [17] define co-ordination
mechanism as a construct of a coordinative protocol (an integrated set of procedures
95
and conventions such as a review, a contract negotiation or a follow-up) and a related
coordination artifact in which the protocol is objectified (minutes of a meeting, a
contract and a status report as examples). Coordination mechanism reduces the
complexity of the dependency and the space of possibilities. Because students learn to
use their own conceptualizations, VPM protocols embody weak stipulations [see 15].
2.3 Reflection processes
A special emphasis on reflection processes is included into VPM. A research team has
been formed in order to augment the traditional feedback given by facilitators and
self-reflection of students. A central part of the research is the utilization of 4Q
framework which provides a holistic view to study distributed work [10]. Bartlett and
Ghoshal [1], for example, have recently noted the importance of human and
intellectual capital and people as key strategic resource of building competitive
advance. This has also been our focus of interest and it has been studied from four
viewpoints:
Personal work focuses on issues around personal tasks of an individual in a
distributed environment. This area covers the following issues: competencies, mental
frameworks, motivation, stress, personal ambitions, individuality, working style,
beliefs, and values. Work with people focuses on general social interactions that,
firstly, take place between persons not having a close or formal relationship, and,
secondly, are often affected by organizational culture, traditions, values, and
experiences. For example, various communities belong to this category. Personal
goals, needs and earlier experiences are often driving forces for collaboration.
Project/teamwork focuses on how people are involved with production as a member
of projects and teams. Here, project goals are controlled and schedules are often tight.
In this area personal values and capabilities are considered against project rules,
processes, schedules, managerial styles and the values of closest teammates. The
fourth viewpoint focuses on the work with data and information. It addresses
capabilities of a person to use and utilize data and information [5]
1
including qualities,
means, and tools.
3 Applying the virtual project model in two seminars
This section presents two seminars applying VPM model. We describe the
arrangements and main differences of the seminars, show the evolution of the co-
ordination layer, and present our experiences and results of analyses.
1
We prefer to use data and information instead of knowledge. Any of the four areas may be
present in knowledge creation processes.
96
3.1 “Coaching model” (Spring 2003)
Virtual Project Model was first applied as a joint exercise between three different
seminars and units: Organizational Psychology (University of Helsinki), IS Project
Management (Helsinki School of Economics), and Knowledge Management
(Tampere Technical University). Altogether there were 49 students participating the
exercise with largest representation from economics. The participants had extensively
variable backgrounds ranging from second year students to those having a several
year experience in work life.
We had a leading facilitator from each university (with assistant facilitators), a
coaching team (7 persons), eight virtual teams and one local “reference” team (4-6
persons per team), and a research team (3 persons). Facilitators delegated the
following responsibilities to the coaching team: collecting a project management
literature database, providing coaching services and consultancy for teams, and
steering the progress and preventing problematic situations of teams.
All virtual teams and local teams were given the same goal of designing a
competence information system (CIS) that will be implemented in a fictional
organization. The process was divided into three sequential tasks:
1. Team identification and project planning (2 weeks)
2. Definition of system requirements and use scenarios (4 weeks)
3. System design and deployment planning (5 weeks)
The following tools were used: learning environment called Optima [3] as a shared
repository and a bulletin board, IP based videoconferencing tools in milestones, and
e-mail and phone for team communication.
A special topic of interest was the virtual start up of teams. We started each period
by giving instructions virtually through the Optima. The first task included a virtual
kick-off and team building. We composed teams, nominated the first project
managers and published the information via Optima. Teams were encouraged to use
available tools for their collaboration instead of meeting face-to-face. The first two
milestones were arranged as a multi-point videoconference, status checks and internal
reviews of deliverables (project plan, requirements specification) as tasks. In the last
session all teams presented their final report of CIS design.
3.2 “Subcontracting model” Fall 2003
The second exercise was arranged together with two seminars from the previous case
“Organizational Psychology” and “IS Project Management”. Altogether there were 36
students, 70 % of them from economics. The following teams were set up: a co-
ordination team (8 persons), five subcontractor teams (4-6 persons in each) and a
research team (5 persons). Next we consider changes and their motivation to the
earlier exercise covering team responsibilities and goal decomposition, kick-off, and
tools used.
The main goal for the exercise was still to design a CIS. But now the coaching
dependency was replaced by co-ordination dependency. The responsibilities of the co-
ordination team were as follows: to be an intermediary between facilitators and sub-
contractors (as in Figure 1), to guide sub-contractors work, and to follow-up teams’
progress. Co-ordination team created the imaginary organization and stated
97
requirements for CIS. They reviewed project plans of subcontractors, signed contracts
with them, steered the progress and accepted design reports. All subcontractors had a
sub-goal to design a part of the CIS (system architecture, database design, user
interface, and two system modules). The responsibility of use scenarios (a part of task
2 in earlier case) was given to the subcontractor designing general UI.
The kick-off was now organized differently. First, all students had the possibility to
select which team to join. Facilitators held a meeting with the co-ordination team in
which the responsibilities were discussed. After the meeting the exercise proceed
mainly as interplay between coordinators and subcontractors. Only minor questions
were asked from facilitators.
Optima was used as a shared repository, a discussion forum and a messaging tool
as in previous case. Now also Optima discussion lists were taken into use for inter-
team communication, for the negotiation and the acceptance processes. In addition,
teams used e-mail and phone for inter-team communication. Videoconferencing
sessions were not arranged.
3.3 Discussion of co-ordination and reflection processes
In the first exercise (“coaching model”) the main focus was to study and learn virtual
work and virtual start-up. We expected a need for coaching services but started with a
simple arrangement without dependencies between teams, all teams designing CIS. In
this exercise the role of the coaching team remained unclear due to minimal demand
for coaching. We believe two reasons affected to the lack of this demand. Firstly,
joined tasks were not planned to occur between teams. Secondly, the process was
divided into three tasks, and teams expected feedback and instructions to continue
from facilitators. Teams reported that without planned milestones their interest was in
their own responsibility and not in to design the whole CIS.
Table 1. Coordination in exercises: coordination dependencies, mechanisms (M) and artifacts
(A)
Dependency “Coaching model” “Subcontracting model”
Task-subtask
relationships
Intra-team coordination (M)
Project plan (A)
Goal decomposition by co-ordination team
(M), Statement of work (A)
Intra-team coordination (M), Project plan
(A)
Producer-
consumer
relationships
Project phases (M)
Project plan (A)
The team is both a producer
and consumer.
Requirements and design made by different
teams (M), Contract and project plan (A)
Dependency management between co-
ordination and subcontracting teams (M)
Simultaneity
constraints
Milestone presentations at
intra-team level (A)
No simultaneity constraints
between teams.
Flexibility in schedules by co-ordination
team (M),
Progress report to coordination team (A)
Shared
resources
“Coaching” aid (M) Interfaces between teams (M)
Requirements documentation and FAQ (A)
Discussion list per dependency (A)
98
The coaching dependency was not totally unrealized. Coaching team used their
role in two cases. Firstly, one of the teams did not start during the virtual kick-off and
this required the re-allocation of students. As a result two projects joined together.
Another case was a team conflict. Team members of diverse background had different
opinions of quality of deliverables. Based on the experiences we concluded to have
more emphasis on co-ordination in the “subcontracting” exercise. Table 1 summarizes
how co-ordination is arranged in two exercises.
The research team used the 4Q framework for creating a questionnaire of 50
questions that monitored collaboration. We used 5 point Likert -scales, where the
anchors for the scales are: 1 = very low and 5 = very high. For the “coaching model”,
the study was done three times (4th week, 7th week, 10th week). The subjects were
instructed to use their earlier responses as a reference in order to guarantee within-
subject reliability. All 49 participants responded to the first survey, 47 to the second,
and 42 to the last one. Research team made an analysis for all teams and presented the
main findings of each teams. For the “subcontracting exercise” the test was
accomplished only at the end of the exercise (last week, 24 responses out of 33) and a
special feedback session was arranged.
We first summarize some findings of the coaching model with 8 virtual teams and
two local teams (including co-ordination team). After the virtual kick-off we
encouraged teams to proceed virtually but meeting face-to-face was not totally
forbidden. As we noted the virtual start-up was problematic for some teams. Another
disturbing factor was team members belonging to the same class and seeing each
other e.g. during courses. Thus, in addition to making a comparison between “virtual”
and local teams we compared persons working virtually with those working in a
mixed mode.
The overall profile of work in the exercises can be characterized by the following
data. 60% of time on average was spent working alone with assigned tasks.
Participants in local teams reported lower share for working alone. Presumably, they
could contribute to the tasks also during their meetings. It also seems that the amount
of face-to-face meetings was not essential for successful collaboration due to slight
differences between “virtual” and local teams. Semi-virtual teams met once in three
weeks and local ones once in two weeks. Probably the crucial factor was that teams
met or there was a possibility to meet when necessary.
One quarter of participants worked virtually, the rest met once per week or rarely
as discussed above. Virtual ones felt that they got less information from project and
other teams, and also were less satisfied with received feedback when compared
against others. However, they felt easier to share their work with others’ and also to
use others’ work. Furthermore, they felt better able to fit project goals and schedule to
their own ones.
99
Fig.3. Questionnaire results visualized as 4Q radar: means of two exercises.
The analysis of the subcontracting exercise is still ongoing and no final conclusions of
differences between two exercises have been done. However, as we expected from the
subcontracting model, the students were committed to the subcontracting course (they
gave the average score of 4.04 on a 5 point Likert scale in the 4Q test on this issue).
They also experienced the new arrangements (Section 3.2) as helpful. When
comparing means of respondents (Figure 3) we may assume to affect teamwork by
changing co-ordination. Differences in the area of “work with people” can be
interpreted as increased inter-team coordination in subcontracting exercise.
4 Conclusions
In organizing virtual collaboration and learning initiatives, there is no ready-made
knowledge of all factors (organizational, team-related, or individual ones) that come
to play in different contexts. We argue that this knowledge is partly tacit and we
consider useful to incorporate one type of research and/or monitoring function to
gather relevant knowledge and experiences from virtual work. This is especially
important in the early phases of projects. Often when projects are terminated the
obtained experiences are soon lost and there is a fast disruption of any shared
knowledge concerning the issues. The Virtual Project Model demonstrates a way to
0
1
2
3
4
5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
subcontracting model
coaching model
Personal work
Work with people
Work with data
and information
Team work
100
study and to understand these problems. Our experiences from two seminars are quite
positive and the research data and other feedback that we received was very
constructive for developing the model further.
This study does not give answers to the questions of the best way to build co-
ordination or reflection. Based on our experiences we can say that if the pedagogical
aspects are not planned or are a free aspect of the tasks they will not be realized. We
see the layers of co-ordination and reflection as a way to support these pedagogical
aspects. The whole exercise is built on the action towards the goal and on interaction.
Thus, the role of an individual student needs to be well defined and clear enough in
the network of teams and layers.
As a next step in VPM we do statistical analyses of the data collected from two
seminars, study the capabilities of students to work in each exercise and analyze
differences between seminars. As we have developed a more coherent co-ordination
layer after the first attempt, our own goals are to improve learning practices for the
seminar. Secondly, based on the experiences we are ready for testing the exercise as a
multi-cultural and organizational exercise.
References
1. Bartlett, C.A. and Ghoshal, S. Building Competitive Advance Through People. MIT Sloan
Management Review, Winter, (2002), 34-41.
2. Cross, R., Nohria, N., and Parker, A. Six Myths about Informal Networks. MIT Sloan
Management Review, Spring 2002, (2002).
3. Discendum Optima. 2003, Discendum Oy www.discendum.fi
.
4. Fisher, K. and Fisher, M.D. The Distance Manager: A hands-on guide to managing off-site
employees and virtual teams. McGraw-Hill, 2001.
5. Galliers, R.D. and Newell, S. Back to the future: from knowledge management to the
management of data and information. Information systems and e-business Management, 1,
1, (2003), 5-14.
6. Hoefling, T. Working Virtually - Managing People for Successful Virtual Teams and
Organizations. Stylus Publishing, LLC, 2001.
7. Kimball, L. and Digenti, D. Leading Virtual Teams that Learn. 2001, Learning Mastery
Press: Amherst, MA. p. 43.
8. Lipnack, J. and Stamps, J. Virtual Teams: Reaching Across Space, Time and Organizations
with Technology. Wiley, New York, 1997.
9. Malone, T.W. and Crowston, K. The Interdisciplinary Study of Coordination. ACM
Computing Surveys, 26, 1, (1994), 87-110.
10. Marttiin, P., Nyman, G., and Lehto, J. Understanding and Evaluating Collaborative Work in
Multi-Site Software Projects - A framework proposal and preliminary results. In 35rd
Hawaii International Conference on System Sciences). IEEE, 2002,
11. McCreery, J.K. Assessing the value of project management simulation training exercise.
International Journal of Project Management, 21, (2003), 233-242.
12. Mills, A. Collaborative Engineering and the Internet. Society of Manufacturing Engineers,
1998.
13. Olson, G.M. and Olson, J.S. Distance matters. Human-Computer Interaction, 15, (2000),
139-178.
14. Raelin, J.A. A model of work-based learning. Organization Science, 8, 6, (1997), 563-78.
15. Renninger, K.A. and Shumar, W. Building Virtual Communities: Learning and Change in
Cyberspace. Cambridge University Press, 2002.
101
16. Rheingold, H. The Virtual Community: Homesteading on the electronic frontier. Addison-
Wesley, Reading: MA, 1993.
17. Schmidt, K. and Simone, C. Coordination Mechanisms: Towards a Conceptual Foundation
of CSCW Systems Design. Computer Supported Cooperative Work: The Journal of
Collaborative Computing, 5, (1996), 155-200.
18. Wenger, E., McDermott, R., and Snyder, W.M. Cultivating Communities of Practice: A
Guide to Managing Knowledge. HBS Press, 2002.
19. Winograd, T. and Flores, F. Understanding computers and cognition. Addition-Wesley,
Reading, MA, 1987.
102