Content uploaded by Mika Yasuoka
Author content
All content in this area was uploaded by Mika Yasuoka
Content may be subject to copyright.
Collaboration Support
in the Initial Intercultural Collaboration Phase
Mika Yasuoka
Dept of Engineering, RCAST
University of Tokyo
4-6-1, Komaba Meguro Tokyo, Japan
+81 3 5452 5289
mika@kid.rcast.u-tokyo.ac.jp
ABSTRACT
Collaboration of participatory design is often carried out by
a wide variety of members from several work cultures
especially in the initial design stage. However, little is
known about how to support work culture difference. This
paper presents an analysis of participatory design case
applied to software development collaboration project. The
analysis is from an intercultural point of view and aims at
finding ways to support intercultural collaboration
activities. Investigation how intercultural communication
errors are recognized, repaired and bring agreements to
participants were made. Surprisingly, our analysis indicates
that the initial stage of a participatory design process
mainly relies on a dynamic and creative process where
participants create expressions together with unique
semantics rather than just transferring static terms from
each others own vocabulary.
Author Keywords
Collaboration Support, Intercultural Collaboration, Design
ACM Classification Keywords
H5.3. Group and Organization Interfaces: Collaborative
computing, Computer-supported cooperative work
H1.2 User/Machine Systems: Human factors, Human
information processing
1. INTRODUCTION
There are several design support techniques that apply to
participatory design such as personas, prototyping, video
card games [1], and design games [2]. All of them aim at
mediating and facilitating collaborative work among
participants such as developers, interaction designers and
uses. However, in current participatory design approaches
and their supplemental design support techniques, other
aspects of participatory design such as cultural, social and
political issues are often neglected [10].
In practice, collaboration activities in a participatory design
process are often carried out by a wide variety of members
from several work cultures. Wenger [13] calls a cluster of
people sharing the work culture communities of practice.
Participants in collaborative work bring together several
communities of practice with collective concerns. Fischer
[4] defines these as communities of interests. When people
form communities of interests, they tend to have different
preferences [10], culture, sense of values and terminology
[3]. Thus in collaborative work, it is believed people from
different work cultures need to understand each others’
cultures through sharing information and knowledge [5].
This paper investigates a collaborative activity in a software
development project. Investigation aims at finding ways to
support participatory design activities from an intercultural
collaboration point of view. We analyze data focusing how
collaborative communication process is carried out.
We expected to see a transferring process of terms with
static semantics brought by each participant to the project
team. Contrary to our expectations, participants seem to
reach agreement via a creative negotiation process where
new terms with unique semantics fitting to the task at hand
are developed. Our results indicate that it may be more
effective to promote creativity in the initial stage of
participatory design rather than supporting understanding
on others to facilitate intercultural collaboration.
The paper is organized as follows. First we introduce the
theoretical background of our approach to intercultural
collaboration support. Next in section three, characteristics
and its analysis of our participatory design case, software
development collaboration, are explained. In section four,
based on the analysis, we propose that the intercultural
collaboration process on participatory design can be
supported as creative design process rather than shared
understanding process. Finally in section five, we conclude
and discuss directions for future work.
In PDC-06 Proceedings of the Participatory Design Conference,
Vol II, Trento, Italy, August 1-5, 2006, under a Creative
Commons License. CPSR, P.O. Box 717, Palo Alto, CA 94302.
http://www.cpsr.org ISBN 0-9667818-4-8
2. INTERCULTURAL COMMUNITY COLLABORATION
It has been discussed that in collaboration among people
from different work cultures, people need intercultural
communication through sharing information and
knowledge. Intercultural communication means more than
129
just translating language of one culture to language of the
other at a symbol level [8] because it is often the case that
different work cultures have different semantics for
identical symbols and representations. This ontology drift
often occurs unconsciously in intercultural collaboration
[9]. People realize the ontology drift when a break down
[11] in communication occurs. Thus, to overcome different
perspectives and different vocabularies [4] is thought to be
the initial challenges of intercultural collaboration.
In collaborative activities among intercultural community
members, external representations often play an important
role as a boundary object [12]. Community members
gradually build ability to collaborate when each member
can associate a word and an object to his/her own
representation world [8]. External representations working
as boundary objects can be used as media either for sharing
or learning at one time, or for creating or negotiating
meanings [13] at another time.
In participatory design, participants bring together different
communities with different work cultures and interests.
Thus, it is beneficial to approach participatory design from
an intercultural collaboration point of view in order to
address social and cultural issues.
3. CASE STUDY
In this section, we analyze a participatory design case. The
overall objective of the analysis is to find ways to support
collaborative work of participatory design. We focus on the
process how intercultural communication errors are
recognized, repaired and bring agreement to participants
through external representations that function as boundary
objects. Through analysis, it is expected to answer about
what kind of external representations can be boundary
objects, how agreement are made through those key
external representations, and what kind of reasons
accelerate or prevent from reaching agreements.
3.1. Observation Experiments
The case is a software development project where four
participants collaborate. Their objective is to design a tool
for protocol analysis for academic researchers. An ethno-
graphical approach was taken for investigation.
The four participants are one client, two programmers and
one interaction designer. Three kinds of data, video taping,
voice and photos, were collected during a span of four days.
The total length of the discussions was twelve hours. Five
photos were taken for recording drawings on a white board.
Protocol was transcribed from video and voice data.
1st meeting: A client explained protocol analysis process to
programmers and an interaction designer. The client used a
projector to show real protocol data while explaining.
2nd meeting: Based on the first meeting, the interaction
designer asked questions, the client suggested desirable
functions of a future analysis system and the programmers
often interrupted conversation in order to show their initial
ideas for functions and approaches. Projectors and personal
computer screens were used to show ideas.
3rd and 4th meeting: Third and fourth meetings were held in
a row in a day. The programmers showed the initial system
functions and mockups. To explain functions projectors and
their personal computers were used. The client asked for
more functions, specified requirements in details, and
confirmed suggested functions.
3.2. Approach
External representations were identified in conversation
protocols as key expressions.
First, the transcribed 12 hours of conversation was analyzed
with the morphological parser for the Japanese language,
ChaSen1 that categorize each morpheme. The results of
Chasen were filtered into nouns, verbs and adjectives.
After that, the frequency of each word was calculated both
for the total conversation and for each participant. We
defined the top 30 words and expressions to be key
expressions (ex. Label, Concept, Coding Scheme).
Next, the transition of the number of usages of each key
expression was analyzed and visualized with colors based
on functions of Popout Prism2. Figure 1 shows a transition
of key expression Label in relation with other key
expression. Co-occurrence relations among key expressions
over time were also calculated and visualized with Polaris
[7] that reads morphological data and visualizes key word
relations in a graph structure based on the co-occurrence
calculation results.
The objective of this analysis is not to define and generalize
key expressions, but to understand how people discuss,
negotiate and agree by investigating usages of key
expressions. As a first step of our analysis, conversation
protocols are investigated among a wide variety of external
representations. Other kinds of multimodal representations
such as figures drawn on the white boards, distributed
documents, and repeatedly used body language are also key
external representations and will be targets for analysis.
1 http://chasen.naist.jp/hiki/ChaSen/
2 http://www2.parc.com/csl/projects/popoutprism/
130
Figure 1. Relations between Label and other key expressions in
Popout Prism
3.3. Analysis Results
In this section, three kinds of findings observed in
visualization results around key expressions are explained.
Negotiation of Meaning
Frequently used expressions throughout all meetings, for
example Label has been used by all members since the first
meeting. For example in the first 1.5 hours of the fist
meeting Label was used 44 times while in the last 1.5
hours of the fourth meeting, the expression was used 39
times. Co-occurrence graphs of each member shows that
co-occurrence relations between Label and other key
expressions differed drastically from person to person in the
first meeting. However, over time each member’s co-
occurrence graph came to have much similarity. Figure 2
shows how the co-occurrence graphs for Label for two
participants (programmer A and the client) become similar
over time.
Protocol analysis [14] shows that members seem to have
exchanged questions and answers as for Label repeatedly.
Negotiation of meanings through confirmations in relation
with another key expression are the most frequent.
Figure 2. Co-occurrence graphs for Label for two participants
Creation of Expressions
Key expressions, for example Segment Table has been used
frequently from the second meeting. The protocol analysis
shows that the key expression, Segment Table was coined in
the beginning of second phase through the negotiation
process shown in the following conversation. The
expression, Segment Table was used by all participants
since this moment.
Programmer A : This name.., how about grouping table?
Client : Well….
Programmer A : How about coding category table?
Client : It is odd. We call it segments.
Programmer B : Ok, let’s call it a segment table.
Convergence of Expression Usages
The key expressions Tree and Ki (means a tree in Japanese)
were used frequently in the beginning of meetings.
However, the usage of Ki decreased over time. For
example, in the first meeting, programmer A and the
interaction designer use an expression, Ki, while
programmer B and the client used the term Tree as tree
structure in their conversation. On the other hand, in the last
meeting, all participants used the expression Tree when
they meant tree structures.
-In the first half of the first meeting-
Programmer A : This Ki is
Client : When I think this as Tree structure
I. Designer : No, you mean the leaf of the Ki-structure
-In the second half of third meeting-
Client : Which part of the tree?
Programmer A : Bottom part of the tree…
4. DISCUSSION
From the analysis of the intercultural collaboration process
through key expressions, three kinds of collaborative
processes leading to agreement were identified.
1. Negotiation of Meaning
2. Creation of Expressions
3. Convergence of Expression Usages
In negotiation of meaning, participants discuss meanings of
key expressions to define exact meanings among
participants. The meanings often become unique to the
participants and could often be different from conventional
dictionary definitions. Observing and analyzing both the co-
occurrence graphs and conversation protocol, it became
clear that a definition of key expressions (Label in the
example) is gradually fixed through iterative interactions
among participants. Every four member’s co-occurrence
graph changed its semantic distance over time and became
similar in the end. Thus, in this collaboration process, exact
definitions of key expressions often don’t exist in the
beginning and they are created by negotiation.
In creation of expressions, participants collaboratively
coined names to new concepts during the discussion
through negotiations. The co-occurrence graphs clearly
show the moment when new expressions were coined. The
131
communication protocol often shows a similar creation
process of new expressions. Before participants started to
use newly created expressions, they often discussed ideas or
concepts that will lead to new expressions afterwards.
In convergence of expressions, when participants have
different expressions that indicate the same concept, the
shared concept often converges to one expression through
negotiations. Cross reference of co-occurrence graphs and
communication protocol shows that when the number of
usages of one expression increases drastically, it is
sometimes caused by convergence of expressions.
The findings shown in this paper indicates that new
meanings or new expressions are added to key expressions
or created in intercultural collaboration over time. In other
words, in our intercultural collaboration case, creations of
expressions or concepts are quite usual. The initial process
of intercultural collaboration might highly rely on this
creative process among members.
These analysis results suggest investigating whether
techniques for initial creative design support could also
facilitate the initial stage of intercultural collaboration. For
example, game approaches [1, 2] may be applied to
intercultural collaboration support. Although the game
approaches have their own aims, the overall purpose,
support collaboration across various interests, is the same.
The idea of using creative support for initial collaboration
support is also attractive for another reason. There are many
similarities between the target group of creative design and
intercultural collaboration communities such as temporary
(“people come together in the context of a specific project
and dissolve after the project” [4]), lack of shared
understanding and potentially creative. All the more, while
it is still not clear how and why important it is to support
shared understanding, creative support provides unique
opportunities for collaboration. So long as intercultural
collaboration is in the initial stage, creative support has high
potential to facilitate intercultural collaboration for
collaboration sake.
5. CONCLUSION
In this paper, we have analyzed a participatory design
process in the form of software specification meetings. We
have aimed at finding ways to support intercultural
collaboration by analyzing communication protocols of the
collaboration process. Our analysis indicates that the initial
phase of intercultural collaboration relies on a creation of
new expressions rather than just transmitting each others’
understandings. We concluded that creativity support have
high possibility to become an effective approach for the
initial stage of intercultural collaboration.
Further investigation is needed in order to apply creative
design approaches to collaboration support. In particular it
should be examined whether the creative design techniques
in participatory design can be applied to intercultural
collaboration support.
ACKNOWLEDGMENTS
Thanks to Kumiyo Nakakoji for supervising the research
presented in this paper.
REFERENCES
1. J. Buur, T. Binder, and E. Brandt, Taking Video beyond
‘Hard Data’ in User Centred Design, Design,
Participatory Design Conference, 2000.
2. E.Brandt, and J. Messeter, Facilitating Collaboration
through Design Games, Proc. Participatory Design
Conference, 2004.
3. K. Bødker, and J. Pedersen, Workplace Cultures:
Looking at Artefacts, Symbols and Practices, Chap6, pp.
121-136, Lawrence Erlbaum Associates, Hillsdale, NJ,
1991.
4. G. Fischer, Social Creativity: Turning Barriers into
Opportunities for Collaboration Design, Proc.
Participatory Design Conference, pp. 152-161, 2004.
5. J. Greenbaum and M. Kyng, (Eds.) Design at Work:
Cooperative Design of Computer Systems, Lawrence
Erlbaum Associates, Inc., Hillsdale, NJ, 1991.
6. S. Nomura, T. Ishida, M. Yasuoka, et al., Open Source
Software Development with Your Mother Language:
Intercultural Collaboration Experiment 2002, 10th
International Conference on Human Computer
Interaction, 2003.
7. N. Okazaki, Polaris: An Integrated Data Miner for
Chance Discovery, Proc. Workshop of Chance
Discovery and Its Management, 2003.
8. J. Ostwald, Knowledge Construction in Software
Development: The Evolving Artefact Approach, Ph.D.
Dissertation, University of Colorado at Boulder, 1996.
9. M. Robinson and L, Bannon, Questioning
Representations, Proc. of the Second European
Conference on Computer Supported Cooperative Work,
pp219-233, Kluwer Academic Press. 1991.
10. K. Rönkkö, M. Hellman, B. Kilander and Y. Dittrich,
Personas is not Applicable: Local Remedies Interpreted
in a Wider Context, Proc.Participatory Design
Conference, 2004.
11. E. Schegloff, Conversation Analysis and Socially
Shared Cognition, Perspectives in Socially Shared
Cognition, L. Resnick (et. al, Eds.), Chapter 8, pp.150-
171, American Psychological Association, 1991.
12. S. Star, The Structure of Ill-Structured Solutions:
Boundary Objects and Heterogeneous Distributed
Problem Solving, Chapter.2, pp.37-54, Morgan
Kaufmann Publishers Inc., San Mateo, CA, 1999.
13. E. Wenger, Communities of Practice: Learning,
Meaning and Identity, Cambridge University Press,
Cambridge, 1999.
14. H. Kaiho and E. Harada, Protocol Analysis, Shinyosha,
1993 (in Japanese).
132
