Beruflich Dokumente
Kultur Dokumente
Harry Brignull
March 2005
Declaration
I hereby declare that this thesis has not been submitted, either in
the same or different form, to this or any other university for a
degree.
Harry Brignull
Contents
i Acknowledgements
ii Preface
iii Abstract
5 Chapter 1
Introduction
14 Chapter 2
Background
55 Chapter 3
Studying the social nature of a large display in a
communal space
76 Chapter 4
The adoption of the Opinionizer Community Display in
a one-shot setting
99 Chapter 5
Informing the design of Dynamo: a Community Display
for on-going usage settings
132 Chapter 6
Preliminary observational study of an on-going setting
145 Chapter 7
The adoption of the Dynamo Community Display in
an on-going setting
190 Chapter 8
Discussion
219 Chapter 9
Conclusions and future work
232 References
260 Appendix 1
Interview questions used in Opinionizer study 2
263 Appendix 2
Materials used in the Dynamo study
Acknowledgements
This thesis is dedicated to the memory of Mike Scaife, who sadly died in
December 2001. Mike helped me become an HCI researcher, so his
influence will stay with me for the rest of my life. Thanks Mike.
I must also thank the other people I collaborated with during the EPSRC
funded Dynamo project. Shahram Izadi, my research colleague, spent a
significant chunk of his life working closely with me on the project, and has
become a great friend. My thanks also go to Tom Rodden, whose distinctive
approach to motivation and management will not be forgotten.
Many other people have helped me along their way with their kindness,
advice and time: Jon Rimmer, Eric Harris, Sam Woolf, Paul Marshall, Jon
Matthews, Rowanne Fleck, and everyone at the Interact lab; Rene and the
AV team; the staff and students at Blatchington Mill; and of course the
support staff here at Sussex University – Christian Catherham, Richard
Grainger, Linda Thompson, Rita Stone, Celia McInnes, and everyone else.
i
Preface
Part of the work in these pages has also appeared in the various forms
below, and was part of the EPSRC funded Dynamo project (GR/N01125).
The papers below and the user-studies reported within them were the
product of collaborative group work between myself and the other Dynamo
project members: Shahram Izadi, Yvonne Rogers, Geraldine Fitzpatrick and
Tom Rodden. My contributions to this collaborative work included the
design and building of prototypes; the planning, design and running of user
studies; and a substantial part of the analysis. However, it should be noted
that the analysis and discussion carried out within this thesis is entirely my
own work, and considerably extends that reported within the papers below.
Brignull, H., Izadi, S., Fitzpatrick, G., Rogers, Y., & Rodden, T. (2004). The introduction
of a shared interactive surface into a communal space In Proceedings of the 2004 ACM
conference on Computer supported cooperative work (pp. 49-58). Chicago, Illinois, USA
ACM Press.
Brignull, H., & Rogers, Y. (2003). Enticing People to Interact with Large Public Displays
in Public Spaces. In Proceedings of Interact 2003 (pp. 17-24). Zurich, Switzerland.
Izadi, S., Brignull, H., Rodden, T., Rogers, Y., & Underwood, M. (2003). Dynamo: a
public interactive surface supporting the cooperative sharing and exchange of media In
Proceedings of the 16th annual ACM symposium on User interface software and
technology (pp. 159-168). Vancouver, Canada ACM Press.
Rogers, Y., & Brignull, H. (2002). Subtle ice-breaking: encouraging socializing and
interaction around a large public display. Paper presented at the Workshop on Public,
Community and Situated Displays at Computer Supported Cooperative Work (CSCW’02),
New Orleans, USA.
Rogers, Y., & Brignull, H. (2002). Designing dynamic interactive visualisations to support
collaboration and cognition. Paper presented at the Information Visualisation '02, London.
Rogers, Y., & Brignull, H. (2003). Computational offloading: Supporting distributed team
working through visually augmenting verbal communication. Paper presented at the
Cognitive Science Conference, Boston.
ii
Abstract
Community Displays are a new genre of large digital wall display system
for the support of informal social interaction in communal spaces. Prior
research shows that encouraging the initial voluntary adoption of
Community Displays in situ can be difficult, and is currently not well
understood. This problem is investigated in this thesis.
iii
Chapter 1
Introduction
4
1.1 Introduction
This thesis investigates the problem space of designing Community
Displays, a new genre of digital wall display system for the support of
informal social interaction in communal spaces. An analysis of prior
research reported in Chapter 2 shows that achieving initial voluntary
adoption by a community is a crucial “first hurdle” that many systems fail
on, even for those that are technically impressive and feature-rich. The
contribution of this thesis is that it provides the beginnings of a lingua
franca or framework for researchers and system designers, enabling them to
better understand the interplay between the settings, the social behaviour,
the voluntary adoption and the design issues that are at work in the design,
use and appropriation of Community Displays. The key concepts put
forward include:
5
This thesis is structured as follows: Chapter 2 provides the background to
this research, detailing a literature review, an analysis of the problem-space,
and the research questions. Following this, three case studies are reported,
beginning with an initial grounding case study in Chapter 3, a main case
study in chapter 4, and the largest case study is reported in Chapters 5, 6 and
7. Chapter 8 consists of the thesis discussion, and Chapter 9, the future work
and conclusions. The rest of this chapter will provide a more detailed
chapter-by-chapter overview of the thesis.
Chapter 2: Background
6
This leads into the specific problem addressed by this thesis: the situated
voluntary adoption of Community Displays. Examples are given from the
literature which demonstrate the existence of this problem, together with a
heterogeneity in the nature of the deployment sites in which it has been
observed. A critical analysis of the properties of these sites gives rise to a
characterization of these sites into two kinds of setting: “one-shot” usage
settings and “on-going” usage settings. Following this analysis, the thesis
research questions are then put forward, followed by the approach and
methodology.
Analysis of the findings from this case study extends the understanding of
the social nature of large displays when used by a group in a communal
space. It introduces the concept of user “flow”, describing the pattern of
how locally mobile, peripatetic members move though space; and also
introduces the “honey-pot effect”, which describes the manner in which the
public availability of interaction with the large display enabled passers-by to
oversee and choose to join in, allowing group congregations to form
spontaneously, without the need for any planning or coordination work.
7
These findings are applied and refined in the following two case studies,
where they provide a focus for the analysis of situated voluntary adoption of
Community Displays.
8
Factors found to deter adoption were found to be a fear of social
awkwardness (feeling “on stage”) and the length of the queue, among
others.
The overall contribution of this chapter is the finding that in on-going usage
settings where a Community Display is used intensively or repeatedly over
an extended period of time, the display estate becomes a limited communal
resource, which introduces management issues and conflicts between users,
and without support is likely to hinder adoption and on-going use.
9
Chapter 6: Preliminary observational study of an on-going
usage setting
10
Chapter 7: The adoption of the Dynamo Community Display in
an on-going usage setting.
This chapter details a ten day long observational study of Dynamo V2 in the
college common room. The findings showed that over the course of the
study, the community progressed from initially treating it like a normal PC,
to eventually developing a recognized set of practices, which are detailed in
this chapter.
As found in the previous case study, learning about the system tended to
occur vicariously, and interaction in relation to the Community Display
occurred at different levels of engagement, from peripheral overseeing
through to direct interaction. However, in this setting, since user exposure to
the system spanned a long duration, this did not manifest itself in the one-
shot studies as two prominent thresholds that user needed to cross. Instead,
it enabled many community members to learn about the system gradually
while going about their other daily activities, through the employment of
low engagement activities.
Also, users were observed to engage with the system and each other in a
wide array of different contingencies of use. For example, they would use it
individually or in groups, with or without various kinds of help (e.g. “side-
by-side support”, “back seat tutoring” or “over the shoulder learning”), with
or without various kinds of devices, and so on. These different
contingencies are defined as “entry points”, and it is suggested that they
should be wide-ranging, to enable gradual buy-in and a wide user base; and
that designers should monitor for closed entry points during evaluation, and
endeavor to open them through iterative re-design.
11
Chapter 8: Discussion
This chapter takes stock from the findings of all the user studies carried out
in this thesis, and incorporates them into a cohesive account of the way the
Community Displays in one-shot and on-going settings studied were
adopted and used, addressing each of the research questions and providing
suggestions for the design of future Community Display systems.
The concepts and interaction models put forward for the two settings are
summarised and compared, drawing attention to the similarities and
differences between the two. Finally, to demonstrate the value of the
concepts and interaction models as analytical tools, they are applied in a
post hoc analysis of user-studies from the Community Display literature.
12
Chapter 9: Conclusions and Future Work
One of the important avenues suggested for future research in this area is
the progression beyond the one-shot and on-going setting characterization.
Through a literature review and an analysis of a number of hypothetical
settings, some suggestions are made for other ways in which these settings
vary beyond those investigated in this thesis. These include differences
between “loose knit” and “close knit” communities; and the effect of
“mixed settings” in which different communities with different needs use
the same communal space together.
The thesis concludes that the concepts and interaction models put forward
provide the beginnings of a lingua franca for researchers and system
designers, enabling them to better understand the setting, the social
behaviour and the voluntary adoption that occurs in relation to Community
Displays.
13
Chapter 2
2.1 Introduction 15
2.2 Adoption: the key to the success of any technology 17
2.2 Informal Social Interaction in Communal Spaces 19
2.3 The social nature of large displays 23
2.3.1 Information Dissemination 25
2.3.2 Awareness 26
2.3.3 Shared Point of Reference functions 29
2.3.4 Serendipity functions 30
2.4 Community Displays 32
2.4.1 Noticeboard service 32
2.4.2 Immediate Display and Exchange service 34
2.4.3 Passive Awareness services 36
2.5 Voluntary Adoption of Community Displays 39
2.5.1 Voluntary Adoption in the ‘one-shot usage’ category 46
2.5.2 Voluntary Adoption in the ‘on-going’ usage category 49
2.6 Research Questions 51
2.7 Approach 52
2.7.1 A Strategy for studying a young field 52
2.7.3 Informing design from observational studies 53
14
2.1 Introduction
Figure 2.0 provides a visual overview of its scope of this thesis. Originating
from an interest in Community Display systems, the problem of voluntary
adoption is found to be an important problem in need of investigation.
Therefore this research lies at the intersection of research on adoption, and
research on Community Displays. Its motivation lies in the improvement
Community Displays, with the aim to develop concepts and uncover
pertinent issues for Community Display system designers, to help them
develop systems that actually get used by their target communities.
This chapter will detail the research that underpins and motivates the thesis.
It begins by introducing the problem of technology adoption and how it is a
fundamental challenge for technologies that require voluntary uptake by end
users. The focus of this thesis, situated voluntary adoption, is described.
Next, the general problem space of this thesis is detailed: the support of
informal social interaction in communal spaces. The nascent field of
Community Display research is then introduced, and it is explained how
these systems are intended to provide support within this problem-space.
15
Some underpinning research is then detailed, describing the inherent social
nature of physically large displays.
This then leads into the specific problem-space of this thesis: the voluntary
adoption of Community Display systems. The widespread nature of
voluntary adoption problems is described, referencing a number user
studies from the Community Display literature. The diverse and varied
nature of these studies is discussed. A critical analysis gives rise to six
dimensions on which studies vary, and from these, two types of setting are
characterized: “one-shot” and “on-going” usage. The differences of
voluntary adoption problems between these two settings are described. This
analysis gives rise to a set of research questions that this thesis aims to
address. Finally, the methodological approach chosen to study these
research questions is described.
16
2.2 Adoption: the key to the success of any technology
“New technology intended to support cooperation often risks initial
rejection” (Francik et al., 1991; p. 53)
“Getting a new idea adopted, even when it has obvious advantages, is often
very difficult.” (Rogers, 1962; p. 1)
Adoption of technology by end users is key to its success: even the most
brilliant and beneficial innovation is worth nothing if it does not get used.
Many technologies suffer from delays in achieving adoption. For example,
when Bell Co. attempted to introduce the telephone into America, it was
delayed for years while consumers simply couldn’t fathom its potential
usefulness (Aronson, 1968; Seely-Brown & Duguid, 2000). More recently,
voicemail was reported to have taken approximately 10 years to achieve
widespread adoption (Francik et al., 1991), while the fax machine took
approximately 50 years (O’Brien, 1989). This shows that it is not just
quality or usefulness that determines successful adoption: the above
technologies were ignored for long periods of time, and only recently
became widely used. Therefore, there are other factors at work beyond a
technology’s usefulness. The many-faceted sociotechnical problem of
achieving adoption is the topic of much discussion within CSCW
(Computer Supported Collaborative Work) and its related fields (e.g.
Grudin & Palen, 1995, Rogers, 1962).
17
the most well known is the problem of achieving “critical mass” (Ehrlich,
1987; as cited in Grudin, 1988). Also known as the “network effect”
(Liebowitz & Margolis, 1998), it can be described as a person’s decision to
use a tool being determined by whether other people they know also use it.
Email, for example, would be of no value to a user if they knew no-one else
who used it. Thus voluntary adoption depends not on the decisions of
isolated individuals, but on the emergent effect of interactions between
people in a community.
18
This perspective enables researchers to address questions about how users
react to a system when they use it for the first time, how they then learn
about the system and its features, how adoption takes place socially between
immediate colleagues or friends, and what social roles it develops through
use. To quote Bradner et al. (1999), “[U]nderstanding adoption requires
careful examination of the interactions between technological features and
the social context of use” (p. 139).
19
are not “communal spaces” since usage in these settings is typically
planned, scheduled, and there is little overlap or mixing between different
people engaged in different activities in the same space.
20
educational establishments and members’ clubs as well as workplaces, and
so offers an unnecessarily narrow view. It is for this reason that this thesis
adopts the more general term “communal space”, which, although less
widely used, is recognised by other researchers in the field (e.g. Churchill et
al., 2004; Russell & Sue, 2001; Huang & Mynatt, 2003; Rogers & Rodden,
2003). Other related terms include “social condenser”, a term from
architecture that describes places which serve socialising functions (e.g.
Hughes, 1991); “transitional space”: which describes spaces that people
move through but do not spend extended durations within (e.g. Sawhney, et
al., 2000); and “interstitial space”, which describes spaces between
architectural structures (Tschumi, 1994), as well as various others.
21
members are involved in selecting the material that is displayed, either
actively, by explicitly putting material on the display, or passively, via
sensors and system interpretation of user profiles (see Figure 2.2 on page 43
for some examples). Community displays are widely believed to have the
potential to play a role in supporting community and enhancing “social
capital”, which can be likened to a stock of altruism and friendly favours
which community members develop and exchange with each other through
ongoing social interaction (Grasso et al., 2003; McCarthy, 2003; Zhao &
Stasco, 2002). However, the scope of this thesis is within the support of
informal social interactions rather than the emergent effect on community as
a whole, which is discussed in detail elsewhere (e.g. Resnick, 2002).
22
planned meetings and work activities, and thus their analysis lies outside the
scope of this thesis.
The field of Community Displays is nascent, and much of its research has
taken place during the course of this five year PhD research (2000-2005).
The motivation to develop and study Community Displays may have partly
emerged from the previously-mentioned research in other settings (e.g.
interactive whiteboards in meeting rooms) and the increasing availability of
the technology. However, aside from this technological motivation, over the
past decade a body of evidence has emerged which shows that all large
displays have an inherently social nature owing to their physically large
sizes. This provides a sociological motivation to develop and study
Community Displays. The following section will critically evaluate this
research.
23
displays as being ‘publicly available’ (e.g. Heath and Luff, 1991; Robertson,
2002), allowing the information displayed to be used by multiple people in
cooperative activities. This is a property which has been observed in many
kinds of physical artefacts (e.g. Robertson, 2002; Hughes et al.,1992; Moran
& Anderson, 1990; Suchman, 1987), but is considered to be particularly
prominent in large displays owing to their size. Together with their
situatedness, it is their physical size that gives large displays their special
nature. An analysis of the literature on large displays was carried out,
revealing them to have four main social properties, shown in figure 2.1,
below.
The following section will detail each social property with reference to
studies and theory in contemporary research. It should be noted that they are
deeply intertwined and as such are hard to tease apart into individual
examples.
24
2.3.1 Information Dissemination
Anecdotally, we have all had experience with the use of large displays for
information dissemination. Noticeboards, railway time-tables, road signs,
billboard advertisements and graffiti are all examples, which take advantage
of the size and positioning of the display to visually broadcast information
to people in the vicinity. Bellotti and Rogers (1997) report an account of a
manager in a newspaper publishing company constructing a large
information display by painstakingly copying details each day from their
group project management software system onto a whiteboard in a shared
office, because it “provided an effective public reminder of what was urgent
and needed doing that day” (p. 282). In a study of a team of programmers,
Whittaker and Schwarz (1995) observed a similar scenario. Here, the
programmers chose to use traditional pin-and-paper noticeboards in a
communal space to plan and coordinate their programming projects, instead
of the project management software assigned to them by their company.
This was found to be not due to a lack of technical expertise in using the
software, but in fact due to a preference for the noticeboards. In both cases,
this effortful use of non-digital tools shows how much they valued the
benefits of large displays over and above the desktop computer network and
project management software.
When interviewed, respondents in both studies generally stated that they felt
digitally transmitted information (e.g. email or shared calendars) was
transient and easier to lose, forget about, or overlook, whereas the physical
board created a sense of realness. To quote Bellotti and Rogers: “When
asked why she laboriously wrote up by hand information that could be
readily accessed by everyone on the network, she replied that […] people
had become desensitized to the many email reminders that such software
applications provide” (p. 283). Similarly, Whittaker and Schwarz (1995)
were told by respondents that the desktop digital representations seemed
“dead on arrival” whereas “the board somehow seemed more real” (p.6).
Thus the physicality and the social setting seemed to give the large displays
more impact and credibility than their desktop digital counterparts.
25
2.3.2 Awareness
Robertson goes on to explain that part of the reason for this “bewildering
array” of awareness terminologies is because they derive from different
philosophies and understandings of human perception and interaction which
are embedded in the design of technologies. This in-depth discussion of
awareness is placed outside the scope of this thesis. Like Pedersen and
Sokoler (1997), this thesis takes a phenomenological approach, explaining
awareness by describing its characteristics rather than attempting to describe
how it works. They define awareness as “our ability to maintain and
constantly update a sense of our social and physical context … [which we
achieve in an] apparently effortless manner and without being aware that
26
we do so” (p. 51). In CSCW, the concept of awareness is usually applied to
awareness of other people, work activities, and the artefacts that are used. In
co-operative activities, awareness allows people to understand what others
are doing and thus to plan and coordinate their own actions (Bellotti &
Dourish, 1993).
27
more than that which we immediately attend to: we have the ability to
perceive things peripherally (e.g. “out of the corner of your eye”) as well as
focally. Research in vision psychology details the physiological and
neurological basis for this (Pedersen et al., 1997). This peripheral awareness
provides us with the ability to be aware of multiple things at the same time,
and therefore to participate in multiple activities.
28
driver on the radio, which the DIA (his colleague, the “Divisional
Information Assistant”) notices:
Since the large displays have a large interaction space, this enables people
to be peripherally aware of them across a wide area – they provide a
medium that people can employ to broadcast information to others. Also,
colocatedness fosters peripheral awareness simply in the sense that people
are able to see what their colleagues are physically doing.
The long history of blackboard, whiteboard and the recent popularity of the
interactive whiteboard (e.g. Pedersen et al., 1993; Mynatt et al., 1999)
suggests a genuine value of large displays for co-present cooperative group
activities. The large public interaction space (O'Neill et al., 2004) provided
by a large display allows multiple people to have a shared visual point of
reference, augmenting their conversation. Various studies of users
collaborating on a task found that having a shared visual artefact improves
the quality of their communication (Fussell et al., 2000; Veinott et al.,
29
1999). Research has found that a shared point of reference can help
establish common ground and mutual knowledge, provide a means to
monitor comprehension, and enable effective communication by allowing
deixis, i.e. pointing or other physical references to the visual representation,
rather than needing a full verbal description (Kraut et al., 2002). Jordan and
Henderson (1995) make a similar point in their analysis of large displays:
Other studies have found related limitations with small displays. For
example, Rodden et al. (2002) found that face-to-face interactions in travel
agents are limited because the desktop monitor used by the agent is oriented
away from the customer, preventing them from having a shared point of
reference and reducing the effectiveness of their interaction.
30
opportunities for social interaction because people are more likely to bump
into each other serendipitously (“by happy accident”). In a field study of a
university library, Twidale et al. (1995) observe this happening:
Isaacs et al. (1996) observe that colleagues often meet at high traffic areas,
where there are shared resources such as coffee machines or water coolers.
Also, Perry et al. (1999) found, in a field study of the management office of
a construction site, that large pin-and-paper displays showing project
information served to encourage and support serendipitous interactions.
Studies of serendipitous interactions in workplaces have revealed them to be
short and frequent in nature, and do much to support work-related
collaborative activities (Kraut et. Al., 1990; Kraut & Streeter, 1995;
Whittaker, et. al, 1994). They are also believed to benefit information flow
in an organization (Kraut et al., 1990), and help people learn and adopt the
social conventions and procedures of a community (Suchman & Wynn,
1984). It has even been suggested that they may contribute to the well-being
of the group (Isaacs, et al., 1997), although this is likely to be hard to
critically evaluate.
31
section will show, these social properties have been exploited by designers
of Community Display systems in varying extents and to various different
ends.
However, little is yet known about exactly how these properties interact
when Community Displays are used in communal spaces, and what the
implications are for voluntary adoption. As such, this presents itself as the
first set of research questions for this thesis:
32
interfaces, such as FXPal’s Plasma Poster (Churchill et al., 2003, Figure
2.2b); XRCE’s Community Wall, Intel’s GroupCast and Outcast, and
Carnegie Mellon’s Messyboard (Fass et al., 2002; Figure 2.2f). Other
methods of posting include email, e.g. Plasma Poster, Community Wall, and
Apple NewsLens (Houde et al., 1998); SMS and MMS, such as SPAM,
ECT (Greenhalgh et al., 2004), the Orange™ SMS board (Figure 2.2d), and
Meshbox™ (Anderson, 2003); paper scanners , such as Community Wall’s
usage of Xerox Dataglyph™ technology), and extensions of drag-and-drop
on the Windows™ desktop (Greenberg & Rounding, 2001).
The Opinionizer system (Brignull & Rogers, 2003) and the Dynamo System
(Brignull et al, 2004), reported in this thesis, both offer noticeboard
services. These will be discussed in detail in Chapters 5 and 6 respectively.
33
2.4.2 Immediate Display and Exchange service
34
submitted media, called the “media soup”. Any user, however, can connect
to the display via laptop or PDA, allowing them to take control of it. This
enables them to control the current media on display (e.g. to browse a
website), or take over the entire display with a copy of the screen of their
personal device (cf. “VNC”, Richardson et al., 1998), allowing them to
show their screen to a larger number of people.
35
2.4.3 Passive Awareness services
Carter et al.’s (2002) “Iconic” and “Lexical” displays aim to “make shared
interests visible in order to support conversation and help build
relationships” (p.1). The system is intended to track users interactions with
each other to detect relationships, and monitor the information they interact
with (e.g. web pages) to detect their personal interests. Users’ presence is
detected via infrared communication between their PDA’s (which they are
36
required to carry) and a beacon on the public displays. “Friend of a friend”
information about common interests and colleagues is then displayed in an
abstract manner, either on the “Iconic” display, using abstract icons, or the
“Lexical” display, using brief textual descriptions. However, initial user
studies revealed problems with the design of their visual representations,
giving users problems in making sense of the information.
37
that they both share a common interest in wine (information that had been
previously gathered by the system at an earlier point in time). McCarthy
claims that by learning a little more about each other, this makes them more
likely to have conversations in the future (p. 287), which he suggests can
promote “a sense of community” (p. 306) and “increase social capital” (p.
284). Other examples of Community Display system that offer similar
passive awareness services include MIT’s “Aware Community Portal”
(Sawhney et al., 2001), IBM’s Fishtank Community Display (Farrell, 2001).
38
2.5 Voluntary Adoption of Community Displays
This section will critically evaluate current research relating to the voluntary
adoption problems of Community Displays. In particular, it will draw
attention to the diversity of these different studies, and propose some
dimensions on which they differ.
39
they tended “not to perceive information on which they do not have reason
to focus”, and that they should be “made aware of the benefits of such
displays” to encourage use (p. 3). Agamanolis (2002) reported after studies
of the MIT Human Connectedness Community Display prototypes, “Half
the battle in designing an interactive situated or public display is designing
how the display will invite that interaction”.
Given the young age of the field, there is a great diversity in the research
carried out on Community Displays –there is even little agreement on the
name of the field. Furthermore, little has yet been done to categorize and
understand the differences between the diverse items of research. For
example, Intel’s Proactive Displays (McCarthy, 2004) were designed for
and studied within a conference setting, Churchill et al.’s (2004) Plasma
Poster was designed for and studied within workplace communal areas; and
Borovoy et al.’s Community Mirror (1998) was designed for and studied
within a social “party” event. Figure 2.2 overleaf shows photographs of six
different Community Display field studies.
40
Figure 2.2: Community Displays in a range of different communal spaces (a)
McCarthy, 2004; (b) & (c) Churchill et al., 2003; (d) Orange™, 2002; (e) Russell,
Drews & Sue, 2002; (f) Fass, et al., 2002
41
knit” community, in which each member knows few of the other community
members, or it may be “tight-knit” community, where they know many
others. These terms are used by social network researchers to give a
generalized description of the community interconnectedness (e.g.
Wellman, 1996; Wellman, et al., 2002; Burt, 2000). Other similar
descriptors are used by social network researchers, such as “sparse” or
“dense” knit (e.g. Guiffe, 1999). Other notable dimensions include the
activities normally carried out in the space, and the duration and frequency
people normally spend there. Finally, the systems themselves will differ in
terms of the services they offer, as described in section 2.4. These six
dimensions are depicted in Figure 2.3, below.
42
Intellibadge Public
Proactive Displays AgentSalon Iconic & lexical Community Mirror Community Wall Palimpsest
Display
Sumi & Mase, Borovoy et al., Agostini et al.,
Reference McCarthy, 2003 Cox et al., 2003 Carter et al., 2002 Agamanolis, 2003
2001 1998 2002
Permanence of
3-4 days Few hours Few hours Permanent 1 day
Communal Space
43
Community Wall
Apple NewsLens Community Wall Plasma Poster Blueboard X-cast trio Notification Collage
(different location)
Churchill et al., Russell & Sue, McCarthy et al., Greenberg &
Reference Houde et al., 1998 Grasso, 2003 Agostini et al., 2002
2003 2003 2001. Rounding, 2001
Workplace shared Workplace: coffee
Type of Communal Workplace coffee Workplace Workplace foyer Research lab
kitchen & other room & other Town Square
Space room communal area area & corridors communal space
locations locations
Permanence of
Permanent
Communal Space
Not specified,
Passed by A few minutes while probably consists
Duration &
A few short visits a day during breaks, occasional “passing through” to use shared momentarily many passing through mainly of short
frequency spent in
resources in the space or simply as a thoroughfare. times during space to and from visits (e.g. 5-10
proximity
working day workstation area. minutes) while
passing through
Passing through,
Activity normally Coffee Break, socializing, discussion Thoroughfare to lab shopping,
Thoroughfare
carried out Thoroughfare space socialising in cafes,
etc.
medium or close
Community
Close knit: established, academic research groups knit: an established
Interconnectedness
village community
Immediate Display Noticeboard,
System Services Noticeboard Noticeboard Passive Awareness Noticeboard Noticeboard
and Exchange Passive Awareness
44
An analysis of tables 2.1 and 2.2 show a marked clustering on two of the
dimensions: “permanence of communal space”, and “community
interconnectedness”. These give rise to the one-shot and on-going
characterizations.
To elaborate, table 2.1 shows the clustering within the “one shot” setting
characterisation. All of the examples in this table are temporary events, such
as conferences (McCarthy, 2003; Cox et al., 2003, Sumi & Mase, 2001),
rather than permanent communal spaces, and are typically used by loose-
knit communities. As the examples show, users are exposed to the
Community Display for only a short period – up to a few hours in total,
before the event finishes. A common activity seen in the majority of the
examples is socializing, which may complement other activities, such as
viewing presentations or exhibits.
Conversely, Table 2.2 shows the clustering within the “on-going” usage
setting characterisation. A typical example in this table is a workplace
common room or coffee area. All of the seven field studies listed share the
properties of being a permanent of the communal space used by a tight-knit,
long term community. The spaces are used regularly on a day-to-day basis
by various community members, either momentarily passing through and
using the shared resources, or spending regular periods there, (e.g. up to an
hour), taking refreshment, relaxing and socializing (Houde et al., 1998;
Grasso, 2003; Churchill et al., 2003, Russell and Sue, 2003; McCarthy et
al., 2001; Greenberg & Rounding, 2001; Agostini et al., 2002). Essentially,
the communal space is a part of normal daily community life, and the
Community Display, if successful, becomes integrated into this.
It should be stressed here that the two setting characterizations are not
intended to be a mutually exclusive dichotomy, in which any real world
example must lie in one or the other. Instead, these characterizations should
be regarded as “landmarks” in a multi-dimensional landscape of types of
communal space. This is depicted in Figure 2.4, below.
45
Figure 2.4: The on-going and one-shot settings shown on two defining
dimensions.
46
participants did not realize what the Community Displays had to offer, nor
that the displayed information could actually be a useful adjunct to their
conversations. This example draws attention to the fact that participants
need to know what a Community Display does before they can decide to
benefit from its functionality, and if this is unclear, adoption is going to be
hindered. The participants only had the duration of the social gathering to
learn this and make the decision to use it, otherwise they would have missed
their chance and may never have been exposed to the exact same
Community Display again.
In general, Community Displays for one shot settings tend to have simple,
limited functionality that are designed to be suited to brief, one time
interactions. For example, the functionality of the Proactive Displays is
rather limited – as users get into range of a Proactive Display, it displays
some information relating to them. A user cannot take control of a Proactive
Display, as with an immediate-display-and-exchange service, nor can they
post up any kind of material they like, as with a noticeboard service.
Crucially, while this limits their utility, this simplicity may have made it
easy for passers-by to comprehend. Also, to use an analogy with
“disposable income”, conference delegates are likely to have only a small
amount of “disposable time” available to them during the conference breaks
to learn about the system and decide to register.
47
Community Display systems. (McCarthy, Personal Communication, March
10 2005), which included making the community more aware of their
existence and making registration easier and more rapid.
48
2.5.2 Voluntary Adoption in the “on-going” usage category
49
advertisement to sell their car, three people independently complained and
asked to have it taken down, because it did not fit the accepted themes of
community research interests or light-hearted humour.
It also should be noted here that adoption problems may have been under
reported in both categories – researchers may have worried that reporting
this kind of information would make their system and field study “look like
a failure”. It is plausible that they instead may have concentrated on
reporting the positive aspects of uptake instead. Alternatively, they may
simply not have considered it relevant to the analysis in their report.
50
2.6 Research Questions
The research questions posed for this thesis are summarized below:
51
2.7 Approach
52
This research was therefore carried out in an iterative manner, drawing on
the spiral design approach (Boehm, 1988), which defines the process as
cycles of analysis, creation and evaluation, progressively leading to
improved understanding and design. The critical analysis of contemporary
research carried out in the previous sections of this chapter raised some
questions about large displays in communal spaces in general, uncovered
two common types of Community Display settings, “on-going” and “one-
shot”, and pointed towards some key differences between the two in terms
of the issues they face in voluntary adoption. These were therefore selected
to be investigated in a series of case studies. This approach is called the
“collective case study approach” (Stake, 2000), and it advocates the use of
multiple, heterogeneous cases, because they can lead to “better
understanding, perhaps better theorizing, about a still larger collection of
cases” (Stake, 2000; p. 437). Proponents of this approach also suggest it can
be used to suggest complexities for future investigations, and help establish
the limits of generalizability between the cases (Stake, 1994).
53
appropriate. Any requirements identified would better be
considered as provisional, being part of an iterative design
process, needing testing, prototyping and experimentation
with various options and trade-offs identified by the analysis.”
(p. 264, emphasis added)
Following this point, implications for design in this thesis are specified as
“design suggestions” rather than “recommendations”, and are intended to be
evaluated in future research, rather than to be considered the final word in
design for Community Displays. This approach ties back into
conceptualising the design of Community Displays as a “wicked problem”,
in which there is no single solution, and in which “the problem is only
understood progressively as solutions are developed” (Fitzpatrick et al.,
1996, p. 122)
54
Chapter 3
3.1 Introduction__________________________________________ 56
3.2 The Field Study _______________________________________ 57
3.2.1 Method ________________________________________________ 57
3.2.2 Observations ____________________________________________ 57
3.2.3 Analysis of study findings __________________________________ 66
3.3 Prototyping and evaluating a digital large display __________ 68
3.3.1 Wall-loader prototype system description ______________________ 69
3.3.2 Wall-loader field evaluation _________________________________ 72
3.4 Discussion __________________________________________ 74
55
3.1 Introduction
The case study reported in this chapter involves an initial investigation into
the social properties of a large display when used in a communal space. Its
contribution is to provide an underpinning understanding of the situated
social interaction that takes place in its public interaction space, before the
specific issue of the voluntary adoption of Community Displays is
investigated in the following chapters. It would have been desirable to have
begun this thesis research with a field study on an actual Community
Display system in an authentic real world setting. However, at the time of
this research – early 2001 –the field was nascent, and these systems were
not readily available. For example, the first conference workshop directly
related to Community Displays was not held until November 2002 (O’Hara
et al., 2002). Therefore, it was decided that grounding observational study
would be carried out on another kind of physically large display in a similar
communal space setting, as this would still serve to help develop
understanding in the general domain. As detailed in the previous chapter,
prior research has investigated and revealed some important properties of
physically large displays – they support information dissemination,
awareness, serendipity, and provide a shared point of reference for co-
present group interactions. However, there is still much more to be
understood about how large displays are used in communal spaces by
locally mobile communities.
56
up an established, long term community who had been working together for
at least a year. This chapter details an observational study of the AV team’s
use of their large display, the prototyping of a digital alternative called
“Wall-loader”, and a one-day field evaluation of this tool.
3.2.1 Method
3.2.2 Observations
57
A team of twelve AV technicians from a company called “Audio-Visual
Headquarters” were hired to deal with the delivery, installation and
technical support of AV equipment for each session. This equipment
included data projectors, projector screens, sound systems, mixers,
microphones, video cameras, teleconferencing equipment, flip charts, and
all the cables, splitters, cassettes and battery packs required for these items
to function properly.
- The team
The AV team consisted of 12 people. There was one team manager, who
was in charge of booking the equipment and staff for each event. He was
also the team’s official point-of-contact at the conference, and in charge of
making any financial decisions. The remaining 11 team members were
technical support administrators. They were in charge of setting up the
equipment and offering technical support to help out with problems as they
occurred. The majority of the team knew each other very well, and had been
working together for at least one year. Three of the team members were
temporary employees, hired locally for this conference. Each team member
carried a two-way radio for remote communication.
The team had a large room in the conference centre which comprised their
shared office and storage space. The photographs in Figure 3.1 depict the
team’s resources: (a) shows the flight-cases containing their equipment –
data projectors, over-head projectors, projection screens, amplifiers,
speakers, video cameras and other miscellaneous equipment; (b) shows a
large trolley, located in the rear middle of the room, on which the chargers
for their two-way radios were located. Inside the trolley drawers was a wide
assortment of tools, cables, and connectors; (c) depicts the office area,
mainly used by the manager. Here they had a land-line telephone, laptop,
printer and internet access; (d) depicts the large paper wall display used by
58
the team in scheduling equipment and technical support, which consisted of
a series of columns of A4 printed time-tables pinned to the walls. In front of
the large display was a shelf area on which various coloured pens and
highlighters were stored.
The need to use communal resources regularly meant that there was a
continual flow of team members in and out of the space during the day,
particularly during busy times. The team members can be described as
“locally mobile” (Bellotti & Bly 1996), since they were continually moving
between conference rooms and their shared office throughout the day.
This section details the team’s work practices, revealing the complex nature
of their work. The following section goes on to detail the key role the large
display played in these work practices. The team had two main
responsibilities: pre-staging the equipment, and offering technical support.
59
Pre-staging involved moving the correct equipment to the correct rooms at
the right time, so that when the conference sessions started, the presenters or
workshop organizers had the equipment that they requested. This was a
tricky task, since the team was often short of equipment. For example, a
typical strategy they used was to take a piece of equipment out of a room
where it had been “pre-staged” but was not currently being used, and move
it somewhere where it was needed immediately. They then later on needed
to ensure that they replaced the missing item with an equivalent piece in
time for the start of that session. This resulted in a complex “juggling”
activity, akin to a critical-path analysis problem. Additional to this was the
fact that equipment was spatially distributed through the floors of the
building, which could create a lot of heavy carrying work. They therefore
tried to minimize this by taking into account stairs and long distances. Often
they would store heavy equipment in rooms on different floors for this
reason. This created the problem of keeping track of their equipment,
particularly the expensive items.
The other aspect of their daily activities involved offering technical support
to the conference delegates who had trouble with the AV equipment. This
sometimes involved replacing faulty pieces of equipment, introducing the
additional problem of keeping track of the faulty items and finding
replacements, in turn introducing knock-on effects for equipment
scheduling through the rest of the conference.
The rhythm of the day involved two main deadlines for the team: preparing
and getting the morning sessions running, and then doing the same later in
the day for the afternoon sessions. One team member described the pace of
work as: “long periods of boredom interrupted by quick moments of panic,
as the quote goes”. In the run-up to the deadlines, a large number of last
minute technical problems would be discovered, which the team had to deal
with on the fly. They described this part of the work as “fire-fighting”, and
the individual problems as “fires”, connoting the fast paced and stressful
nature of the work. These problems generally consisted of complaints from
the session organizers and presenters. To quote one of the technicians:
60
“When you’re setting the show up, the people in the room,
their communications are passed on to us, and then we have to
communicate back, trying to find out where the thing went
wrong, and try to solve the problem. And that happens for
about an hour period, generally starting about thirty minutes
before the show. That’s kind of key… You kind of have to see
the whole thing. Its like a theatre show. Half the folks in the
house have no idea what it takes…”
Messages were transmitted from the session organizers to the team either
via a student volunteer runner, the conference organizers office, or directly
to an AV team member. They arrived either via radio, telephone or face-to-
face communication. If the message arrived already on a piece of paper, it
would sometime be stuck up on the wall by the large display. The messages
mainly consisted of equipment problems (e.g. malfunctions, usage
problems), equipment requests (e.g. missing equipment, extra equipment
needed) and occasionally requests to set up unscheduled events. During fire-
fighting, equipment was often moved between rooms in an emergency,
which had knock-on effects for the equipment schedule, which they
continually attempted to keep up to date. However, sometimes non-team
members moved equipment between rooms without notifying them, which
introduced problems. To quote from the interview transcripts:
On some occasions, the same complaint arrived in the office multiple times,
even after it had been solved. For example, in one instance a session
61
organizer panicked at the beginning of the session because their wireless
microphone was causing a buzzing sound. They then told everyone they
could to pass on the message to the AV team: a student volunteer, the
conference organizers, and also an AV team member. In such a situation
when the team received the same message multiple times, they needed to
engage in detective work to find out whether the reported problem was
current and required attention (e.g. “Has it been solved previously?” “When
and who by?”, “Has it re-occurred?”), as show in the following example:
62
manager worked this into an initial schedule, drawing up four or five A4
tables for each day using Microsoft Excel™, printing them out and taping
them up to the wall in vertical strips (one for each day) to make a large
display (see figures 3.1d and 3.2). The font used was 12 point Arial,
meaning that this text could be read only while standing directly in front of
it. Figure 3.2, below, shows how the large display was located in the corner
of the room and was highly visible from all parts of the room.
During each day, considerable effort was made to keep the information on
the large display up to date, showing both the current state of affairs and the
upcoming scheduled work. Team members constantly made amendments to
it, which they did by hand using pencils, coloured pens or highlighters. At
the end of each day, when sheets were typically very cluttered with
annotations, the manager would consolidate the amendments and print out a
new sheets to replace the current ones. During particularly busy periods, the
information on the display became messy and disorganized looking,
occasionally causing team members to have problems in reading and
updating it.
63
The team was asked why they bothered with the extra work of printing,
annotating and updating the large displays, when they could have avoided
this by just using the digital version in Microsoft Excel on the laptop. They
commented that it allowed them to scan-through all the schedules at a
glance. For example, one day’s schedule typically spanned three A4 sheets,
and they could look between them just using their eyes, rather than needing
to scroll with the mouse. The manager also commented that this way he
could easily see what changes had been made without needing to sit down
in front of the laptop. They also jokingly commented that some of the team
members might have trouble using Excel and make a mess of it for
everyone else. Team members were observed to prefer to hold discussions
around the large display, even though they had the capability to do so
remotely, via two-way radio. One team member commented that that
solving complex scheduling problems over the radio could be very difficult,
and it was easier for everyone involved to see the timetable being worked
on.
Figure 3.2 shows the room in use, during fire-fighting before a morning
session. While one team member (left) loads up a trolley with equipment to
take out to a room, another (middle) is retrieving tools and cables from the
drawer unit, in order to go and fix a problem, while another (right) is
checking that the display is up to date, and finding out what equipment
needs to be moved next. The remaining nine team members were out
roaming the corridors, either looking for “fires” to “put out” or in the
process of dealing with one.
While passing by, team members often walked over to the large display and
glanced at it, apparently to check for recent annotations, or to make an
annotation of their own. If a team member passing by the large display
noticed one of their colleagues working on it, they would often stop to see if
their help was required, in solving, for example, an equipment distribution
problem (e.g. “I’m short of a data projector for room 206 in the afternoon
session, where can I borrow one from?”). Similarly, if a team member
working at the large display found they needed help or information, they
64
often requested it from passers-by. It was also observed that congregations
often emerged in this way from a single user solving a problem, into three
or four people working on a problem together in front of the display,
discussing their ideas, gesturing to it and making amendments. A popular
activity during quiet periods was to congregate around the display as a
group, and try to re-plan the movement of equipment for the next session, so
that the least amount of footwork would be required.
Users of the display often used conspicuity in their annotations to grab their
colleagues’ attention. Unsolved problems would be highlighted using
fluorescent highlighter pens, by scrawling question marks and circles to
draw attention to recent changes. In some cases extra sheets of paper would
be stuck up by the large display. For example, in one case a team member
was observed to stick an extra sheet of A4 paper up stating “Need 4 cameras
for 302 PM session” in very large, bold writing, when they were unable to
find any spare cameras according to the information on the large display.
This served to get their colleagues’ attention, and, with the additional
knowledge, solve their problem.
It was noted that during fire fighting, when it was particularly busy, the
large display often became messy and disorganized in appearance, with
many crossings-out, and a “spaghetti” of arrows and annotations crammed
into increasingly smaller amounts of space. This tended to caused the team
problems in reading the display and understanding the current equipment
schedule status, requiring extra conversations and interpretation work, and
occasionally causing errors in deployment.
The large display was a physical hub for team activity within the office. All
team members regularly passed it by while bringing or taking away
equipment, and there would often be team members nearby working on the
equipment or taking refreshment breaks. Usage of the large display often
took place in groups.
65
3.2.3 Analysis of study findings
This study confirms and extends many of the things already learned in
section 2.3 concerning the inherent social properties of large displays. For
example, the team chose to use the pin-and-paper large display instead of an
adjacent laptop containing exactly the same information in a digital form,
even though this required extra editing and transcription work. Analysis
suggests that this was because it offered a number of other benefits which
outweighed the cost of this work. The benefits the team members explicitly
stated were that it provided “at a glance access”, overview of the timetables,
and ease of use. These benefits have been observed in a number of other
studies on large displays, such in Rogers and Bellotti’s (1997) newspaper
publisher study, and Whittaker and Schwarz’s (1995) software developer
study (detailed previously in Section 2.3). Further analysis of the
participants’ behaviours reveals some other important phenomena relating
to the nature of the way the large display was used.
The nature of the continual movement of the team members in and out of
the communal space can be described as flow. The range of shared
resources (equipment storage, office space, and the large display), and its
central location meant that the communal space was the “hub” of this flow,
and was the most likely place for team members to serendipitously bump
into each other, a finding consistent with various other field studies (e.g.
Twidale et al., 1995; Isaacs et al., 1996; Perry et al., 1999), as discussed
previously in sections 2.3.4 and 2.2.
The prominent location of the large display in the communal space was
therefore very sensible, since it made its displayed information readily
available, improving its information dissemination function. The
information shown on the display was an important resource for the team,
and as such, it attracted team members to it regularly as they passed it by.
Group congregations would often spontaneously form around it – which can
be described as a “honey-pot effect”. This term is borrowed from the term
“honey-pot” in human geography, where it describes physical locations,
66
such as national parks, that have attractive features which draw people to
them like bees around honey (Johnston et al., 2000). Here, “honey-pot
effect” is used to describe the phenomenon of the formation of social
gatherings in the vicinity of the large display.
67
However, as well as offering important benefits which gave the large
display a key role in their work activities, the paper-based nature of the
large display also caused them problems, since editing of annotations
involved crossings-out, arrows and re-writing, which became confusing
during peak times. This meant that equipment sometimes became lost or
deployed to the wrong rooms at the wrong times, and the team often held
conversations trying to work out what was written down. While they
aggregated the edits and re-printed the timetables every night, this could not
be done during the day since it was too time-consuming.
68
centered design attention. A number of different formats were experimented
with, using “quick and dirty” user testing on paper-based prototypes,
including “Chernoff faces” (Chernoff, 1973). This is reported in detail in
Rogers and Brignull (2002). The aim was to design a representational
format that allowed to easily interpret the current state of affairs (the
“fires”), and to decide which needed immediate attention. A suitable
representational format for displaying the “fires” was found to be a Gantt
chart representation, since it allowed effective parallel visual comparison
between the items (cf. Rogers et al., 2002; Rogers & Brignull, 2003).
69
Figure 3.3: Photo of Wall-loader in use
70
Figure 3.4: Two snapshots showing part of the Wall-loader display.
Technician icons are superimposed on problem boxes as part of a chart, with location and time elapsed as its axes
71
3.3.2 Wall-loader Field evaluation
One key problem lay in the learning requirements of the system. After the
demonstration, one of the temporary employees commented that. “The next
guy might have trouble learning this… My briefing took 10 minutes!” They
commented that the paper time-tables on the large display hardly took any
explanation, since most people already knew how to read and write on a
timetable. During every event, the team consisted of approximately one-
quarter new temporary employees, so this rapid and easy learnability was
therefore a very important feature of any large display they would use.
72
display estate of an average desktop computer monitor (1024 by 768 pixels:
note that display estate resolution is independent from physical size).
One problem of Wall-loader was its bulkiness: since it used a data projector,
it had a large footprint. As soon as the demonstration was finished, the team
packed it away to make space for their work, so that it would not interrupt
the flow of people through the space. Future re-designs could benefit from
considering the proportions of the destination space and the way it is used,
to select the most suitable large display technology to fit that space. More
compact alternatives to front projection include rear projection using
mirrors, LCD screens, plasma screens, or other forthcoming technologies
(such as electronic ink™ or OLED displays).
73
another way, this flexibility could be described as allowing the community
to appropriate the tool and devise their own style of using it, which may not
be exactly how the designer originally conceptualized it (cf. Dourish, 2003).
Wall-loader overly constrained their use of the display, and prevented this
from happening.
3.4 Discussion
Findings from this field evaluation confirm many of the things already
known about the social properties of large displays, including their support
of information dissemination, awareness, serendipity, and provision of a
shared point of reference for co-present group interactions. This chapter has
described how these properties were effectively exploited by a team of
locally mobile technicians using a large display to support their work.
A key finding from this case study is that interaction with a large display in
a communal location is spatially distributed. Interaction does not take place
just in front of the large display, it involves the movement of people around,
into and out of its public interaction space (O’Neill et al., 2004). This
understanding led to the development of the concept of “flow” to describe
the movement of people in relation to the large display, and of the “honey-
pot effect”, to describe the way in which spontaneous group congregations
tended to form in large display’s public interaction space.
To elaborate, the large display did not just provide publicly available
information in the pinned-up sheets – the interaction of people around it was
publicly available, which provided an important awareness function,
allowing passers-by to “oversee” the usage (Heath & Luff, 1991; Luff &
Jirotka, 1998; Bentley et al., 1992). Overseeing created the opportunity for
passers-by to notice what their colleagues were doing, and stop and help if
appropriate. This enabled the team to have group gatherings around the
display as and when they were needed, without requiring any planning
effort. This can therefore be described as an informal mechanism for
74
coordinating group gatherings, resulting from the public availability of the
display and the interaction in its vicinity.
75
Chapter 4
4.1 Introduction__________________________________________ 77
4.2 Supporting the activity of socializing _____________________ 78
4.3: Designing a Community Display to support socializing
in a one-shot setting _____________________________________ 81
4.4 Opinionizer System Description _________________________ 82
4.5 Field Studies _________________________________________ 83
4.5.1 Study 1: The Book Launch Event ____________________________ 84
4.5.2 Study 2: A Welcome Party _________________________________ 86
4.6 Analysis of findings ___________________________________ 90
4.7 Discussion __________________________________________ 94
76
4.1 Introduction
The type of one-shot setting selected to be studied in this chapter is a social
gathering event – specifically, a party. Other similar one-shot events occur
at conferences, trade-shows and festivals. These are unlike gatherings that
occur at familiar venues, such as a local pub or workplace common room,
because they are special one-off events. The event is short, typically ranging
from a few hours to a day long, the venue is typically borrowed (e.g. a
conference center), and when it finishes, the visitors may never return to the
same venue again. Community members often come together from remote
places for the duration of the event, making up a composition of friends,
colleagues, “familiar strangers” (Milgram, 1992) and unfamiliar strangers.
The event provides them with the ability to meet other community members
in person for the first time or to maintain existing relationships.
77
The findings of the previous chapter show the way a social gathering can
emerge around a large display and facilitate social congregation and
interaction, without the need for planning activities (the “honey-pot effect”).
This suggests that Community Displays may be useful in supporting
socializing activities. Unlike the large display investigated in the previous
chapter which was used for work activities, adoption of a Community
Display is voluntary. As detailed in Chapter 2, the voluntary adoption of
Community Displays is the core concern of this thesis. This chapter aims to
ascertain the role of situated social interaction in voluntary adoption within
a one-shot setting. Drawing on the findings of the previous chapter, the
concepts of flow and the honey-pot effect will be used as an analytical
frame and expanded upon within this context.
78
their personal social networks for information, work and favours, and thus
they play a very important role in their work lives. Nardi et al. (2000) state
“The old adage, "It's not what you know, but who you know," could,
paradoxically, be the motto for the Information Age.” (p. 1) Social capital
theory, described earlier (Section 2.2) suggests that people keep track of the
favours and good-will they pass on to each other, and draw upon it in a
manner akin to trade. Therefore building and maintaining a personal social
network is an activity of high value, even though it can involve apparently
frivolous socializing activities.
One of the problems for an attendee trying to meet new people at a social
gathering event is when the attendee is a new arrival to the gathered
community, and is not well linked within the social network. This means
that the attendee has to start conversations with strangers “in the cold”,
rather than being able to meet them via introductions through existing
colleagues. This activity can be difficult, and can hinder a person’s
socializing activities and thus their integration into a community (Borovoy
et al., 1998). Erving Goffman (1963), in his book on behaviour in public
places, suggests that a widespread ritual which enables strangers to meet is
via an introduction from a third party. Clark’s (1996) theory of Common
Ground is relevant here – it describes how the mutual knowledge, beliefs,
and suppositions shared between two people are prerequisites for interaction
to occur, in that all communication depends on an understanding of one’s
audience. Clark goes on to claim that the establishment and development of
common ground is crucial to the development of acquaintedness. However,
the development of common ground can be a difficult and time consuming
process.
It is for this reason that name badges, ice-breakers and other such activities
are used at social gathering events (cf. Newstrom & Scannell, 1995;
Epstein, 2001), to facilitate the development of common ground.
Researchers such as Krauss and Fussell (1990), and Borovoy et al. (1998),
have postulated whether technology could be a useful aid for this. As well
as Community Displays, a large number of other kinds of tools have been
79
developed in recent years for this purpose, which fall under the banner of
“social software” (Tepper, 2003). For example, during 2004 alone, over 100
web-based “social networking” and “friend of a friend” services have
appeared, such as Friendster™ and Orkut™, which enable people to make
new acquaintances on-line by being able to view and correspond with their
friends’ friends (Boyd, 2004). With the emergence of mobile technologies
such as WAP and MMS on mobile phones, a number of similar mobile
services have emerged, such as Jabberwocky (Paulos & Goodman, 2004),
MamJam™ and Dodgeball™, which enable people to discover if friends are
nearby (e.g. in the same area of town) and meet up with them, or to meet
new people with shared interests who happen to be co-located (e.g. in the
same café). Other bespoke mobile technologies offer similar services, such
as nTag™, SpotMe™, and in the research domain, Intellibadge (Cox et al.,
2003) and Thinking Tags (Borovoy et al., 1998).
At the outset of this work (early 2002), only a small amount of research had
looked at the use of Community Displays to support the activity of
socializing at social gathering events. For example, Borovoy et al. (1999)
developed “Community Mirror”, a Community Display which showed
aggregated user information, based on the use of infra-red badges called
“thinking tags”. Sumi and Masse (2001) developed and tested “Agent
Salon”, a community Display which used small cartoon avatars representing
nearby users, that engaged each other in speech bubble conversation about
their shared interests. Since then, a number of others have appeared,
including the “Intellibadge” Community Display (Cox et al., 2003),
McCarthy’s “Proactive Displays” (2004), MIT’s “Palimpsest” (Agamanolis,
2003), as well as efforts by Carter et al. (2002) and Agostini et al. (2002).
80
4.3: Designing a Community Display to support
socializing in a one-shot setting
Drawing upon the analysis of the nature of socializing in Section 4.2, the
problem of supporting socializing between strangers was selected to be
explored through the prototyping of a system called “Opinionizer”. This
was another use of the vertical prototyping approach, in which a prototype
is designed to support one aspect of a problem in a fully functioning
manner. Socializing between strangers is also a particularly important aspect
of the problems of socializing, since strangers do not have the linkage of
common colleagues or friends. Section 4.2 shows that it is crucial for people
to establish common ground when attempting to hold a conversation (Clark,
1996; Borovoy et al., 1998). As such, “establishment of common ground”
was taken to be a motivation for this Community Display’s functionality.
Specifically, the Community Display was designed to provide a publicly
available surface displaying topical themes for discussion, and the means
for adding opinions to the surface. For this reason, it was named
“Opinionizer”. Opinionizer was not intended to be like a chat room or
discussion board, through which participants hold on-line discussions, but
instead it was intended to be a public resource which people could
contribute to, and use to initiate face to face conversations. Emphasis was
placed on ensuring the themes were topical and relevant, and that the
interaction was lightweight and fun. Drawing from the literature review of
one-shot Community Display studies in Section 2.5.1, some initial design
suggestions were made:
81
• Low effort interaction: ensure that interaction involves minimal
effort and time from users (McCarthy, 2004; Cox et al., 2003)
These initial design suggestions were taken on board and used in the
specification and development of the following system, which was coded
using Macromedia Director™.
82
Figure 4.1: A screen shot from Opinionizer taken at the book launch
party (study 1)
83
4.5.1 Study 1: The Book Launch Event
The setting of this study was an evening book launch party during the CHI
2002 conference, located in a large room at the top floor of the Minneapolis
Hilton Hotel. The room was octagonal, and offered a long buffet table, eight
round tables with chairs (seating for 48 people) and a bar. The Opinionizer
Community Display was set up on a raised platform next to the bar, as
depicted in Figure 4.2.
The size of the Opinionizer screen was approximately 6’ wide by 4.5’ tall.
The text on the screen was legible from approximately 16’ away, while the
screen itself was visible from anywhere in the room. The laptop served as
the input point, and was positioned on a table close to the display. The
projected display mirrored the laptop’s, providing an identical but larger
view. A helper stood by to explain to people what Opinionizer was and how
to use it. The party lasted for 2 hours, during which approximately 300
people passed through, milling around the area near the display (usually on
84
their way to and from the bar). Opinionizer was deployed for the full 2 hour
duration of the party. Activities were observed and observational notes were
taken. Video was not used since the event organizers expressed privacy
concerns.
- Observations
At the beginning of the party, the first people to arrive tended to congregate
near the buffet table, a few meters from the Opinionizer screen (in the centre
of the room). This left an empty space around Opinionizer. At this point,
few people came forward to try out the Opinionizer, and when people were
invited to come over and try it out, they seemed rather shy of doing so. Fake
opinions were added by the researcher to encourage interaction. However,
this alone did not seem to encourage usage.
Figure 4.3: The crowd around the Opinionizer input point (laptop),
pictured bottom right.
As the room filled up, a congregation developed around Opinionizer, and
more people started to try it out. People were observed watching and talking
to others using it. Groups of people were also seen talking about it, (i.e.
85
looking at it, gesturing, laughing) without actually interacting with it, as
shown in Figure 4.3 above. The congregation seemed to create a honey-pot
effect as found in the previous chapter, whereby the more people who
interacted with it, the more other people followed suit. This seemed to result
in an increase in rate of contributions to the Opinionizer over time, as shown
in Figure 4.4 – the comments began spaced out, and progressed to be closer
together as the event wore on (Total number of contributions: 38. Data
gathered from system log).
As time passed, it was observed that people were able to interact with the
Opinionizer without needing any explanation from the helper, either
learning by observation, or being shown or told how to use it by other users.
Some of the users became proponents of the system and were observed
going to fetch friends to view and try it.
The setting for the second study was a welcome party for postgraduates
entering a school at a university. The same Opinionizer system was used but
86
with different themes, intended to be more relevant for the community of
newly starting postgraduate students at Sussex University. An example was
“What do you think of the food at Sussex University?” The projection
screen was the same size, with the same legible and visible distances
(legible from approximately 16’, visible from anywhere in the room,
occlusion from crowding not withstanding). As shown in figures 4.5 and 4.6
overleaf, the room was a different shape, being rectangular. Approximately
150 people passed through the space during the study. Several had met
before and formed groups but there was a sizeable number of people who
were there by themselves. The party lasted for about five hours, with the
Opinionizer active for the first two and a half of those. Two video cameras
were deployed near the display to record group behaviours and track
people’s movement. They were placed high up to be unobtrusive as
possible, so as not to affect people’s behaviours. As well as video recording,
two assistants helped capture other data: one person was employed as a
roaming interviewer, another as a observer, taking notes and photographs.
Interview questions are shown in Appendix 1. A third person stood on-hand
next to the input point, to help participants and explain the system if asked.
Figure 4.5: Photo of the set-up used at the welcome party event
87
Figure 4.6: Floor-plan of the set-up at the welcome party
- Observations
Similar to the book launch event, at the beginning of the party people
congregated some distance away from the Opinionizer, collecting food and
beverages from other tables. One person commented, later "Nobody really
knew what it was when they came in and there was a whole kind of fear
because it was something new". This can be seen in Figure 4.7, where many
people are standing some distance from and face away from the Community
Display. As time wore on, the room filled up. People began to congregate
around the display, paying it attention, watching others using it and talking
about it between themselves. This is shown in figure 4.8 (p. 94).
88
In total, out of approximately 150 people who passed through the space
during the study, approximately 60 interacted with Opinionizer, i.e. 40%
(deduced from system logs). During the study, 23 people were interviewed
(see appendix 1 for questions), selected at random from the room to gain a
cross section of people who had used and had not used Opinionizer (mean
age 29; 14 males, 9 females). Over half of those interviewed made positive
comments about Opinionizer, and stated that they felt comfortable
socializing around it and talking with others about the opinions displayed.
Over three-quarters of respondents stated that they had socialized with
people they hadn’t previously met while standing near Opinionizer. Hence,
it proved to be a highly effective ice-breaker.
89
bring down the tone of the comments, and encourage facetious or even
offensive comments. One participant interestingly commented: “It's
perhaps safer [to input remotely] but not all the fun of going round the table
together".
90
People would often then move to “focal awareness” activities. This involved
standing in view of the Community Display, looking at it in more detail, and
socializing with people in relation to it (e.g. having conversations about the
displayed themes or comments). To quote one interviewee from study two:
“I didn’t see people using it at first, but I did see people standing around it
so I stood beside and watched it for a while”. Transition from peripheral
awareness to focal awareness activities would often take place when other
conversations had finished, and people were apparently looking for
something interesting to occupy themselves with. Also, people who arrived
at the event alone often progressed into focal awareness activities on their
own, and were observed to often engage with others who also were looking
at the Community Display.
From focal awareness activities, people often then moved into direct
interaction activities – moving to interact with the Community Display’s
user-interface via its mouse and keyboard. Although ultimately only one
person could add comments to the display at a time, group efforts were
91
observed where the adding of an opinion was cooperative and where
different individuals took turns to be “the driver” and enter data via the
keyboard. People engaging in direct interaction activities typically had an
audience, who were watching their interaction.
The honey-pot effect observed has a parallel with that observed in the AV
team case study, in which the team members would oversee their colleagues
using the large display and would stop to help if needed. In both cases, the
public availability of the display and the interaction around it provided a
mechanism for enabling serendipitous group congregations. However, in the
case of Opinionizer, socializing was the aim rather than solving work
problems, and the congregations were much larger, up to approximately
eight people rather than two or three, as in the case of the AV team.
92
“multifocused” in nature (p. 91). The flow of people in the space can be
described as mingling, whereby people moved between “foci”, i.e. clusters
and resources such as food and refreshments (rather than moving directly in
and out of the communal space as in the AV team case study). As well as
this mingling, a flow was observed in the movement of people towards and
away from the Community Display.
Another issue was the delay involved in the queuing – some interviewees
stated that they found this off-putting. This is consistent with predictions
about the nature of one-shot settings- that attendees are only there for a
short amount of time, and are likely to spare only a small window of time to
93
interact with a Community Display system. Also, given the many other foci
of activity (other conversation circles, food and drink tables, etc), the
Community Display is in direct competition for the attendee’s time.
4.7 Discussion
As noted in the second study and shown in Figure 4.8, activities in relation
to the Community Display can be described in three categories: peripheral
awareness, focal awareness and direct interaction – each representing a
different level of engagement with the Community Display. Goffman
(1963) suggests this is common in public gatherings: “…a differentiation is
sometimes found among full fledged participants and various grades of
onlookers” (p. 18). This has a parallel with the findings from the AV team
case study, in which team members could go about their own activities, yet
be peripherally aware of what their colleagues were doing, owing to the
public availability of their actions.
94
it. These thresholds are defined here as the “threshold to attention” and the
“threshold to interaction”. These concepts of flow, activities and thresholds
make up a model of interaction with the Community Display, as shown in
the schematic diagram in Figure 4.9, overleaf. The term threshold is used
here in a psychological rather than physical or spatial sense. Psychological
“hurdles” seem to occur at these two thresholds, with notable differences in
their nature.
95
system while they are involved in another activity, such as conversing with
other people. Then, having chosen to, they will give the display more
attention, and perhaps move to stand in a position with a clearer view.
Once they are attending to the Community Display, they may then cross the
threshold to interaction, deciding whether to actually interact with the
Community Display. Essentially, the two thresholds represent two levels of
granularity of detail. At the threshold to attention, the person will judge the
system based on the broad-grained details available to them at a distance
and from brief glances, essentially asking themselves “Is it worth taking a
closer look?”. Then, at the threshold to interaction, the person has
committed to spend some time and effort finding out more about the
Community Display. Here they will be able to discover more fine-grained
details, such as “vicariously” learning the details of interaction with the
system by observing others using it (Reber & Reber, 2001), including the
length of the interaction, and exactly what the system does and the benefits
it offers. Here they can also develop a detailed expectation of what the
experience of using it would be like, including its comfort, or conversely,
social awkwardness.
96
Similar strategies are also used by shopping psychologists in placing
information displays, advertisements and plinths in shops (e.g. Underhill
2002).
Vicarious learning, i.e. learning by observing others (Reber & Reber, 2001),
was a key mechanism by which onlookers learned these things, and it
allowed the spread of usage to become “self perpetuating” to an extent.
Here the suggestion for designers is to ensure that the system has a visual
user interface that is readily observable. However, this observability also
brings the problem of on-lookers’ fear of social awkwardness and
evaluation apprehension, which was noted in both studies. Opinionizer’s
clear and simple function of requesting users to write a one-sentence
opinion may have reduced this fear – it can be postulated that more complex
functionality and user interfaces could have been a greater deterrent.
In all, the findings of these studies confirm the applicability of the design
suggestions posed at the beginning of this chapter, which were used to
inform the design of Opinionizer: present immediately obvious functionality
and benefits, ensure rapidity of learning, low-effort interaction, and
relevance of content so that members of the local community find it
interesting. Returning to the concept of disposable time, a person’s spare
time and effort in a one-shot setting is limited, and can be likened to a
97
currency which they will only spend if they see themselves benefiting over
and above the other alternative things they can “spend” it on, such as
conversing with other people elsewhere in the social gathering.
98
Chapter 5
99
5.1 Introduction
The previous chapter investigated the nature of adoption and usage of a
Community Display in one-shot “social gathering event” settings,
uncovering some pertinent design issues. This chapter will present the
design process carried out on Dynamo, a Community Display for on-going
usage settings, i.e. places like workplace common rooms and coffee areas,
or public cafes and recreation venues which have regular patrons. Voluntary
adoption problems in on-going usage settings are different in nature to those
in one-shot usage settings, such as those reported in the Opinionizer studies.
Rather than being a foreign and temporary artefact, when adopted in an on-
going usage setting a Community Display can go on to become integrated
into community life, and involved in social activities, norms and practices.
This on-going process of adoption is often referred to as “appropriation”, as
defined by Torpel et al. (2003):
100
For example, Agostini et al. (2002, p. 711) describe Community Wall users
as having re-invented the Community Display during a longitudinal field
study. They observed that since the system did not provide search facilities
in Italian (the national language at the deployment site), users found this
frustrating. Through time, members of the user community took on the role
of “Italian search engine” using their own knowledge to help others to
search for information and to suggest related items (p. 701). Also, Churchill
et al. (2004) found that in a 14-month study of Plasma Poster, social norms
started to emerge that specified the styles of acceptable use, community
members were observed to complain about when broken. These implicit
rules share a parallel with explicit rules such as FAQs or “codes of practice”
often used by established virtual communities, such as USENET or web
discussion boards (Preece, 2000). As well as social norms, emergent social
activities can also give rise to practices – these are activities which become
established and understood customs within a community (Torpel et al.,
2003). In contrast, in one-shot settings, the process of appropriation is
curtailed due to the short duration of user exposure to the system. While
users may engage in creative social activities, they cannot develop
community practices and norms, because they do not use the system
repeatedly and regularly with their peers.
101
session to review its suitability, and in a laboratory study, to test its user-
interface and interaction models. The findings of these studies are then used
to recommend some iterative re-designs for the system, which were
integrated into Dynamo Version 2 (V2).
102
Izadi et al. (2003) detail and expand greatly upon these principles and their
underpinning motivations.
With existing Community Displays for on-going settings it can be seen that
designers typically aim to provide a range of open-ended and flexible
functions which users can employ in various different ways. In essence, this
is design for appropriation. For example, the designers of Blueboard
(Russell and Sue, 2003) provide a range of “immediate-display-and-
exchange” functions (see Section 2.4.2), such as the means to enter text,
display images, arrange them, annotate on top of them using the
touchscreen, and share them in various ways. This functionality is flexible
and can be used in a “freeform” manner. For example, it could be used in a
group meeting to support the discussion of architectural plans, it could be
used as a brainstorming tool, or it could be used for socializing, e.g. by
someone who bumps into a colleague and decides to show them some
photos from a recent trip. Community Wall (Grasso et al., 2003) provides a
range of noticeboard functions (see Section 2.4.3), allowing users to post up
103
various media types and annotate them in different ways (e.g. via web
interface or touchscreen). Again, this could be used in any number of ways,
depending on what the community themselves choose to use it for. For
example, it could be used to display a firm’s stock and productivity figures
in the foyer area for clients to see. Alternatively, it could be used to display
humour and jokes for socializing in a less public shared kitchen area.
Although the functionality is the same, the social role given to the
Community Display is very different.
The technical goal specified for Dynamo to offer a surface for general
“display, sharing and exchange” purposes is consistent with the need to
provide open-ended and flexible functionality for an on-going setting. For
example, Dynamo will be appropriable as a noticeboard, a presentation
screen, an information display, a surface for note-taking, or even an ice-
breaking socializing tool, (among many other possibilities), depending on
the users’ needs.
In the Opinionizer studies it was found that vicarious learning, i.e., learning
by observing others, was a key process by which community members
104
learned about Opinionizer. Drawing upon this, it is suggested that Dynamo
should offer a visual user interface that is easily observable by onlookers, to
facilitate the vicarious learning process. For example, keyboard shortcuts
and small gesture-based commands will be avoided, and instead iconic
palette based interfaces will be used.
105
of acquiring a particular type – the findings of the Opinionizer studies
suggest that requirement of time and effort can outweigh the perceived
benefits of the system and deter people from using it. Furthermore, some
community members may not even own a personal device, and if they do,
they may not carry them at all times. The system should be inclusive to such
people since they are potential users. In summary, a wide range of personal
devices will be supported in order to not exclude any potential users.
Furthermore, personal devices will be conceptualized as an optional benefit,
rather than a pre-requisite to interaction with Dynamo.
106
It was also hoped that simultaneous multi-user interaction will allow open-
ended and flexible range of group uses. For example, a single small group
could gather round and each of them have a mouse and keyboard.
Alternatively, individuals could arrive alone, and use the Community
Display independently, or in pairs (etc). Or instead, a group of 12 people
could gather in the room, and congregate in groups of four, with one mouse
and keyboard on each table, among many other possibilities. In other words,
this interactional and spatial flexibility could provide new opportunities for
social use, and thus encourage a honey-pot effect.
The AV team case study showed how team members often highlighted text,
and used large text sizes to draw attention to particular items on their large
display. This served an important function of disseminating information to
passers-by at a distance, and also of indicating the recency and importance
of these large items. To interpret this into a design decision for Dynamo,
user-controlled scaling of media windows (e.g. photos, videos) and font
sizes will be supported, since it is flexible function that community
members could potentially utilize to disseminate information and indicate
importance.
Access control was a question posed in the initial Dynamo project group
discussions: should media (e.g. graphics or documents) placed on the
Dynamo surface be “lockable” so that other users can’t move it? Should
media be specifiable as “private”, and accessible only by the owner, or
“public”, and accessible by anybody? Findings from the AV and
Opinionizer case studies suggest that an open access, laissez-faire model
can be effective, as this keeps the system simple. This thereby facilitates
learning and adoption, while access control, if needed, can be socially
negotiated, and even supported through the development of community
107
norms and practices. For example, participants in the AV team knew when
it was and was not appropriate to alter someone else’s annotations on the
large display. Also, participants in the Opinionizer study tended not to
manipulate other people’s comments, because they knew it was not socially
appropriate. Recent research in the area also implies support for this user
experience principle – for example, in Dix et al.’s (2004) user studies of the
“Hermes” situated door display (for the booking of shared rooms in an
office environment), they found that abuse of the open-access model tended
not to occur, because on-lookers provided a function of “neighborhood
watch”, to quote:
Similarly, O’Hara et al. (2004), in user studies of another similar office door
display system, found that:
108
5.3 Dynamo Version 1 System Description
Drawing upon the project group’s technical goals and the user experience
principles put forward in the previous section, the Dynamo system was
designed and implemented. It should be emphasized that the resultant
system is just one instance of many systems that could potentially have been
created on the basis of these goals and principles: it is exploratory research.
• The device hub: a USB (or Firewire) hub where users’ mobile devices
are plugged in and accessed as media storage units
109
Figure 5.1: Schematic diagram of the Dynamo system
110
5.3.1 The Dynamo Display Surface
Interaction points are the means by which users interact with Dynamo.
“Base” interaction points are provided for users who do not have a
Windows laptop, while those with their own laptop can utilize it as a
111
“mobile” interaction point if they wish. In a normal set up, three base
interaction points are provided. Any combination of base and mobile
interaction points can be used. Additional interaction points can be added
(or taken away) at any time during usage.
If a user has a laptop which they can put on the Local Area Network (via a
wired or wireless connection), they are able to run the Dynamo Laptop
Application. This allows the desktop of the users’ laptop to tile with the
Dynamo surface, so that when a user moves their mouse to the top of the
screen, it appears on the Dynamo surface as a colour-coded pointer. Users
are able to drag and drop media files from their laptop desktop to the
Dynamo surface, and vice-versa.
112
USB Flash Drive (termed “Pen Drive”). Designed to be held on a key-
ring, these devices serve as small storage devices. They plug directly
into a USB port without requiring a cable.
MP3 Player: Many serve the double function of music playback and
audio recording.
Digital Still Camera: Many have the capability for recording video and
audio clips as well as still photos.
Multimedia Jukebox: This new genre of device are for storage, viewing
and playback of multimedia, including video, images and audio.
PDA: As with smart phones, many PDAs have capabilities for taking
photographs, video, audio, document editing and storage.
Windows Laptop: When connected via LAN and running the Dynamo
Laptop Application, the user is able to drag and drop media to and from
their desktop and the Dynamo surface.
Figure 5.2: Dynamo Compatible Devices
113
5.4.1 Feedback Session
- Study Setting
Dynamo V1 was deployed in the foyer area of the hotel at which the
conference was being held. A diagram of the floor plan is shown in Figure
5.3a below, and a photo of the layout is shown in Figure 5.3b, overleaf.
114
Figure 5.3b: Annotated photograph of the system layout
- Findings
Overall, Dynamo was well received and the feedback showed that it had
promise in being an effective Community Display. However, problems in
ownership and access to media and devices caused concern for users. This
section will detail these findings.
115
approximately 30 interacted with the system; some as individuals and others
in small groups, while others watched. Some users commented that they felt
“on stage” or under the spotlight. One commented, “I hope I don’t break it
in front of everyone”. This social awkwardness and apprehension of using
technology for the first time in front of an audience is consistent with the
analysis made on the Opinionizer studies in Chapter 4 (cf. Rogers &
Brignull, 2003).
However, despite these benefits, a notable problem was observed in the lack
of representation of ownership and access rights of devices and media on
Dynamo’s surface. Of all the study participants, five users tried their own
pen-drives on Dynamo, browsing or sharing their personal media. In one
interaction, a pair stood together with an interaction point each, one of them
with a pen-drive plugged into Dynamo. When the owner tried out Dynamo,
opening and showing some documents to his colleague, the colleague
started browsing through the disk further. The colleague found an
application form on the disk for a grant proposal, which initiated a
conversation on the topic. In this case the pen drive owner did not mind his
colleague “nosing” through the contents of his disk. However, another
observer fervently complained to us about the possible breach in privacy
that could have occurred. Using the above example, he stated that that the
item found could have been something private, such as a secret job
116
application or a compromising photo which he would not have wanted his
colleagues to see.
The aim of this second study was to observe the effectiveness and suitability
of Dynamo’s user-interface and interaction models, when used intensively
by groups. Specifically, a co-operative task was given to groups of
volunteers which they were asked to carry out using Dynamo. This study
was carried out on Dynamo Version 1.
- Participants
117
study, they were given a consent form, explaining that participation was
voluntary, the data collected was confidential, they could ask any questions
they liked, and that they were free to leave at any time without needing to
give a reason. Participants were paid £5 each.
- Study design
The given task involved the group creating an “interactive poster” on the
Dynamo surface, effectively preparing it to be a community noticeboard for
a hypothetical community who would use the space. This involved them
browsing through and choosing media from a mixed selection which they
were each given on different pen-drives as they arrived at the study. Each
participant’s selection was mixed in terms of relevance, aesthetics and
physical proportions, to engender discussion and decision-making. The task
was to spatially arrange the media on the Dynamo surface, with the option
of creating text and of displaying web pages if they wished. The scenario
given to them was that they were to create a poster which aimed to
encourage prospective students to come and study in Brighton. They were
told that the poster was to be displayed in a public space during an open
day, and passers-by had the option of interacting with the media and
downloading items to their own USB disks if they wished. The content was
chosen to be relevant to our participants (who themselves were based in
Brighton) and to be familiar enough to encourage a relaxed atmosphere and
discussion.
Four groups of four participants were studied. Within each group, two
participants were given wireless mice and keyboards (connected to PCs
running the telepointer application), one was given a laptop running the
Dynamo laptop application, and the fourth was given no device of their
own, and was instructed to contribute through discussion and by sharing
devices with the others. Three pen-drives were given to three of the users,
each containing a different selection of multimedia (images, video clips,
PowerPoint™ and Word™ documents). A further selection of multimedia
118
was placed on the desktop of the laptop, allowing the laptop user to drag
and drop this material onto the Dynamo surface. Each session lasted 45
minutes. Before each session, 15 minutes were spent talking the users
through a detailed set of instructions explaining how to use the system.
Users were seated at two tables arranged in a wide “V” shape in front of the
Dynamo display, to allow users to be able to easily look at each other and
Dynamo. They were videoed from two angles – one camera was pointed at
the group, to record their physical behaviours, while another was pointed at
the Dynamo display, to record their behaviour on screen. A still of the video
capture set-up is shown in Figure 5.4. After the sessions, participants were
debriefed and interviewed about their experience of using Dynamo.
Figure 5.4: a still of the video capture set-up used in the laboratory
study (Dynamo is shown picture-in-picture, top left)
- Findings
Overall, although each of the four groups successfully completed the task,
they often experienced problems with the “open access” access control
model, in that they often inadvertently got in each others’ way,
119
manipulating each others’ windows and working in each others’ space. This
section will detail these findings (also reported in more detail in Izadi et al.,
2003).
In general, when users started out using Dynamo, many of them had
difficulties in understanding the etiquette of using the communal surface.
Some users were very polite and asked the group before they opened any
documents on the public surface, while others were more forward and
dominated the display estate with less regard for others’ needs. Once they
became more familiar with Dynamo, users often helped each other and
worked together without using much discussion about how to cooperate
using the interface. This is shown in Vignette 5.1, below, where “P” and
“R” assist “L” who is having trouble grabbing and moving a window
containing a photo of an old lady from an overlapping pile of windows:
Vignette 5.1:
L: Can anybody move this one? [has trouble grabbing title bar
of a window] I’m trying to get the old lady to the front. [takes
hand off mouse momentarily]
P: [immediately grabs an occluding window and moves it out
the way, helping L]
R: [immediately grabs another occluding window and moves
it out the way – leaving L’s desired window visible.]
All the groups successfully completed the cooperative task, and expressed
satisfaction with the end result. However, in all the groups, when interaction
at the interface was most intensive, it was common to see users hindering
each other by overlapping, closing or “stealing” each others’ windows; or
using space on the surface that another user had considered their own. In
Vignette 5.2, below, ‘J’ closes a window that ‘R’ was using, and apologizes;
while in Vignette 5.3, ‘H’ and ‘M’ bicker over the control of two windows.
Vignette 5.2:
J: [makes some space for his new window by closing one in
its way.]
R: Oh, you closed my Events!
120
J: Oh sorry R, Did I?
R: That was mine that was!
J: Oops! I do beg your pardon!
Vignette 5.3:
H: See I can’t move your thing, look!
M: You can...
H grabs his window while he is talking and moves it while M
is trying to use it
M: [angrily] No!
M [resizes his window to cover H’s]
H: Oi! [He grabs the offending window and moves it out of
his way]
To quote one user in the debriefing interview: “I think it’s actually quite
chaotic unless you work to some sort of plan, because basically it’s like
having a desktop but harder because you can’t control it, you know, and
when you put something there, somebody’s going to put something over
what you’ve just put down.” Even though they were trying to cooperate,
users found it hard to stay aware of what the others were doing and, more
specifically, the windows and spaces the others were using.
The two studies detailed above showed that the user-interface and
functionality of Dynamo V1 shows promise for group usage. In the
feedback session (Section 5.4.1), participants gave positive feedback about
the system and a honey-pot effect was observed. In the laboratory study
(Section 5.4.3), users achieved cooperative activities during a group task,
and were generally satisfied with the end result. However, some important
problems were also uncovered with Dynamo V1’s support for the
management of sharing and ownership of resources. In summary, the
feedback session showed that Dynamo V1 needed to support the ownership
of personal devices when plugged into the system, to avoid privacy
concerns. The laboratory study showed that Dynamo V1 needed a
mechanism for managing the sharing of the communal resources on the
121
display, i.e., the real estate and the media items, to avoid conflicts between
users, particularly during periods of intensive use. This requirement created
a tension with the original user experience principle of “allowing open
access to all resources” in order to keep the interaction with the system
simple and thus facilitate learning and adoption (detailed in Section 5.2.1).
In other words, how could the system be kept simple to use for first-time
and novice users while also providing these additional features of access
control, which would add complexity to the system? The following
subsections detail how this tension was addressed in the redesign of
Dynamo, i.e. version 2 (V2).
To briefly revisit the origin of the design tension of ease-of-use and access
control, in Section 5.2.1 a number of pieces of evidence were found to
support the design of an “open access” access control model into Dynamo,
since they showed that social protocol could be sufficient in providing
122
access control rules (Specifically, the case studies in Chapters 3 and 4; Dix
et al., 2004; O’Hara et al., 2004). Furthermore, an “open access” model
enabled the system to be easier to learn and use, and for this reason such a
model was designed into Dynamo V1. However, the laboratory study
reported in Section 5.4.3 shows that during intensive group usage, conflicts
between users became highly evident, in the form of accidental overlapping,
closing or stealing each others’ windows or space. In order to address the
tension between providing ease of use while also providing a means of
access control, facilities for access control were designed to be optional,
defaulting to the open access model for first time and novice users, while
offering advanced access control features via registration for those users
who desired them. In addition, the screen real-estate was re-designed to be
significantly larger, as detailed in section 5.5 and figure 5.6.
The mechanisms for access control designed for Dynamo V2 were called
“Carving”, and “Parcels”, inspired from other multi-user systems. For
example, many multi-user document sharing systems allow users to “check
out” documents they are working on and check them back in when they are
done, to prevent two people from accidentally working on the same item at
the same time (e.g. Macromedia Dreamweaver™). Also, multi-user shared
document editing tools which allow users to simultaneously work on the
same document (e.g. SubEthaEdit™), often use a technique of colour
coding to visually demarcate the section each user is working on, to prevent
accidental clashes. This inspired the design of the “carving” mechanism,
detailed in the following section. Also, various operating systems provide
means to administer access control of folders, allowing users to aggregate
media and store it iconically on their desktop, serving to provide an
organizational function, and freeing up valuable screen estate. This inspired
the development of the “Parcels” mechanism, also detailed in the following
section. Additionally, the size of Dynamo’s display estate was substantially
increased to address this problem, to provide more space within which
multiple users can position their media.
123
5.5 Dynamo Version 2: System Description
This section describes the changed and additional features introduced into
Version 2. Repetition of the unchanged features stated in Section 5.3 is
avoided for the sake of brevity. A detailed video demonstrating the features
of Dynamo Version 2, is available in Izadi et al., (2003b). Figure 5.5
overleaf shows screen shot of the system in use. Following this section,
“Dynamo V2” will generally be referred to as just “Dynamo”, for the
purpose of brevity.
124
Figure 5.5: Screenshot of the Dynamo (V2) system in use
125
- Increased screen estate
Figure 5.6: comparing the screen estate size of Dynamo V.1 with the
intended size of Dynamo V.2
- System registration
To register with the Dynamo system, the user needs to own a compatible
personal device, such as a pen-drive, a digital camera or a laptop.
Registration with the system can be carried out away from the Dynamo
surface, via a dialogue box on Dynamo or by running a command-line
registration application on any Windows™ computer. This deposits a
configuration folder onto the user’s device: this includes an XML document
specifying their full name and their password (encrypted). Their chosen
personal icon is also placed here, as an image file. This is customizable by
the end user at any time.
- Logging in
A registered user logs into Dynamo by plugging in their device, and then
clicking on their personal palette and entering their password into the small
dialogue box which appears adjacently. The system then associates their
126
chosen interaction point (wireless mouse and keyboard) as owned by that
user. Non registered users are able to use the system at any time as
anonymous “guest” users. Unlike registered users, any window they open is
publicly accessible, and they are not able to create or belong to any carve-
regions (see later for details), since the system cannot identify them as
unique individuals.
Users can take copies of media items by grabbing a window’s content and
dropping it onto their personal icon on their personal palette. Similarly,
users can give other users copies of media in the same way.
127
- Addressable Parcels
Parcels provide users with a means of storing media on the dynamo surface
for future access. Parcels are like “folders” seen in normal GUI
environments (such as Microsoft Windows™), except that they have an in-
built and readily accessible visual access control tool. Parcels have two
states – iconic and open. When iconic, a parcel is shown as a parcel icon
(shown as either “sealed” or “previously opened”, as depicted in Figure
5.8), and has a textual label. When open, a parcel is shown as a media
viewer window, which allows users to view the contents of the parcel and
open each item. It also offers facilities for setting access control, sealing the
parcel, and deletion. An access control pane offers simply a list of registered
users, which can be selected via the mouse, as well the option to make the
parcel “public”, i.e. publicly accessible by any user.
Figure 5.8: Parcel icons, shown sealed (left) and opened (right).
128
right of the carve, and dropping it onto the target user’s personal icon on
their personal palette. This causes the joining user’s icon to appear with the
owner’s in the top left of the carve, as depicted in Figure 5.10, overleaf.
Access can be provided for users who are not present at the time by putting
the carve’s key icon into a parcel and addressing it to them. Access can be
revoked by dragging their personal icon out of the carve and dropping it
onto the Dynamo surface. A user who does not have access is shown this by
their cursor developing a “shield” icon around it, as depicted with the green
cursor in the red carve region, in Figure 5.9. A carve region can be
discarded by clicking the “discard” button in the bottom left of the carve
region.
Figure 5.9: Detail of two carve regions. (from Brignull et al., 2004)
129
Figure 5.10: Adding (right cursor) and revoking (left cursor) a user
from a carve region (from Izadi et al., 2003)
5.5 Discussion
This chapter has presented the design process carried out for Dynamo, a
Community Display for on-going usage settings. Evaluation of Dynamo V1
revealed it to be promising as a Community Display, yet it suffered from a
problematic lack of facilities for the management of sharing and ownership
of the communal resources on the display, i.e. personal devices, the media
displayed, and the screen real estate itself. Analysis suggests that such
facilities will be important for usage in on-going settings. To draw an
analogy of Dynamo’s resources with the tables and chairs in a typical on-
going communal space, these are both intended to free for anyone to use as
they wish. However, for this communality to work, there is also a need for
signifying temporary ownership. With tables, chairs, and personal
belongings, people mark their ownership physically, with their bodily
presence. The mechanism of carving in Dynamo is intended to be the digital
equivalent of marking ownership, in the absence of physicality.
130
members to become concerned about ownership. Indeed, such lack of
concern is consistent with the design goals aimed for with Community
Displays for one-shot settings. However, in on-going usage settings, where
owned, non-trivial content may be used (as observed in the field
evaluation), and community members may invest more time and effort in
usage of the display (as observed in the laboratory study), issues of
managing sharing and ownership come to the forefront of the users’
concerns.
131
Chapter 6
132
6.1 Introduction
This chapter describes a one-week long preliminary observational study of
an on-going usage setting – the common room of a 6th form college. The
aim of this study was to investigate its suitability as a potential deployment
site, and provide a grounding description of the community’s existing
activities, practices and flow, to offer a point of comparison with the effects
of Dynamo after it is deployed.
The site selected was the common room of Blatchington Mill 6th form
college, used by a student community. An initial visit suggested this site
would be suitable for the following reasons: it was used regularly by its
community members; the space was large enough to house Dynamo; the
members showed an interest in the capabilities of the system; and they were
willing to be videoed during the study.
The common room had 300 square meters of floor space, and contained
enough tables and chairs to seat 44 people. As detailed in Section 6.2.4, the
common room contained a number of resources, including lockers, pigeon
holes, vending machines and so forth. The rhythm and flow of usage is
detailed in the following section (Section 6.2.2).
The community was composed of 150 students aged 17-19 years old. All
members were studying for their A-level exams, and had been doing so for a
duration of 1-2 years. Community members were, in general, on familiar
terms with their peers, although some cases were noted where students did
not know each others’ names. Most students treated the common room as
their base on the college campus, and used it as a place to hang out, work,
socialize, read, listen to music, and locate friends. Approximately 50
individuals spent most of their spare time in the common room: from an
133
hour and a half, to three hours in total a day. The majority used it for less
than an hour a day, to access resources such as the tea bar, their lockers, or
to hang out while eating meals or having coffee with friends.
All students shared the same general timetable structure. Specifically, this
meant that they shared a twenty minute morning break at 10:50 am, and an
hour long lunch break at 1:10pm, during which time a tea bar in the
common room was open, and students could buy hot food and drinks, which
drew a flow of students into the room – at these times, the room was at its
fullest. Also, lesson change-overs occurred every hour, at which time many
students would return to the common room to use their lockers and have
passing conversations with their friends. During these change-overs, the
room would fill and empty rapidly – within a period of 5-10 minutes.
Different students had different timetables, owing to the fact that they were
all studying different combinations of subjects. They also often had long
gaps between their timetabled lessons, time they would often spend in the
common room. Some students enjoyed spending time in the common room
so much, that even if they had no more lessons that day, they would stay
there until the end of the day at 3:50 pm. Other streams of flow occurred in
which students accessed other communal resources in the space. As shown
in Figure 6.1, these included vending machines, a photocopier, a
noticeboard, their pigeonholes, and the deputy headmaster’s office which
was often used for student meetings.
134
Figure 6.1: Floorplan of the common room
The common room was used by the community for a range of different
activities, as depicted in Figure 6.2. Figure 6.2a depicts a student reading
135
literature quietly for their coursework, while on the other side of the table,
another student has left their folder and pencil case at the table, in the midst
of working on an essay. Figure 6.2b depicts a gathering around a table,
where one reads, two eat their lunch and another has their work on the table.
In the background someone is putting on a CD, as shown close up in Figure
6.2c. Mixed in with these different activities is typically number of
simultaneous threads of informal conversation. These threads would move
and interweave as people joined in and left, reflecting a continual
undercurrent of socializing and mingling activities. Figure 6.2d shows a
student giving a musical performance on their guitar. Other similar
performances involved students telling stories and jokes, and playing music
on the stereo, such as newly released albums. The common room would also
be used as a convenient location for group work, for example, students
working on group projects or proof-reading each others’ essays.
Overall, many of the activities carried out in the common room had the feel
of performance for the benefit of onlookers – the common room was a place
to see and be seen. For example, mobile phone conversations were often
held loudly; students would often announce text messages to their friends
136
that they had received or were sending; and the act of entering would often
be carried out with a flourish, a loud “hello” and handshakes or hugs. This
resonates with Goffman’s (1969) sociological analysis of social life being a
“theatrical performance” involving performer and audience.
137
stating that fancy dress was optional and everyone was invited. On the left
hand side, the pin board is predominantly used by the staff to display
“official” notices about lesson cancellations and room changes, among other
things.
Figure 6.3b also shows the use of the walls as a surface for displaying
notices (on the left is a request to keep the room tidy, on the right an
advertisement for another evening social event). Figure 6.3c shows a
makeshift cardboard noticeboard constructed by the student committee (a
group of students involved in organizing events and representing the
community in staff meetings). This makeshift noticeboard was evidently
constructed for lack of available existing display surfaces, and was used to
display a list of people who had not paid for an upcoming theatre trip. Once
paid, their names were crossed off, providing a public record.
Also shown in Figure 6.4 is a magazine left on one of the tables – students
tended to leave magazines out for others once they had read them.
Magazines and other artefacts were often used as conversational props, i.e.
providing a resource for conversational topics and a shared point of
reference. Other artefacts used as conversational props included CD cases,
text and photo messages on mobile phones, and digital cameras. The
138
following vignette is an example of this practice – specifically, it involves a
group of students passing around a digital camera and chatting about the
photos. Figure 6.5, overleaf, shows a series of video stills from the event,
with corresponding descriptive text.
139
(1) Cathy & Penny (middle) talk about
and laugh at the photos on the camera,
while Becca (right) studies. Mary (left)
notices their gathering and
approaches…
140
The small screen of the digital camera poses some limitations to their group
viewing activity – which they remedy with the “work around” of passing the
camera around (e.g. frame 3), leaning in (e.g. frame 2), and holding the
camera out (e.g. frame 5). During the 4 minutes and 22 seconds of the
vignette clip, the camera was moved and the people re-arrange themselves
22 times- an average of approximately once every twelve seconds, which
may be an unnecessary amount of ‘re-arranging’ work if instead a
Community Display like Dynamo was used. However, there may be other
benefits to using the camera, such as privacy of visibility within the group.
Frame 7 shows how the sharing of the digital photos had to be deferred for
two days, since the camera alone did not provide this function, and the
owner did not know how to carry out the CD-recording procedure without
help. This also suggests that Dynamo may be suitable in this setting, since it
allows sharing and exchange to be easy and spontaneous.
This discussion of technology usage leads onto a survey carried out on the
community’s usage of digital technologies, reported in the following
section.
141
Figure 6.5: Personal devices regularly used at college, as indicated by
survey respondents.
In rank order, the most popular to least popular devices were: non-mms
mobile phones (60%), followed by mms (photo-capable) mobile phones
(35%), digital still cameras (28%), recordable CDs (26%), zip disks (21%),
mp3 walkmans (16%), floppy disks (14%), pen-drives (9%), laptops (7%) ,
digital video cameras (7%), and PDAs (2%). Of particular note here is the
prevalence of physically small pocket-sized devices, of low value –
typically under £100. This is likely to be due to the fact that students had
little money to spend on expensive hardware, and traveled long distances on
foot each day, therefore needing their bags to be light. This data also
suggests that Dynamo would be particularly suitable for this community
since many of them regularly used personal devices that were compatible
with Dynamo.
The survey also asked the students to select from a list which digital
information sharing activities they regularly undertook. Results of this
question are shown in Figure 6.6, below.
142
Figure 6.6: Digital information sharing activities regularly undertaken,
as indicated by survey respondents.
In summary, the most dominant practice was mobile phone voice calls, at
95%, followed by textual emails (91%), SMS (91%), emails with
attachments (88%), instant messaging (53%), MMS (picture, video or
sound) messaging (28%), web discussion boards (19%), photo-sharing
websites (e.g. shutterfly.com) (9%). The rarest practices were Bluetooth file
sharing (5%) and use of shared drive or ftp servers (5%). This shows that
the respondents engaged in a range of digital information sharing activities,
which implies a further potential suitability of Dynamo to this community.
6.3 Discussion
To summarize the findings of this preliminary observational study, the
common room was found to have many of the hallmarks of an “on-going”
communal space. It was used by an established community and members
were generally familiar with one another, having studied together and used
the space for a period of 1-2 years. They used the room on a daily basis, and
a subset of approximately 50 used the room for over 1.5 to 3 hours a day.
The room was used for informal social interaction (predominantly
socializing rather than work-related), and contained a number of resources,
143
contributing to its centrality and offering opportunities for serendipitous
interaction. In general, the common room was observed to be a comfortable
place in which people passed time and interacted with others (Oldenburg,
1989) for the purpose of enjoyment.
144
Chapter 7
145
7.1 Introduction
In general, the findings show the progress of appropriation over the course
of the deployment, detailing how the community progressed from complete
naivety regarding the system and treating it like a normal PC, to gradually
developing a shared understanding of the usefulness of the system for their
needs, in the development of activities, practices and norms. Practices they
146
developed included using Dynamo as a prop to entice social interaction with
peers; using it as a stage to give entertaining performances to large groups
during breaks; using it as resource to support socializing and mingling in
“multifocused gatherings” (Goffman, 1963); and engaging in policing and
tidying activities to maintain it as a communal resource that benefited the
majority. Vignettes describing these practices are analyzed in detail.
The vignettes reveal many parallels with their prior practices. For example,
prior to Dynamo, they would give performances, e.g. playing guitars or
telling jokes; they would socialize using magazines and digital cameras as
props; and they would maintain the room’s communal resources (e.g. tables,
chairs and the stereo) through policing and tidying carried out by particular
community members. This emphasizes how they appropriated Dynamo to
meet the needs of their own community setting, i.e. to do familiar things in
new ways, and this shows the importance of providing flexible and open-
ended functionality to enable them to do this.
Two community members were very keen and adopted the system on the
first day, acquiring and registering pen-drives as soon as they could.
However, a majority of the community “bought in” (i.e. invested interest,
time and dedication) to the system and its related practices gradually and at
their own pace. In general (though not exclusively) they would begin by
engaging with the system in a predominantly peripheral manner, learning by
overseeing, while doing other things in the locality. Initially, they would
“dip in” to Dynamo-related group interactions on occasions, and then, over
time, the length and regularity of these occasions would increase. Also,
within these interactions, they would progress from a predominantly
147
conversational role, to gradually using interaction points and devices more
and more often. Furthermore, they would often get support and help from
their friends, in the form of “over the shoulder” learning, side-by-side
tutoring and “back seat” tutoring, which was facilitated by Dynamo’s multi-
user functionality.
The structure of this chapter is be as follows. First, the study details are
reported, including the arrangement of the room, the instructional materials,
and the data collection approach. Then the study findings are reported,
including an overview, patterns in the system log data, and a series of five
vignettes. An analysis of these findings is then reported, and finally, the
discussion section describes the contribution of this chapter as a whole.
148
7.2 Study Details
7.2.1 Duration
Two digital cameras and a number of pen drives were provided on a free
loan basis. Students could simply sign out items and borrow them as they
pleased. Also, pen drives were offered for sale to students who wished to
buy them.
149
Figure 7.2: Floorplan depicting the spatial configuration of Dynamo.
150
7.2.2 Participant briefing and instructional materials
As well as the students, the teaching staff were introduced to Dynamo, and
were informed of the possibilities of using Dynamo for student learning and
group activities. Three teachers designed activities involving Dynamo and
advocated them to their students as voluntary activities for their spare time.
The photography teacher asked students to scan their portfolios and put
them on Dynamo to get critical feedback. She also created a public parcel
containing work of a number of famous photographers, and questions about
technique and style, which she hoped would stimulate verbal discussion and
annotation using Dynamo’s text note facility. The Drama teacher created a
page of web links relating to the drama students’ coursework on Romeo and
Juliet, and posted it on the Dynamo surface in a parcel. This was done with
the hope of students browsing the pages, learning useful information and
discussing the material in groups in their spare time. The geography teacher
also advocated use of Dynamo for group web browsing of course-related
materials, and reminded them of this possibility a number of times.
Interestingly, these teacher-advocated activities failed and were not engaged
in by students, as reported in Section 7.4.1.
151
7.2.4 Participant consent
Permission was gained from the college allowing us to observe and record
the students’ use of Dynamo. This blanket agreement was possible since the
college held the legal position of “in loco parentis”, i.e. acting as the
students’ parents while they were at the college. As a further courtesy to the
students, they were also given participation consent forms (see appendix 2
for more details). Only 2 students expressly did not wish to participate in the
study at all. A further 5 students agreed to participate but requested that any
video and photos of themselves remain confidential and unpublished. At the
same time as the consent forms were distributed, a disclaimer form was also
distributed, which students had to sign in agreement in order to participate.
This stated that they agreed not to use Dynamo in an anti-social or illegal
manner, and that if they did, they took full personal responsibility for their
actions. Copies of these documents are available in appendix 2.
The “screen camera” was positioned half way down the side of the room,
pointing towards the screens. This mainly recorded the screen activity and
152
the users’ behaviours in the immediate vicinity. The “behaviour camera”
was pointed from the side of the plasma screens outwards towards the users
and the room behind. Finally, the mobile camera was hand-held and was
moved or zoomed around the room in order to record the detail of any on-
going user interactions.
Each day, 9 hours of video was captured from three simultaneous angles,
producing a total of 270 hours of video. This video was captured digitally,
and the different camera angles were synchronized using Final Cut Pro™.
Three months were spent transcribing the video footage into a written report
for each day of the study. Using Jordan and Henderson’s (1995) Interaction
Analysis approach, these reports were analyzed according to a set of defined
foci. These included chunking the reports into events, locating the
beginnings, endings, and segmentation within these events, the rhythm and
periodicy of daily activities, spatial organization of interaction, and trouble
and repair in these interactions (Jordan and Henderson, 1995). In addition to
Jordan and Henderson’s foci, “flow” and the “honey pot effect” were also
used as foci, drawing upon the findings of this thesis so far. This led to a
condensed set of vignettes (reported in sections 7.3.4 – 7.3.8), which
provide insights into the key issues uncovered by this analysis.
In addition to qualitative data capture, log data was also captured by the
Dynamo system, recording the occurrence of feature usage during the study.
Specifically, this log data detailed high level interactions, such as plugging
in or out a device, logging in or out, dragging media onto or off the surface,
creating, moving and resizing windows, parcels, notes and carves, and use
of palettes. Low level interactions, such as telepointer movements, were not
recorded.
Finally, a post-hoc survey was carried out after the study, in order to
ascertain their perceptions of Dynamo and their opinions of its impact on
their daily lives. This was distributed to all 150 community members in
their pigeon-holes, and 39 responded.
153
7.3 Study Findings
The findings are reported in the following manner. First, an overview is
provided, to give a feel for the findings as a whole. Then, the log data is
analyzed, giving an understanding of the progression of feature usage over
the duration of the study. Some aspects of the post hoc survey are also
reported here to show the uptake of Dynamo and the progress of registration
over time. Then, a series of five illustrated vignettes will be detailed.
Finally, the findings are condensed into a set of key principles.
154
analysis about the important role of “low engagement” and peripheral
interaction with Dynamo is carried out in Section 7.4.1.
It was noted that that the common room became progressively busier
through the study. Feedback from the post hoc survey tallies with this
observation. In response to a question asking about Dynamo’s effect on life
in the common room, selected participants responded:
Quantitative results from the post-hoc survey also indicate some broad
patterns in participant behaviour. The graph in Figure 7.4, below, shows the
timing of first time usage among survey respondents. The average first time
usage was on the third day (S.D. 2.40), although the majority (14) was on
the first day, as a result of the briefing session then. There was a wide
spread in this data (indicated by the standard deviation): participants
155
continued to try Dynamo for the first time from day 1 right up until day 9 of
the study. In total, 21 students registered with Dynamo, of which 16
responded to the post hoc survey.
Figure 7.4: Graph showing the occurrence of first time use and system
registration (data from post hoc survey responses)
156
7.3.2 Analysis of log data
157
(Legend: White=0, light grey = 1-200, medium grey = 201-400, dark grey= 401-600, Black= 601+)
Figure 7.5: Aggregate usage through the study (Brignull et al., 2004).
158
The patterns shown in Figure 7.5 indicate a rhythm of usage that followed
the overall daily timetable, with usage peaking around morning break and
lunch break. The log data also captured the occurrence of usage of different
media types over the course of the study. Figure 7.6, below, shows a bar
graph depicting the amount of media displayed on the Dynamo surface. The
most popular media types throughout were images (shown in turquoise)
followed by video (shown in maroon), reflecting the popularity of activities
relating to use of digital cameras (both those provided and the students’
own). Internet connectivity was not available in the first week due to college
network issues, but in the second week it was fixed, which explains the
sudden upsurge in HTML content in the second week (shown in yellow). It
is interesting to consider the popularity of community generated media
(photos and video from cameras), in comparison to other kinds of media,
including web-sites in the second week.
159
keeping, and which were blurred or poorly framed. This accounts for the
large amount of images and videos displayed on the surface, but not
downloaded to personal devices.
Another interesting facet of the log data was the patterns of feature usage
over time. Figure 7.7 shows the parcels and carve usage over time – two of
the most advanced features which allow ownership, storage and access
control. For parcels this was the number of times a new parcel was created,
or an existing parcel was opened. For carving this was the number of times
a user carved a region and added a window (or set of windows) to it.
160
Figure 7.8: Adoption of the “parcel” and “carve” system features
(Brignull et al., 2004).
As the graph in Figure 7.8 shows, carving (in blue), was used more
frequently in the first week than in the second, where it dropped to a
consistent low. Together with the observations of user behaviour, this can
be interpreted meaningfully: in the first week, carving was typically used
exploratively, as users familiarized themselves with the interface, and as a
mechanism to play and socialize with others – by either carving over
another user’s active window or carving over the free space on screen to
deny access to others. This unanticipated and playful usage helped users
strengthen their familiarity with this feature. By the second week carves
were used in more targeted ways, when users found a genuine need to
control access to shared content such as notices and images. In comparison,
the adoption of parcels (shown in red in Figure 7.8) started low, and showed
a progressive increase in use as time went on.
In order to find out more about whether parcels were being used for actual
viewing rather than being opened and closed playfully or exploratively, data
was aggregated to show the average length of time media items were
displayed, comparing media within parcels to media sourced outside (e.g.
from a personal device or a palette). The pattern exhibited by image media
files is particularly apparent, as shown in the bar graph in Figure 7.9.
161
Figure 7.9: bar graph showing the increasing use of parcels to display
images (Brignull et al., 2004).
The bar graph in Figure 7.9 shows that in the first couple of days, images
were displayed on the surface in a fairly ephemeral manner. The duration of
image displays on the surface steadily increases in the first week. This is
because students had filtered out the popular photos and tended to keep
them on the surface for longer. However, even at its peak these photos did
not remain on the surface for long (less than 1 hour). Towards the end of the
first week, students were often observed being frustrated when they returned
to Dynamo to find that the pictures they had put on display had disappeared.
This seemed to encourage students to experiment with parcels. Initial
evidence of parcel usage is seen in the first week, while in the second week,
the adoption of parcels increases rapidly. This represents a development in
the community’s understanding of Dynamo, moving from conceptualizing it
as being just “like a PC”, into a communal surface for storage, organization
and asynchronous sharing. Associated with this was the development of
practices associated with the use of parcels. For example, following a group
session on Dynamo in which a group had viewed an array of media, users
learned to put the most popular items into a parcel for later viewing by other
parties. Similarly, the practice of “catching up” emerged, whereupon
arriving in the room, people would browse the new parcels to see what new
and interesting things had been going on in their absence. Contained in the
162
parcels was often photographs of students out of college on their breaks, or
their activities from the previous evening or weekend, and thus this “catch
up” activity provided an awareness function.
The video data and observational notes from the study were analyzed using
Jordan and Henderson’s (1995) Interaction Analysis approach, (see Section
7.2.5) to create an index of notable events and social behaviours that took
place each day of the study. From this index, a selection of five vignettes
were taken (shown in Figure 7.10), chosen on the basis that each provided a
rich and insightful description of the different kinds of practices engaged in
with Dynamo.
163
Figure 7.10: Overview of vignettes
Vignette (c) (figure 7.10c) shows a typical example of low engagement with
Dynamo. The room is busy, and usage of Dynamo is in the background,
164
while the majority of people go about other activities such as studying or
talking between themselves, only being involved with Dynamo through
low-engagement interactions such as peripheral observation and shouting
comments.
165
7.3.5 Vignette (a) enticing social interaction
Vignette (a) took place on day 8 of the study (10.30am), and is detailed in
Figure 7.11, overleaf. This vignette illustrates the use of Dynamo as an
audio-visual conversational prop in enticing socializing, i.e. using it to
tempt people into a conversation, to furnish content and provide common
ground. It is also an example of a planful use of Dynamo - Alan (the
individual) collected a selection of media from his home computer in
advance to show his acquaintances at college. Having some spare time, and
seeing some familiar faces at the back of the room, Alan prepares some
media on the Dynamo screen, and then initiates a conversation with them
about it. His choice of media is also of interest – he plays the group a series
of tracks of electronic music he composed himself at home. He gets positive
feedback and, as a result of then putting his tracks in a public parcel, the
tracks are re-played repeatedly over the following few days, are talked about
by various people, and are downloaded 5 times. As well as disseminating
knowledge of Alan’s work and skills, this may have given Alan a sense of
satisfaction, popularity and kudos within the community.
Finally, a honey-pot effect occurs while Alan shows his cartoons, as people
arrive and sit down to watch his activity. Someone joins Alan and strikes up
a conversation about the cartoons, and then progresses to interact with Alan
on Dynamo.
166
(1) Alan arrives alone and sits at the front table (left).
He plugs in his pen drive, and browses some of his
media, opening it on the screen. He then glances to the
group behind him (right).
(2). He turns to the group at the back and calls over:
“Hey, Donny, my home-made music”. He starts an
MP3 playing. Anna comments: “wow! It’s amazing!”,
and Don says “You made that?” – Alan replies “On
fruity loops… and cubase.” Don goes over and stands
next to him, appreciatively miming a drumming action
with the pen in his hand.
(3) Alan opens another tune “This one is the newest
one of mine”. Anna comments “I love it!”. She goes
over to stand next to Alan and Don, and appreciatively
claps along with the music.
167
7.3.6 Vignette (b) concurrent group use for socializing
Also of note in this vignette is the distribution of expertise. The two most
novice interactors (pictured far right in the blue t-shirt and far left in the
grey shirt) are paired up and share an interaction point with a more
experienced user. Finally, there is the involvement of the community
members in the background. As shown in frame 3, a group sits drinking tea
and chatting behind the Dynamo group, as do other group of people out of
frame. They can see Dynamo being used and thus learn about the practice of
using it in groups for fun socializing. Also, during the crossword game
(frame 4 onwards), people in the background join in, shouting out answer
suggestions from time to time.
168
(1) Seated left to right: Joe, Rich, Mike, and Bob (also
Colin and Charlotte, out of frame on the left). Colin, Rich
and Bob have interaction points, which they share with the
people sitting next to them. Two simultaneous
conversations occur: Mike and Bob talk about iPods, while
Colin, Joe and Rich talk about mountain-biking, and look
at Rich’s own website detailing his mountain-biking
hobby.
(2) [Mike] “What else shall we look at?” [Bob] “Oh I
know!”- Mike and Bob move on to looking at a site about a
cartoon from their youth (the web browser window
pictured left)..
169
7.3.7 Vignette (c) low engagement group interaction
Vignette (c) occurred on Day 3 of the user study, in the morning (9.45am),
as detailed in figure 7.13, overleaf. In this vignette, usage of Dynamo was in
the background while most people studied or sat in clusters around tables
talking quietly, and only becoming involved in Dynamo interactions in
passing from time to time. It is the nature of this low engagement,
background participation that is of particular interest in this vignette.
Around the room, people sit in circles around tables, either working or
chatting. They are generally oriented away from Dynamo, preoccupied with
what they are doing. Nina decides to put on some music from her walkman
(frame 1), and she goes over. She asks a question loudly and Anna, who was
quietly working in the background overhears and shouts “yep!” in reply.
This helpful involvement required minimal effort or disruption to Anna,
enabled by her peripheral awareness of the environment.
Another example is when Peter helps Nina open her media (frame 4 in
Figure 7.13), there is very little attention from other people in the room –
they are all busy, going about their own activities. However, when he puts
on her song, people visibly demonstrate their peripheral level of
participation – one girl shouts out a request to “turn it up”, while others
express their appreciation physically, by bobbing their heads and dancing
along with the music in their chairs.
The point being made here is that engagement in activities around Dynamo
occur at a number of different degrees along a scale, from low engagement,
to high. As cited previously, Goffman (1963) suggests this is common in
public gatherings: “…a differentiation is sometimes found among full
fledged participants and various grades of onlookers” (p. 18). The
implications of this low engagement, peripheral interaction is discussed in
detail in Section 7.4.
170
(1) Nina decides to put on a song from her personal
MP3 player. She has never connected her MP3 player
before. She asks if she is plugging it into the right
place, and someone shouts “yep” from the background.
171
7.3.8 Vignette (d) opportunistic use in a transition
Vignette (d) took place on day 6 of the study (12.10pm), during a change-
over between lessons. As depicted in Figure 7.14a, it shows how Dynamo
was sometimes used for rapid interactions during brief windows of
opportunity, such as the five minutes between lessons or before the bus
arrives. It also shows how Dynamo was given the role of an open forum in
which public involvement was expected and welcomed.
At the end of the interaction, Peter puts the files in a parcel like Alan in
vignette (a), another example of the practice of leaving a public archive of
media from a group interaction as a gift to the wider community. Days later,
Peter and Charlotte return and conclude their interaction. Together they
cooperatively drag the selected files from the parcel to Charlottes’ pen
drive.
172
Figure 7.14a: panoramic view of vignette
(2) Charlotte (pictured standing) and the other girls (out of shot)
wait in the area by the door, in view of Dynamo from a wide
angle. [Peter] “Wait! I’ll show you! Wait stay there and I’ll show
you… Are you ready?”
(4) Peter logs in and finds the photo from his pen drive. [Peter]
“Are you ready for this Charlotte? … Are you ready for this?”
Peter opens the picture up and enlarges it, in place of where
Heather’s was. [Sally] “What is he wearing?”
173
7.3.9 Vignette (e) individual performing to a large group
Vignette (e) took place on day 7, at the beginning of the lunch break
(1.10pm). As depicted in Figure 7.15a, this vignette emphasizes how the
community as a whole benefit from the system without them all needing to
be active Dynamo users.
As shown in Figure 7.15b, the room is busy since it is a cold day. Leo
arrives in the room with some friends and word goes round that Leo is going
to show some videos on Dynamo. After Leo starts his interaction, no-one
else touches an interaction point, and it becomes a clear “presenter-
audience” style of interaction. He selects media that is entertaining to the
audience and provides awareness about community members, particularly
information that is normally hidden, e.g. people’s activities in the holidays
and on weekends, and in one case, photos of a community member who had
not been seen in college for months. The audience talk between themselves
during his performance, often laughing loudly, and shouting things out to
him or the room in general. This demonstrates audience participation, yet
the level of effort and investment on their part is very small in comparison
to Leo’s.
174
Figure 7.15a: panoramic photo of vignette
(3) Leo plays a series of home videos showing him and other
community members “hanging out” on holiday (which he
edited at home to include credits and music using Windows
Movie Maker™). In between each video he gives a
humourous introduction. During the show people laugh and
chat between themselves.
(5) There is now talk of going for a cigarette break. Leo logs
out and the audience disperses. Later on that day, Leo
removes his carve to make space for other uses, but leaves up
the publicly accessible parcel containing all the media he had
shown.
175
7.4 Analysis of findings
Upon initial deployment of Dynamo, it was a new and foreign artefact, but
over time, the community laid claim to ownership of the system. They used
it for socializing and recreation activities and completely rejected all of the
work-related activities advocated by the teachers, such as using it as a
discussion forum for Geography A-level related web sites (refer back to
Section 7.2.2 for details). One telling example of this developing sense of
community ownership was on day 6 of the study, when a teacher put up a
notice advertising a concert he was playing in on the following weekend.
Shortly after he left the room, a student closed his window to make space
for some photos of their peers that they were was posting up.
- Persistence
When Dynamo was first introduced, users tended to habitually close down
all the windows open at the beginning or end of an interaction, in a manner
176
similar to the way people use normal desktop PCs when they log in and log
out. However, as time progressed, they began to realize the potential of
persistence in leaving up media, notes and parcels as a means for showing
and sharing with the wider community. This was most prevalent in the form
of leaving public parcels and public, open windows on the surface, as shown
in the log analysis detailed in Section 7.3.2. Initially, people would leave
items in the center of the display, but as time went on, they tended to
become more considerate, and would leave persistent items sized down to a
compact size at the far left and right edges of the displays, leaving ample
open space for public usage and avoiding occlusion of the palettes.
Interestingly, other users learned not to close these items down, by judging
whether they looked like they had been intentionally left up for public
consumption. In doing this, they would also often close down media that
looked like a “waste of space” (such as extraneous file browser windows) or
that had been on display a long while. This provided a mechanism for
“cleaning up” and clutter management, as detailed in the following section.
In the post hoc questionnaire, participants were asked to list a bad aspect of
their experience with Dynamo (see appendix 2 for the questionnaire). One
responded “When people take up the screen with a load of crap and you
can't get rid of it”. This refers to the challenge of dealing with clutter on the
communal surface. As previously stated, users tended to remove and tidy up
publicly accessible media. However, registered users were able to leave
locked media that no-one else could move. This gave rise to the practice of
“policing”, i.e. keeping track of users who left locked windows, and
reminding them to close them down. One user in particular (called Gemma,
[name anonymised]) took on this role.
177
sometimes putting newer or personally favoured items first, and older items
last. Again, one particular user (called Crispin [name anonymised]) took on
this role, as shown in Figure 7.16, below.
Figure 7.16 Crispin tidying parcels on days 5 (left) and 9 (middle), and
a paper notice (right)
Both the users who took on maintenance roles with Dynamo were also
members of the student union: a small group of students who had a position
of responsibility and were partially in charge of managing the upkeep of the
common room. For example, Crispin not only took responsibility for parcel
tidying, but also was responsible for putting up the posters for keeping the
room itself tidy. Therefore there was a strong parallel between the practices
and roles of responsibility, tidying and maintenance in the physical common
room, and on the Dynamo surface.
- Performances
178
stage for their performances, which they gave using digital media where
previously they used traditional delivery methods.
179
web together, as detailed in vignette (b) (Section 7.3.6), using it as a
resource to spontaneously find and display media to socialize about. These
progressed from being occasional activities to familiar practices which
people engaged in, becoming part of the community’s “vocabulary” of
things to do with Dynamo.
People in the vicinity of Dynamo were able to “oversee” (Heath and Luff,
1992) interactions while going about other things, i.e. without needing to
give it their full and undivided attention. As detailed in Chapter 2 (Section
2.3), O’Neil et al. (2004) refer to this property of visibility in the
surrounding area as a “public interaction space”. This property also
supported a model of “informal performer and audience” where the
performer or performers can be keen Dynamo users, putting effort and time
into the interaction, while the informal audience of people in the room can
dip in and out of their own activities (e.g. studying, reading, chatting),
watch the performance and interact verbally. This meant that the wider
community could benefit from the social interaction, disseminated
information and awareness information provided, without having to go to
much effort to do so.
This section will describe how Dynamo’s support of these various levels of
engagement enabled community members to engage with Dynamo in their
own chosen manner, and buy-in to it gradually, at their own chosen pace,
providing a path for initially less motivated community members to become
users. As exemplified in vignette (c) (Section 7.3.7), the level of
engagement a person invests in an interaction can vary from low to high
levels. This is shown in Figure 7.17 below, which lists a series of example
activities representing different levels of engagement. It should be
emphasized here that these levels are not intended to be conceptualized as
“steps” that a user must pass through, in order to reach a “gold standard of
high engagement”. Instead, they simply describe the level of engagement
180
for a particular interaction, or, as Goffman puts it, the “various grades of on-
lookers” (p.18).
Beginning at very low engagement, the scale begins with “being present”.
Although this in itself is not explicit interaction, it alters the ecology of the
room and provides the opportunities for further interaction. Next on the
scale is “bodily reaction”. This includes re-orientation towards the screens
(e.g. craning neck or turning chair to watch), and “bobbing” or dancing in
time with music. Next on the scale is “chirping”. This term was coined
during analysis to describe brief, low effort and low commitment
vocalizations, such as the helpful “Yep” shouted out in reply to a question in
vignette (c) (Section 7.3.7, Figure 7.13). Other “chirps” observed elsewhere
included “Ahh!” or “Oooh!”, or one-word statements such as “Yes”, “uh-
huh” and “Wow”. At a slightly higher level of engagement, “Shout outs”
were observed, in which users sometimes shouted out instructions to people
directly interacting with Dynamo, such as in the vignette of “low
engagement group interaction” (Section 7.3.7; frame 6 of Figure 7.13),
where someone in the background shouts “turn it up!”. Again, in this
example these low engagement interactors are still going about their own
activities and are not investing much effort in their involvement. Further up
the scale still, at a fairly high level of engagement, is chat and discussion
with people interacting with Dynamo, which typically consisted of talking
about the displayed media. At the top of the scale is direct interaction with
Dynamo. Here people interact either individually or as a group with
Dynamo, paying a great deal of attention to the system and conversational
interactions relating to the system and the media displayed.
181
The significance of these levels of engagement was that it enabled “gradual
buy-in” in that over the course of the study, a number of users tended to
move from typically low engagement or occasional use of Dynamo, to
progressively higher levels, and in doing so, their expertise of Dynamo
increased. In other words, people learned about the system piecemeal and
while going about their normal daily activities in the common room, without
them necessarily intending to do so. Although two community members
were highly motivated, and became dedicated users of Dynamo from the
first day onwards, the post hoc survey data shows that the majority of
community members took their time. As previously discussed in Section
7.3.1, there was a wide spread in the respondents’ day of initial use, and for
those who registered, there was a wide spread in the length of time between
initial use and registration. The following interview quote (from an
interview with Charlotte after the study) illustrates the nature of the gradual
buy-in that many community members engaged in:
The interviewee clearly states how she had learned by observing others
(very low engagement interaction) before her first usage of Dynamo, and
that upon her first usage, she was supported by two of her friends (high
engagement). This led to her acquisition of a pen drive and registration with
the system, and she went on to interact with Dynamo on various occasions
throughout the study (occasional high engagement).
182
Another example can be given with a short description of Heather’s
progress of buy-in to Dynamo. The point of note here is that nowhere along
the line did Heather explicitly set out to learn about Dynamo, her expertise
effectively “crept up on her”:
2. Days 3-4: She started to use it occasionally with her friends to pass
time between lessons (occasional episodes of high engagement).
4. Days 6-10: From there on, she used Dynamo more often, bringing in
and taking home media, putting up photos in carve regions, and so
on (i.e. regular episodes of high engagement).
183
interaction to fit into this window. In this sense, the entry point was “open”
and enabled this interaction. Had lengthier interaction been required by the
system, or had they needed to queue to use it rather than being able to just
grab a free interaction point, this would have prevented their interaction, i.e.
the entry point would have been “closed”. Other factors these entry points
varied upon included:
Other examples of closed entry points included when users attempted to use
the web functionality of Dynamo in the first week, and gave up when it did
not work. Conversely, in the second week, this became an open entry point
as the college system administrators enabled the internet connection. On the
184
Tuesday of the second week, a novice user was observed entering the room
alone during a quiet period in the afternoon, and, upon seeing a familiar
Google search page that happened to be open on the surface, sat down and
started browsing the web. Shortly afterwards a friend entered the room, also
a Dynamo novice, and they started co-browsing together and learning more
about Dynamo.
One of the most frequently observed closed entry points was that
registration with the system required the user to have a personal device
(owing to the way the system was implemented). This meant that if a user
did not own one, or happened not to have one with them at that point in
time, they could not register. In the following quote, Martha sees Daniel
making carve regions, (a facility only available to registered users), and
learns about their access control function from playing with him:
Had it not been for the need to have a personal device, Martha may have
registered with the system at this point. As such, it can be considered a
barrier. If the system was to support registration without a device, and
allowed addition of a device to the account at a later time if desired, this
would remove the barrier and facilitate further adoption.
185
them to the system and showing them activities it could be used for. It is
interesting to consider this against the fact that out of the 100 instructional
leaflets that were provided adjacent to the Community Display, only seven
were taken. This shows that the mechanism of informal learning and support
from community advocates was more prominent than the use of formal
instructional materials. Agostini et al. (2000) make a similar observation in
field studies of the Campiello Community Display system:
For novice users, there was also the possibility of being helped by other
users through the user interface, or alternatively, by sharing an interaction
186
point with a more experienced user, as was the case for the pairs in vignette
(b) (Section 7.3.6). The various possibilities of supported interaction
through the multi-user Dynamo interface can be conceptualized as a set of
open entry points, since they provide a number of paths by which a person
can enter into an interaction with Dynamo. However, upon closer inspection
of the observations, the multi-user interface was found to also present some
closed entry points.
Ben sees Nick him and invites him over: “[…] You’ve got to
try this, it’s cool, we both get a mouse!”
Nick goes over, stands next to him, grabs a free mouse and
starts interacting with Dynamo. He cannot gain access to any
of Ben’s windows and misdiagnoses the problem:
Nick: “It doesn’t work, I can’t click on anything […] The
mouse is broken.”
Ben: “It worked earlier […] Look these are the photos from
last week”
After being shown a few photos, Nick says he is going outside
for a cigarette and leaves.
This clearly shows how Nick almost had his first ever interaction with
Dynamo, which would have been tutored and probably quite beneficial to
his understanding. However, the system’s default access-control model
187
prevented him from achieving the tutored interaction. Later in the study,
some registered users discovered that they could put items in a public parcel
to give access to unregistered users, or log out to interact freely with
unregistered users. However, these work-arounds were not always clear,
and users either gave up on such interactions or interacted independently of
each other, unable to take advantage of the possibility of tutoring by means
of simultaneous interaction.
7.5 Discussion
The bulk of the discussion of the findings will be reported in depth in
Chapter 8. As such, this section provides an overview of the findings and a
discussion summary.
188
While the phenomena of “overseeing”, “vicarious learning” and the “honey-
pot effect” were evident here, they manifested themselves differently to that
observed in the one-shot setting studies. These differences can be
summarized in the following three concepts, as put forward in Section 7.4.
“Levels of engagement” describes how people were able to be involved
with interactions relating to the Community Display without having to be
completely immersed in the interaction. This meant that they could be
involved in interactions while going about their normal daily activities in
the space – e.g. hanging out, talking to friends, drinking coffee, studying,
and so forth. This enabled incidental and vicarious learning, which was the
most important way in which community members became users. The
second key concept put forward was “gradual buy-in”. This describes how
most community members tended to decide not to become fully-fledged
users on the spot, but adopted the system in a piecemeal fashion, in a
gradual manner that suited them. The third key concept put forward was
“entry points”. These are defined as the various contingencies in which
users attempt to initiate an interaction using the Community Display. For
example, they might attempt to interact unregistered, with a certain personal
device, a certain arrangement people giving help to each other in a certain
way, engaging in a particular activity. This is but one possible combination
of contingencies amongst many others. Thus, the wide array of entry points
helps enable gradual buy-in, since users can engage in a variety of
unplanned interactions, with a variety of different kinds of help or support
from their peers.
The following chapter will discuss these concepts in depth and draw
comparisons between them and those put forward in the other case studies.
It will also put forward a number of design suggestions for Community
Displays.
189
Chapter 8
Discussion
190
8.1 Introduction
This chapter will address the research questions posed for this thesis in
Chapter 2. It will do this by reviewing and consolidating the findings of the
user studies, and will draw out design suggestions, which will be shown as
grey highlighted boxes throughout the body of the chapter. These design
suggestions are clarifications and improvements on the suggestions and user
experience principles put forward earlier in the thesis. They are intended to
enable Community Display system designers to browse through this chapter
and quickly locate the parts most relevant to them. The research questions
are repeated below.
191
8.2 Understanding the Phenomena of Situated Social
Behaviour around Community Displays
This section will address research questions 1a, 1b and 1c by discussing
them generally:
Common to all three of the case studies were phenomena relating to the
public availability of usage of the display and people in the vicinity
overseeing these activities, which resulted in mechanisms that supported
usage and adoption. This section will detail these phenomena in each case
study.
Here, the large size of the display and its conspicuous positioning in the
shared office meant that when a team member used it, for example while
working on a scheduling problem, their team members could see them doing
it as they passed by. This was because there was a continual flow of people
in and out of the office, owing to its centrality in the local environment and
the collocation of shared resources there (e.g. storage of the audio-visual
equipment). Thus usage of the display was a “double duty” activity (Heath
and Luff, 1991; Luff and Jirotka, 1998; Robinson, 1993). On the one hand
the user was working on their task, while on the other, they were sending
out a tacit invitation to passers-by that they were working on a problem on
the large display, and that they might benefit from the help and extra
knowledge their team members could offer them. This often created
192
spontaneous gatherings of three or four people who would stand together
and work on the large display, discussing it, gesturing and annotating on it.
What was particularly beneficial about this mechanism was that it enabled
co-operative group work to occur, without requiring any effortful
coordination or articulation work (cf. Schmidt and Simone, 1996, on
articulation work). For example, the person in need of help did not need to
radio out a message to the team requesting it and negotiating a time to meet
in front of the display. The quote below from Jordan and Henderson (1995)
postulates that a similar mechanism may occur with large displays used in
industrial process control rooms.
193
this self sustained as people seemed to be stimulated to find out more about
what the crowd was involved in and to join in the social activities. This
resonates with Whyte’s (1980, 1988) analysis of urban parks and plazas in
the USA, in which he states that above and beyond the features of the space
itself, what attracts people most is other people. A further and more detailed
analysis of the facilitators of interaction this setting is carried out in the
following section.
In the audio-visual team case study, the benefit of the honey-pot effect was
to save them coordination work, enabling them to collaborate with less
effort. In the other case studies, the benefit of the honey-pot effect was
different – the activities in these settings was not group work but relaxation,
recreation and a mix of other activities depending on the specifics of the
setting and the individuals themselves.
194
between them. In reflection, one of the benefits here was that it allowed
people to engage with the Community Display and with each other at their
own discretion: it was up to them who they wished to freely associate and
mingle with. Thus the Community Display helped the community engage in
group interactions without requiring any formality or pre-arrangement. This
is particularly relevant in this kind of one-shot setting, where there is a
strong emphasis on recreation and enjoyment in interaction with other
people (cf. Goffman, 1963) - an experience which is highly subjective.
In the on-going setting studied, the activities engaged in and the topics of
discussion were very ad hoc. For example, in vignette (b) reported in
Chapter 7 (Section 7.6.3), the conversation held by the group was
“meandering” and much of the interaction was spent looking for an
enjoyable activity or topic of conversation. As such, it would not make
sense to try to planfully coordinate this sort of open-ended interaction.
Instead, on-lookers and passers-by judged for themselves at any particular
moment whether they found an interaction to be interesting, and could join
in “off the cuff”. For example, in vignette (a) in Chapter 7 (Section 7.3.5)
someone enters the common room, sees a friend interacting with Dynamo
displaying cartoons, and decides to go and join them for a conversation.
Also, in vignette (d) in Chapter 7 (Section 7.3.8), two on-lookers join into
an interaction that began between two people, their interest stimulated by
the content being shown (stills and music from a recent pop concert).
Coordinating such gatherings would be effortful since the community
members tended to have very different schedules and were often
geographically spread out (e.g. in different rooms on campus, working from
home, on a break in town, etc), not to mention potentially expensive should
it be carried out using SMS messages or mobile phone calls.
195
any knowledge of its existence. To quote one interviewee “Nobody really
knew what it was when they came in”. The members would progress in their
understanding of what the Community Display is by observing others using
it. To quote another interviewee “… I did see people standing around it so I
stood beside and watched for a while” These observations allowed them to
learn about the system vicariously – allowing them to ascertain what the
system did, how long it seemed to take, whether it looked easy or socially
awkward, and so on. Based on these observations, they would evaluate
whether or not to try out the Community Display themselves.
196
G4: General Design Suggestion (4)
Use a visual interface to support vicarious learning
If the interface is highly visual and displayed on the Community Display,
an observer can learn the steps of interaction with the system by watching
others using it. “Hidden” interactions like key combinations, gestures or
interaction on an associated private display is hidden and will prevent this
mechanism of learning. However, this observability has also been
documented to cause evaluation apprehension in one shot settings which
can deter usage. This is a trade-off that needs to be considered carefully
(see Section 8.5.1)
Seely-Brown and Duguid (2000) recount the manner in which Bell Co.
stimulated the adoption of the telephone in the late 1800s. At the time, the
telephone was a completely novel technology and its benefits were not
immediately apparent to end users. For example, Western Union in the US
and the Post Office in the UK both declined Bell Co.’s offers to sell the
patent to them at a very low price. Indeed, the chief engineer at the British
post office famously replied that they did not need the telephone because
“…we have plenty of messenger boys.” (Preece, 1876; as cited in Cosier
and Hughes, 2001, p. 9). As a result, Bell Co. instead took a grass roots,
bottom up strategy of getting the telephone directly into the hands of the end
users. This involved putting telephones in hotel rooms for calling the front
desk, and in offices as intercoms. But most interestingly, they also put
phones near lunch counters in diners and lunch rooms, to quote Seely-
Brown and Duguid: “…that way, it reasoned, people who didn’t know how
to use them would be likely to see people who did know how and in this way
learn about the phone system.” (p. 7)
197
This strategy has a great deal in common with the design suggestions drawn
from the findings of the case studies in this thesis (see the grey boxes on
previous pages). By positioning by a lunch counter, Bell Co. are likely to
have considered public availability of the telephone usage, the highly
concentrated flow of people past the telephone as they queued to buy their
food, and the stage-like “performance for an audience” nature of this spatial
arrangement. Indeed, the choice of setting was a communal space in which
people were on breaks from their work and likely to have the disposable
time available to try out a new technology. It is interesting to consider how
Bell Co leveraged situated voluntary adoption to facilitate marketplace
adoption. Indeed, many modern marketing promotion techniques used in
encouraging marketplace adoption have similarities, for example, television
advertising and product placement in films, in which interaction with a
technology is depicted as a performance for potential users to observe and
learn from (Kotler and Armstrong, 2004).
198
Having discussed the findings that relate to research questions 1a, b and c,
the questions relating to the two different types of setting arise. However,
before these are moved onto, another design suggestion first needs to be
made. Referring back to Chapter 2, an analysis was carried out on of a
number of published Community Display research studies which were
aggregated on a set of dimensions. From this analysis, two marked clusters
appeared: “one-shot” and “on-going” settings, which were defined on the
dimensions of “permanence of communal space” and “community
interconnectedness” (Section 2.5). The findings of the Opinionizer and
Dynamo case studies provided further evidence of their existence and added
detail to the understanding of their nature. Most importantly, the phenomena
of situated behaviour and voluntary adoption were substantially different,
requiring different design considerations for each setting. As such, this gives
rise to an important question that Community Display designers must ask
themselves:
199
For example, the culture of the community and the practices normally
carried out the space are also likely to have a big impact (cf. Orlikowski,
1992). Further limitations of the one-shot and on-going characterizations are
discussed in detail in Section 9.2 of Chapter 9.
To begin with the most obvious and most important characteristic of one-
shot settings, as implicit from the definition, people will see the Community
Display system for first time, and may never be exposed to it again. This has
important implications for design, as the following design suggestion states:
200
Community Display is likely to be limited, and may even be as short as a
few minutes (as seen in the Opinionizer field studies, Chapter 4). Finally,
when considering using the Community Display, they are likely to be
apprehensive of being evaluated by on-lookers, and be concerned about the
possibility of social awkwardness or embarrassment (cf. Section 4.6).
The model of interaction put forward in the one-shot setting case study
(Chapter 4) describes the steps a user takes in deciding to interact. This is
summarized in Figure 8.1. The first step in this process is achievement of
initial visibility, i.e. the individual becoming aware of the Community
Display’s existence. Therefore it is highly important that people in the
communal space are likely to actually see it. A strategy to achieve this is to
position the Community Display in a way that offers line of sight to the
main points of flow which community members pass through, e.g. the
entrance to the space. Furthermore, if these are places where users spend
time queuing (e.g. at a bar or food buffet), it is desirable that there is also
line of sight from this location, since this is “dead time” in which people
will have the energy and attention to look at other things in the environment
(Underhill, 2000). Gehl (1987) makes a similar point in specifying design
recommendations for plazas in urban spaces:
"If people do not see a space, they will not use it. […] Unless
there is compelling reason, an open space should never be
sunk [...] sunken plazas are dead spaces." (Gehl, 1987; p.99)
201
Figure 8.1: The interaction model derived for one-shot settings. Key
thresholds (decision points) shown in yellow.
202
Following the achievement of initial visibility (the first decision point
shown in Figure 8.1), the community member then needs to be stimulated to
take a closer look and find out more about what the Community Display
does. This decision point is defined as the “threshold to attention”, and is
depicted in yellow in Figure 8.1 (a re-representation of the interaction model
put forward in Chapter 4). Encouraging a community member to cross this
threshold entails a number of design considerations, since the display must
entice the user forward, as described in the following design suggestion:
If the person does decide to find out more and move closer to the
Community Display, this is the first stage of success, since they are now
likely to have moved to within the immediate vicinity of the Community
Display. This opens the door to possible social interactions in the vicinity.
In other words, an observer may enter into conversation with other people in
the vicinity, using the media displayed on the Community Display to “boot
strap” the conversation. This “non-direct” form of interaction in relation to
the Community Display is a legitimate form of interaction since it clearly
fulfils the intended function of supporting informal social interaction.
However, this alone is only part of the success, since some users need to
decide to actually come forward, interact with the system and furnish it with
topical content. This is the second decision point depicted in yellow in
Figure 8.1.
In making this decision, the observer evaluates the activity of using the
system. They ask themselves questions such as “How long will it take?”,
“Does it look easy or will it be awkward?”, “Will it be fun?” and so on.
They need to conclude that the benefits of interacting will outweigh the
203
costs, e.g. the time and the effort. The system designer needs to facilitate
this decision-making process by communicating the key information about
the system to the observer in a clear and concise manner, as detailed in the
following design suggestion:
If the individual then decides to interact with the system, the system has
succeeded in its goals with that particular user – in fostering social
interaction around the Community Display and encouraging direct usage. In
the case of the Opinionizer studies, this observable usage progressed into a
self sustaining honey-pot effect, which was highly important since, as stated
204
earlier, a major attractor of people is other people. Therefore design
suggestions G1 to G4 are important here.
On the other hand, if people were tending to look at the system, but not
interacting with it or talking about it between themselves, the problem
would then seem to lie at the threshold to interaction, i.e. the point at which
they evaluate the system. Since Carter et al.’s (2002) system used abstract,
unlabelled icons to represent information, it can be postulated that this
confused observers. Unable to make an informed evaluation of the system,
they may have been left in doubt as to whether it was worth an investment
of effort. The Opinionizer study findings indicate that when observers see a
Community Display for the first time, they may tend towards a negative
evaluation (“There was a whole kind of fear because it was something
new”) , and that they tend towards a fear of social awkwardness. Carter et
al’s system may have failed to achieve adoption because the system was not
designed to dispel these suspicions and encourage the observers to interact.
205
Community Display. McCarthy et al. took the step of making the
community members aware of the existence of the system in advance, in an
email sent out 4 days prior to the event, in which a link was provided to an
explanatory web-site (McCarthy et al., 2004; McCarthy, Personal
Communication March 9, 2005). This served to temporally extend the
stages of interaction. Referring back to Figure 8.1, this served to enable
users to pass through the thresholds of “attention” and “interaction”. A keen
user therefore could decide to buy-in to the system having read about it, and
the only remaining step would be for them to locate the displays in the
venue.
However, McCarthy does not report a great deal of success with this
approach. In fact, he observed that people were most motivated to become
users after having seen others using the system – the order which the
interaction model (depicted in Figure 8.1) proposes. It seems that witnessing
usage of the system first hand is a potent motivator to adoption. To quote
McCarthy:
206
interaction model would concur, since this would entice more users to pass
through the threshold to interaction.
Had this process been streamlined, this 3 stage process could possibly have
been made a single, brief step. For example, the web spider software could
have been used to harvest registration information in advance from personal
web-sites, rather than making the users wait while it did it for them on the
spot (cf. Stasko & Zhao, 2002; Carter et al., 2002). Then, if a user wanted to
be involved, they could simply just collect the RFID tag that had been pre-
made for them.
207
8.3 Understanding Adoption in on-going settings
This section will address research questions 3a, 3b, 3c and 3d:
208
freeform text entry, drawing, media display, arrangement, and sharing. Such
functions can be exploited in a range of different activities, e.g. formal
meetings, brainstorming activities, noticeboard display, socializing, and so
forth. This leads into the following design suggestion:
This is contrary to the logic used by many UbiComp researchers, who tend
to try to develop systems which use only the newest and most cutting edge
technology to investigate the possibilities of systems that will be used one
day in the future in real world environments (cf. Abowd and Mynatt, 2000;
Ndiwalana et al. 2003). For Community Displays to be evaluated, they need
209
to be deployed – and adopted – in real world environments, which means
they need to be backwards compatible with legacy technology. Borovoy
(2001) makes a similar point, and states that when users took the “Thinking
Tag” devices home with them (away from the “Community Mirror”
Community Display system), they were not compatible with normal PCs or
any other devices, rendering them useless. He refers to this as the “lost
horizons” problem, after Capra’s 1938 film (Borovoy, 2001; p.44).
210
in which people spend time – such as seating areas, rather than just the focal
points of flow as is the case in one-shot settings.
211
number of users tend to pass through, not as a prescriptive progression –
users move through this at their own discretion.
Within the pale orange box on the right hand side is a representation of
usage of the display. A user can engage in an interaction at any level of
engagement that suits them, and through any available entry point. The
main difference compared to one-shot settings is that members here use the
space regularly in their daily lives, and thus are exposed to it while doing
their everyday activities. It may be the case that upon their very first
exposure to the Community Display, the one-shot interaction model (Figure
8.1) applies to an extent, but this first time exposure makes up a very small
proportion of their use of the space thereafter.
212
Figure 8.2: Interaction model for on-going settings
213
Overseeing and peripheral involvement was observed to occur slightly
differently in the on-going setting as compared to the one-shot setting.
There was a diversity of other activities that people carried out
simultaneously in the communal space. For example, people would study
quietly, read books, work in groups, socialize, drink coffee together, eat
lunch or snacks, or have meetings at the same time, creating “multifocused
gatherings” (Goffman, 1963) which occurred on a regular basis. While
being involved in these other activities, community members would often
be involved with the Community Display interactions at various different
levels of engagement (shown in orange in Figure 8.2, and detailed in Section
7.4.2 of Chapter 7). For example, low levels of engagement included bodily
reactions such as turning to face the display when something interesting was
being shown, or “bobbing” appreciatively in time with music being played.
Higher levels of engagement included brief, low effort vocalizations such as
“ahh”, “ohh” and “wow”, which were termed “chirping”; and at a slightly
higher level of engagement was “shout outs” where users shouted
instructions to the people interacting with Dynamo, such as “turn it up”
(referring to music they enjoyed). High levels of engagement included
conversations with people interacting with Dynamo, and direct interaction
using a mouse and keyboard. Thus low levels of engagement enabled people
to be involved with interactions but while maintaining low effort and low
commitment, allowing them to carry on with what they were doing. The
significance of this was that it enabled “gradual buy-in”, in that many users
tended to move from typically low engagement or occasional use of
Dynamo, to progressively higher levels, and in doing so, their expertise of
Dynamo increased. In other words, people learned about the system in a
piecemeal fashion, and, most importantly, they were able to do so at a pace
that suited them. This is shown in green in Figure 8.2.
From this, the question arises, what was it about Dynamo that facilitated this
gradual buy-in? Analysis of the study findings show that it was partially due
to the wide array of different ways a user could interact with Dynamo, i.e.
its many “entry points” that enabled this. For example, a user could be
214
registered, or not registered; they could initiate an interaction with any kind
of compatible personal device, or with no personal device at all; they could
enter an interaction with a group and gain a range of different kinds support
from people using the other interaction points, or they could interact alone
and independently. These entry points allowed a user to engage in an
interaction with Dynamo that suited their level of expertise and the level of
engagement that they sought. It also allowed spontaneous usage, since it
enabled a wide range of usage scenarios, rather than requiring users to plan
an interaction around a limited set of entry points. For example, if every
user needed to be registered to interact, achieving a group interaction would
require planful-ness and effort to arrive at the entry point in which each user
is registered, which could deter usage. This leads on to the following design
suggestion:
215
tea time, a daily scheduled time when people gather in the
lounge area.” (p. 1)
Houde et al. carried out a field study on themselves and the other people
they shared a building with, by deploying the NewsLens in the shared
kitchen area. They reported a number of positive findings. For example,
community members expressed that it was fun to read, they liked the fact
that they didn’t need to go to a special place, either physically or “virtually”
(on the intranet) in order to discover news, and they also preferred it to
receiving group emails. However, the field study also revealed adoption
problems. After an initial wave of enthusiasm, usage dropped off a great
deal: “Not many stories were posted… approximately two stories per day”
(p. 3). They were left unsure as to the exact cause of this, postulating that
the system may have simply have been too time consuming for the
community to fit into their busy working days.
If the system is critiqued on the basis of the interaction model and design
suggestions put forward in this section, its shortcomings seem to relate to
the problem of appropriation. Put simply, perhaps the community were not
motivated to use the system in the prescribed way. If the system had offered
the flexibility for other uses, this may have allowed them to devise their
own uses and practices that suited their needs more effectively.
Firstly, items that were submitted for display were required to be news
items – they were even formatted in the appearance of a newspaper’s front
page. This may have communicated the implicit prescription that items
posted should be newsworthy journalism – an activity that users may have
evaluated as both time consuming and as a source of evaluation
apprehension. Reflecting on their study findings, this is something that
Houde et al. concur: “Writing short, informative stories for the Newspaper
is a skill that does not come naturally to most users.” (Houde et al. 1998;
p.3).
216
Also, the submission to the display had to be done via email. Given the
early year of this research (1998), it is not surprising that they did not offer
additional facilities for technology interconnection, Indeed, devices such as
PDAs, digital cameras and wireless laptops were not even in popular use at
that time. Nonetheless, this limitation had a certain impact on the adoption
of the system, since users were required to return to their workstations and
write the news item from there. This meant that users needed to leave the
communal space and return to their workstations in order to add an item to
the Community Display. The effort involved may effectively have closed an
entry point for many of the users who, upon seeing the information
displayed or upon having an interesting conversation in the communal
space, might “there and then” wish to post up an item onto the display, with
their colleagues immediately available to help and give feedback.
Finally, the items submitted had to be text only. This ruled out the use of
pre-made content, such as existing web-pages, documents, photos, video
clips, and so forth. This contrasts with more recent research systems, such
as Dynamo (Brignull et al., 2004), Plasma Poster (Churchill et al., 2003)
and Community Wall (Grasso, 2003), which all offer the ability to display a
range of media types.
In all, if NewLens had offered the scope for more flexible kinds of usage,
the community may have evolved their use of it into a tool for a different
kind of purpose. For example, it may have become less of a tool for
217
dissemination of journalistic news stories, and more for more for informal
and low effort media sharing. This analysis has shown that the concepts and
interaction model put forward for on-going settings can be useful in making
sense of study findings and suggesting improvements for iterative redesign.
This chapter has discussed the contributions of this thesis in depth, first
looking at the general phenomena of situated social behaviour around
Community Displays, and then looking consecutively the specific voluntary
adoption issues relating to one-shot and on-going settings. The following
chapter will provide suggestions for future work and the final conclusions of
the thesis.
218
Chapter 9
219
9.1 Introduction
This chapter will begin by reflecting on the work as a whole, recounting the
research narrative, and making suggestions for different ways it could have
been carried out, given the benefit of hindsight. Secondly, this chapter
discusses some important avenues for future research in this area. Thirdly
and finally, this chapter will draw the final conclusions, summarizing the
main contributions of the thesis to field of Community Display research.
9.2 Reflection
The outset of this research began with the general motivation to understand
Community Display systems and the sociotechnical issues relating to their
usage. From this, the problem of voluntary adoption revealed itself to be an
important concern – contemporary researchers in the area were producing a
range of technically impressive Community Display systems, yet many
systems suffered from voluntary adoption issues (e.g. Carter et al., 2002;
Agostini et al., 2002; Houde et al., 1998; Churchill et al., 2003; Greenberg
& Rounding, 2001). Without adoption, all the time and effort put into the
development of these systems could render them into mere ornaments,
regardless of their potential usefulness. As such, this was chosen as the
central motivating research issue for this thesis. In other words, why were
these adoption problems happening? What was the nature of the adoption
problems? And would looking closely at the situated social behaviour
people engaged around these systems reveal a better understanding of the
problems?
Prior to the outset of the practical research, a detailed analysis was carried
out on user-studies reported in the Community Display literature by
comparing the studies on a number of different dimensions. Given the
degree to which these studies varied in their social settings, an emphasis
was placed on the social aspects of the communities and the communal
spaces in which the studies were carried out. From this, a marked clustering
was found, giving rise to the characterization of two distinct types of social
220
setting: “one shot”, which were temporary settings used by loose-knit
communities; and “on-going”, which were permanent settings used by tight-
knit communities.
In hindsight, more initial testing of the system during development (e.g. low
fidelity prototyping and brief field evaluations) could have brought this to
light sooner, and resulted in a system that was more suited to the setting,
perhaps even to the extent that the team could have used it to support their
work during a conference. However, this is a minor criticism since the
findings made were still valuable.
The second and third case studies consecutively looked at one-shot settings,
and on-going settings. Two systems were developed: Opinionizer for one-
shot settings, and Dynamo (as part of a collaborative project) for on-going
settings. These were studied in their target settings. In reflection, it would
also have been revealing to have observed the adoption of the Opinionizer
and Dynamo Community Display systems outside of their target
221
deployment settings. Thus Opinionizer could have been deployed in an
ongoing communal space, and observed over a period of weeks; while
Dynamo could have been deployed at a one-shot event such a conference.
These additional studies could have provided an even better understanding
of the interplay between design, functions, voluntary adoption and types of
setting. However, given that this would require substantially more work, it
is more a case for future work.
The following section will go into more detail on some of these issues,
discussing the limitations of the research reported in this thesis and
proposing a number of avenues for future research in this area.
The characterization of “one shot” and “on going” settings has shown itself
to be fruitful, since the two types have been found to vary considerably in
their nature and in the type of situated voluntary adoption that occurs in
them. However, this characterization only aims to represent the first steps of
research in this area, and is by no means intended as universally applicable
to any kind of setting. This section will hypothesize some of the ways in
which communal spaces vary outside of the aspects investigated in this
thesis, and will discuss other issues that arise.
222
like a one-shot setting. However, the community are familiar with each
other, and with all the established social norms, activities and practices that
are carried out in their normal setting This produces a different blend of
requirements for the Community Display. For example, it may be beneficial
to offer some of the characteristics of a Community Display for on-going
settings, allowing users to engage in freeform activities with each other,
rather than encouraging them to engage in a simple pre-packaged activity to
encourage them to associate with each other. However, it is also important
for them to discover and learn how to use the Community Display within
the time available to them, which is a characteristic found in one-shot
settings. This would no doubt present a number of dilemmas to the system
designers, and is a valuable avenue for future work.
Another complexity not considered in this thesis is that the people present in
a communal space may be from a number of different communities and sub-
communities. For example, in the common room observed in the Dynamo
study, the community as a whole was composed of two parts: students and
teaching staff. For the students, the permanence of the common room would
have been 1-2 years in total from joining to leaving the college. For the
teaching staff, this permanence may have been much longer. Similarly, at
the social events observed in the Opinionizer studies, the service staff would
have used the space regularly as part of their daily working life while for the
visitors it was their first and probably only visit. Oldenburg (1998) describes
the phenomenon of how some people are not loyal to a particular location,
but to their social group, basing this on observations of groups of university
students who rotated the bars and other venues they visited around a city.
He refers to the venues used solely in this way as “bring your own friends
venues” (p. 171), and describes them as containing a number of separate
cliques that socialize mainly within themselves.
This concept of mixed settings, in which there are multiple communities and
sub-communities, means that a Community Display system designer may
need to design for the different needs of the different communities within a
single Community Display system, which is a new challenge. Study of the
223
area of community social network structure and interconnection (e.g.
Wellman and Berkowitz, 1988; Wasserman and Faust, 1994) presents itself
as an interesting avenue for this future work, providing an alternative way
which to look at settings, design issues, and the effects a Community
Display may have on a community.
224
engage in a private interaction, but if they were at a social event at the same
venue at another time, they might be happy using the large wall display and
might welcome open interaction from the people around them. In such a
scenario, the displays in the environment might offer a generic display
service, which users could appropriate and run Community Display
applications on if they wished to, as proposed by Black et al. (2002, 2003).
This presents itself as an interesting avenue for future work, i.e. looking at
how Community Displays would exist in a wider world of Ubiquitous
Computing devices, and the implications this would have for their design,
usage and voluntary adoption.
225
effects, and the long term processes of appropriation. In such studies, a
different observational methodology would be used in comparison to the
Dynamo field study. Instead of a 10 day intense study in which video is
recorded continuously, a spread out method of sampling would be more
suitable, in which occasional video or observational sessions are carried out,
(e.g. twice a month spread over a year), and system logs are used to provide
the day-to-day detail.
These longer term studies will be an important addition to the mix of studies
being carried out on Community Display systems. Indeed, as Andriessen et
al. (2003) state below, maintaining a healthily mix of different styles of
studies, theories and methodologies is important to the development of a
well rounded understanding of the phenomena of adoption:
226
the field of Community Display research is nascent, this possibility is still a
long way off.
227
• Marketing and Consumer Psychology
Concerned with facilitating the uptake of commercial products with
a view to commercial success (e.g. Kotler & Armstrong, 2004;
Moore, 1991; Underhill, 2002).
This thesis has focused on the problem of achieving the voluntary adoption
of Community Display systems, has contributed to a better understanding
228
this problem, and has proposed suggestions for minimizing it through
attention to the social setting and the system design. Overarching conceptual
contributions which were found to apply across all the setting studies were
the concepts of “user flow” and “the honey-pot effect”. As people moved
through the space and engaged in other activities (“flow”), they would
oversee usage of the Community Display, owing to its large size and
prominent positioning, which often would serve to spontaneously entice
them into a social interaction with those people (“the honey-pot effect”).
This served to foster spontaneous group congregations in which informal
social interaction would take place. Overseeing also played a very
important role in enabling vicarious learning, a process by which
community members would learn about the Community Display by
watching others using it.
Another key contribution of this thesis was the characterisation of two kinds
of communal space setting: “one-shot” and “on-going”. One-shot usage
settings include one-off social events (Opinionizer, Chapter 4), conferences
(McCarthy, 2003;) and festivals (Agamanolis, 2003). Such settings are used
briefly, for a few hours or days, and may not be re-visited by members of
the user community. On-going settings, on the other hand, include common
rooms (Dynamo, Chapter 8; Houde et al., 1998), cafés (Churchill et al.,
2003), and relaxation areas (Grasso, 2003), and are used regularly by an
established community over a period of months or years. The differences
between these settings have been shown to have substantial effects on
situated user behaviour and voluntary adoption; and as such they demand
different kinds of Community Display system to cater for this.
229
to learn more about it, and the second, the “threshold to interaction”
involves them deciding to interact with it directly. These thresholds are key
points at which a user may decide whether or not to interact, and are posed
as a key focus of attention for designers of any Community Displays for
one-shot settings. However, this model is particular to one-shot settings,
owing to the short window of time a user has available to discover,
understand and choose to use a Community Display. The design suggestions
put forward relate to encouraging users to cross these two different
thresholds, and achieving the all-important first time usage of the system.
In the on-going setting studied, adoption was often (though not exclusively)
observed to progress gradually from low to higher engagement over a
period of days, described as a user’s “buy-in progression”. Also, usage
situations were observed to vary in a number of different ways, such as
spontaneous or planned, brief or lengthy, and independent or supported.
These various potential contingencies were defined as an array “entry
points”, and it is suggested that these should be wide-ranging, to cater for
the different possibilities of interaction a user may attempt to engage in. The
design suggestions put forward relate to supporting the full range of
potential users and usage situations, enabling “gradual buy-in” by not
230
requiring any unnecessary jumps in investment in the system as they learn,
and also and offering flexibility of the system functions to cater for
community appropriation, enabling them to use the system for the activities
that they prefer.
To conclude, the interaction models and concepts put forward in this thesis
provide the beginnings of a framework or lingua franca for researchers and
system designers, enabling them to better understand the interplay between
the settings, the social behaviour, the voluntary adoption and the design
issues that are at work in the design, use and appropriation of Community
Displays.
231
References
232
Abercrombie, N., Hill, S., & Turner, B. S. (1988). The Penguin dictionary
of sociology (2nd ed.). London: Penguin.
Abowd, G. D., & Mynatt, E. D. (2000). Charting past, present, and future
research in ubiquitous computing ACM Trans. Comput.-Hum. Interact. , 7
(1), 29-58.
Agostini, A., De Michelis, G., Divitini, M., Grasso, M. A., & Snowdon, D.
(2002). Design and deployment of community systems: reflections on the
Campiello experience. Interacting with Computers, 14(6), 689-712.
233
Aronson, S. H. (1979). The Sociology of the Telephone. In G. Gumpert &
R. Cathcart (Eds.), Inter-media : interpersonal communication in a media
world. New York; Oxford: Oxford University Press.
Bellotti, V., & Bly, S. (1996). Walking away from the desktop computer:
distributed collaboration and mobility in a product design team In
Proceedings of the 1996 ACM conference on Computer supported
cooperative work (pp. 209-218). Boston, Massachusetts, United States
ACM Press.
Bellotti, V., & Rogers, Y. (1997). From Web press to Web pressure:
multimedia representations and multimedia publishing In Proceedings of the
SIGCHI conference on Human factors in computing systems (pp. 279-286).
Atlanta, Georgia, United States ACM Press.
Benford, S., Bederson, B. B., kesson, K.-P., Bayon, V., Druin, A., Hansson,
P., et al. (2000). Designing storytelling technologies to encouraging
collaboration between young children. In Proceedings of the SIGCHI
conference on Human factors in computing systems (pp. 556-563). The
Hague, The Netherlands: ACM Press.
234
Benford, S., Bowers, J., Fahlen, L., & Greenhalgh, C. (1994). Managing
Mutual Awareness in Collaborative Virtual Environments. In Proceedings
of VRST'94, (pp. 23-26), Singapore.
Bentley, R., Hughes, J. A., Randall, D., Rodden, T., Sawyer, P., Shapiro, D.,
et al. (1992). Ethnographically-informed systems design for air traffic
control In Proceedings of the 1992 ACM conference on Computer-
supported cooperative work (pp. 123-129). Toronto, Ontario, Canada ACM
Press.
Bier, E. A., & Freeman, S. (1991). MMM: a user interface architecture for
shared editors on a single screen In Proceedings of the 4th annual ACM
symposium on User interface software and technology (pp. 79-86). Hilton
Head, South Carolina, United States ACM Press.
Black, J. A., Hong, J. I., Newman, M. W., Edwards, W. K., Izadi, S.,
Sedivy, J. Z., and Smith, T. (2002). Speakeasy: A Platform for Interactive
Public Displays, Retrieved February, 2005, from
http://www.appliancestudio.com/cscw/black.pdf
Black, J. A., Edwards, W. K., Newman, M. W., Sedivy, J. Z., & Smith, T. F.
(2003). Supporting Extensible Public Display Systems with Speakeasy. In
K. O’Hara, M. Perry, E. Churchill & D. Russell (Eds.), Public and Situated
Displays. Social and Interactional Aspects of Shared Display Technologies.
London: Kluwer.
Blaine, T., & Perkis, T. (2000). The Jam-O-Drum Interactive Music System:
A Study in Interaction Design. In proceedings of Designing Interactive
Systems 2000 (pp. 165-173). New York, USA.
Bly, S. A., Harrison, S. R., & Irwin, S. (1993). Media spaces: bringing
people together in a video, audio, and computing environment Commun.
ACM 36 (1), 28-46.
235
Bly, S. A., Harrison, S. R., & Irwin, S. (1993). Media spaces: bringing
people together in a video, audio, and computing environment Commun.
ACM 36 (1), 28-46.
Borovoy, R., Martin, F., Vemuri, S., Resnick, M., Silverman, B., &
Hancock, C. (1998). Meme tags and community mirrors: moving from
conferences to collaboration In Proceedings of the 1998 ACM conference on
Computer supported cooperative work (pp. 159-168). Seattle, Washington,
United States ACM Press.
Bradner, E., Kellogg, W. A., & Erickson, T. (1999). The adoption and use
of “BABBLE”: a field study of chat in the workplace. In Proceedings of the
Sixth European conference on Computer supported cooperative work (pp.
139-158). Copenghagen, Denmark: Kluwer Academic Publishers.
Brignull, H., Izadi, S., Fitzpatrick, G., Rogers, Y., & Rodden, T. (2004).
The introduction of a shared interactive surface into a communal space In
Proceedings of the 2004 ACM conference on Computer supported
cooperative work (pp. 49-58). Chicago, Illinois, USA ACM Press.
Brignull, H., & Rogers, Y. (2003). Enticing People to Interact with Large
Public Displays in Public Spaces. In Proceedings of Interact 2003 (pp. 17-
24). Zurich, Switzerland.
236
supported cooperative work, (pp. 335-355). Dordrecht: Kluwer Academic
Press.
Brown, J. S., & Duguid, P. (2000). The social life of information. Boston,
Mass.: Harvard Business School; New York : McGraw-Hill.
Carter, S., Mankoff, J., & Goddi, P. (2002). Representing and supporting
action on buried relationships in smart environments. Paper presented at the
Workshop on Public, Community and Situated Displays at Computer
Supported Cooperative Work (CSCW’02), New Orleans, USA.
Cheverst, K., Clarke, K., Fitton, D., Rouncefield, M., Crabtree, A., &
Hemmings, T. (2003). SPAM on the menu: the practical use of remote
messaging in community care In Proceedings of the 2003 conference on
Universal usability (pp. 23-29). Vancouver, British Columbia, Canada
ACM Press.
Greenhalgh, C., Izadi, S., Mathrick, J. Humble, J. & Taylor, I. (2004). ECT:
A Toolkit to Support Rapid Construction of Ubicomp Environments. Paper
presented at the System Support for Ubiquitous Computing Workshop,
UbiSys04, Nottingham, England.
237
Public, Community and Situated Displays at Computer Supported
Cooperative Work (CSCW’02), New Orleans.
Churchill, E. F., Nelson, L., Denoue, L., Helfman, J. I., & Murphy, P.
(2004). Sharing Multimedia Content with Interactive Displays: A Case
Study. In Proceedings of the 2004 conference on Designing interactive
systems DIS’04 (pp. 7-16). New York, USA.
Churchill, E. F., Nelson, L., Denoue, L., Murphy, P., & Helfman, J. (2003).
The Plasma Poster Network. In K. O’Hara, M. Perry, E. Churchill & D.
Russell (Eds.), Public and Situated Displays. Social and Interactional
Aspects of Shared Display Technologies (pp. 233-260). London: Kluwer.
Cleland, C., & Carmichael, M. (1997). Banners that move make a big
impression. Retrieved February, 2005, from
http://adage.com/interactive/articles/19970913/article3.html
Crabtree, A., Hemmings, T., & Rodden, T. (2003). The social construction
of displays. In K. O’Hara, M. Perry, E. Churchill & D. Russell (Eds.),
Public and Situated Displays. Social and Interactional Aspects of Shared
Display Technologies (pp. 170-190). London: Kluwer.
238
de Souza, C. S., & Preece, J. (2004). A framework for analyzing and
understanding online communities. Interacting with Computers, 16(3), 579-
610.
Dillenbourg, P., & Traum, D. (2000). The long road from a shared screen to
a shared understanding. Retrieved March 13, 2005, from
http://tecfa.unige.ch/tecfa/publicat/dil-papers-2/Dil.7.3.29.pdf
Dix, A., Cheverst, K., Fitton, D., & Friday, A. (2004). The auditability of
public space - approaching security through social visibility. Paper
presented at the 2nd UK-UbiNet Workshop, Security, trust, privacy and
theory for ubiquitous computing., University of Cambridge, UK.
239
supported cooperative work (pp. 107-114). Toronto, Ontario, Canada ACM
Press.
Edwards, W. K., Newman, M. W., Sedivy, J., & Izadi, S. (2002). Challenge:
recombinant computing and the speakeasy approach. In Proceedings of the
8th annual international conference on Mobile computing and networking
(pp. 279-286). Atlanta, Georgia, USA: ACM Press.
Ellis, C. A., Gibbs, S. J., & Rein, G. (1991). Groupware: some issues and
experiences Commun. ACM 34 (1), 39-58.
Elrod, S., Bruce, R., Gold, R., Goldberg, D., Halasz, F., Janssen, W., et al.
(1992). Liveboard: a large interactive display supporting group meetings,
presentations, and remote collaboration In Proceedings of the SIGCHI
conference on Human factors in computing systems (pp. 599-607).
Monterey, California, United States ACM Press.
Epstein, R., & Rogers, J. (2001). The big book of motivation games : quick,
fun activities for energizing people at work and at home. New York;
London: McGraw-Hill.
Fass, A., Forlizzi, J., & Pausch, R. (2002). MessyDesk and MessyBoard:
two designs inspired by the goal of improving human memory In
Proceedings of the conference on Designing interactive systems: processes,
240
practices, methods, and techniques (pp. 303-311). London, England ACM
Press.
Ferscha, A., Kathan, G., & Vogl, S. (2002). Webwall -an architecture for
public display www services. Paper presented at the Eleventh International
World Wide Web Conference, Hawaii, USA.
Foster, G., & Stefik, M. (1986). Cognoter: theory and practice of a colab-
orative tool In Proceedings of the 1986 ACM conference on Computer-
supported cooperative work (pp. 7-15). Austin, Texas ACM Press.
Fox, A., Johanson, B., Hanrahan, P., & Winograd, T. (2000). Integrating
Information Appliances into an Interactive Workspace IEEE Comput.
Graph. Appl. , 20 (3), 54-65.
Francik, E., & Akagi, K. (1989). Designing a computer pencil and tablet for
handwriting. Paper presented at The Human Factors Society 33rd Annual
Meeting.
Francik, E., Rudman, S. E., Cooper, D., & Levine, S. (1991). Putting
innovation to work: adoption strategies for multimedia communication
systems Commun. ACM 34 (12), 52-63.
241
Proceedings of the 2000 ACM conference on Computer supported
cooperative work (pp. 21-30). Philadelphia, Pennsylvania, United States
ACM Press.
Gaver, W. W., Sellen, A., Heath, C., & Luff, P. (1993). One is not enough:
multiple views in a media space In Proceedings of the SIGCHI conference
on Human factors in computing systems (pp. 335-341). Amsterdam, The
Netherlands ACM Press.
Gehl, J. (1987). Life between buildings : using public space. New York:
Van Nostrand Reinhold.
242
Greenberg, S., & Rounding, M. (2001). The notification collage: posting
information to public and personal displays In Proceedings of the SIGCHI
conference on Human factors in computing systems (pp. 514-521). Seattle,
Washington, United States ACM Press.
Grudin, J. (1988). Why CSCW applications fail: problems in the design and
evaluation of organization of organizational interfaces In Proceedings of the
1988 ACM conference on Computer-supported cooperative work (pp. 85-
93). Portland, Oregon, United States ACM Press.
Guimbretiere, F., Stone, M., & Winograd, T. (2001). Fluid interaction with
high-resolution wall-size displays. In Proceedings of the 14th annual ACM
symposium on User interface software and technology (pp. 21-30). Orlando,
Florida: ACM Press.
Hall, E. T. (1966). The hidden dimension. Man's use of space in public and
private: London.
243
Hardin, G. (1968). The Tragedy of the Commons. Science, 162, 1243-1248.
Houde, S., Bellamy, R., & Leahy, L. (1998). In search of design principles
for tools and practices to support communication within a learning
community SIGCHI Bull. , 30 (2), 113-118.
Huang, E. M., & Mynatt, E. D. (2003). Semi-public displays for small, co-
located groups In Proceedings of the conference on Human factors in
computing systems (pp. 49-56). Ft. Lauderdale, Florida, USA ACM Press.
Huang, E. M., Tullio, J., Costa, T. J., & McCarthy, J. F. (2002). Promoting
awareness of work activities through peripheral displays In CHI '02
extended abstracts on Human factors in computing systems (pp. 648-649).
Minneapolis, Minnesota, USA ACM Press.
Hughes, R. (1991). The shock of the new : art and the century of change.
London: Thames and Hudson.
244
Hutchins, E. (1995). Cognition in the wild. Cambridge, Mass.; London: MIT
Press.
Huysman, M., Steinfield, C., Jang, C.-Y., David, K., Veld, M. H. I. T., Poot,
J., et al. (2003). Virtual Teams and the Appropriation of Communication
Technology: Exploring the Concept of Media Stickiness Comput. Supported
Coop. Work 12 (4), 411-436.
Inkpen, K. M., Ho-Ching, W., Kuederle, O., Scott, S. D., & Shoemaker., G.
B. D. (1999). This is fun! We're all best friends and we're all playing :
Supporting children's synchronous collaboration. Paper presented at the
Computer Support for Collaborative Learning (CSCL) '99, Stanford,
California.
Izadi, S., Brignull, H., Rodden, T., Rogers, Y., & Underwood, M. (2003).
Dynamo: a public interactive surface supporting the cooperative sharing and
exchange of media In Proceedings of the 16th annual ACM symposium on
User interface software and technology (pp. 159-168). Vancouver, Canada
ACM Press.
Izadi, S., Brignull, H., Rodden, T., Rogers, Y., & Underwood, M. (2003b).
Dynamo: a public interactive surface supporting the cooperative sharing and
exchange of media [Video Recording] In Proceedings of the 16th annual
ACM symposium on User interface software and technology Vancouver,
Canada ACM Press.
245
Jacobson, R. E. (1999). Information design. Cambridge, Mass.; London:
MIT Press.
Johanson, B., Fox, A., & Winograd, T. (2002). The Interactive Workspaces
Project: Experiences with Ubiquitous Computing Rooms IEEE Pervasive
Computing 1 (2), 67-74.
Jones, M., Marsden, G., Mohd-Nasir, N., & Boone, K. (1999). Improving
web interaction on small displays. Computer Networks, 31, 1129-1137.
Kerka, S. (2000). Incidental Leaning: Trends and Issues Alert no. 18.
Retrieved March, 2005, from http://www.cete.org/acve/docs/tia00086.pdf
Koch, A., Monaci, S., Cabrera, A. B., Huis Veld, M., & Andronico, P.
(2004). Communication And Matchmaking Support For Physical Places Of
Exchange. Paper presented at the Intl. Conf. on Web Based Communities
(WBC2004), Lisbon, Portugal.
246
(Eds.), Intellectual teamwork : social and technological foundations of
cooperative work. Hillsdale, N.J.: L. Erlbaum Associates.
Kraut, R. E., Fish, R. S., Root, R. W., & Chalfonte, B. L. (1990). Informal
communication in organizations: Form, function, and technology. In S.
Oskamp & S. Spacapan (Eds.), Human Reactions to Technology: The
Claremont Symposium on Applied Social Psychology. Beverly Hills, CA:
Sage Publications.
Kraut, R. E., Gergle, D., & Fussell, S. R. (2002). The use of visual
information in shared visual spaces: informing the development of virtual
co-presence In Proceedings of the 2002 ACM conference on Computer
supported cooperative work (pp. 31-40). New Orleans, Louisiana, USA
ACM Press.
Luff, P., & Jirotka, M. (1998). Interactional Resources for the Support of
Collaborative Activities: Common Problems in the Design of Technologies
to Support Groups and Communities. In Community Computing and
Support Systems, Social Interaction in Networked Communities (pp. 249-
266): Springer-Verlag.
247
Mankoff, J., & Dey, A. K. (2003). From Conception to Design: A Practical
Guide to Designing and Evaluating Ambient Displays. In K. O’Hara, M.
Perry, E. Churchill & D. Russell (Eds.), Public and Situated Displays.
Social and Interactional Aspects of Shared Display Technologies. London:
Kluwer.
McCarthy, J. F., McDonald, D. W., Soroczak, S., Nguyen, D. H., & Rashid,
A. M. (2004). Augmenting the social space of an academic conference In
Proceedings of the 2004 ACM conference on Computer supported
cooperative work (pp. 39-48). Chicago, Illinois, USA ACM Press.
Milgram, S., Sabini, J., & Silver, M. (1992). The individual in a social
world : essays and experiments (2nd ed. / edited by John Sabini, Maury
Silver. ed.). New York; London: McGraw-Hill.
248
Moran, T. P., & Anderson, R. J. (1990). The workaday world as a paradigm
for CSCW design In Proceedings of the 1990 ACM conference on
Computer-supported cooperative work (pp. 381-393). Los Angeles,
California, United States ACM Press.
Moran, T. P., Saund, E., Melle, W. V., Gujar, A. U., Fishkin, K. P., &
Harrison, B. L. (1999). Design and technology for Collaborage:
collaborative collages of information on physical walls In Proceedings of
the 12th annual ACM symposium on User interface software and technology
(pp. 197-206). Asheville, North Carolina, United States ACM Press.
Morris, M. R., Ryall, K., Shen, C., Forlines, C., & Vernier, F. (2004).
Beyond "social protocols": multi-user coordination policies for co-located
groupware In Proceedings of the 2004 ACM conference on Computer
supported cooperative work (pp. 262-265). Chicago, Illinois, USA ACM
Press.
Myers, B. A., Stiel, H., & Gargiulo, R. (1998). Collaboration using multiple
PDAs connected to a PC In Proceedings of the 1998 ACM conference on
Computer supported cooperative work (pp. 285-294). Seattle, Washington,
United States ACM Press.
Mynatt, E. D., Huang, E. M., Voida, S., & MacIntyre, B. (2003). Large
displays for knowledge work. In K. O’Hara, M. Perry, E. Churchill & D.
Russell (Eds.), Public and Situated Displays. Social and Interactional
Aspects of Shared Display Technologies (pp. 80-102). London: Kluwer.
249
Nardi, B. A. (1993). A small matter of programming : perspectives on end
user computing. Cambridge, Mass.; London: MIT Press.
Nardi, B. A., Schwarz, H., Kuchinsky, A., Leichner, R., Whittaker, S., &
Sclabassi, R. (1993). Turning away from talking heads: the use of video-as-
data in neurosurgery In Proceedings of the SIGCHI conference on Human
factors in computing systems (pp. 327-334). Amsterdam, The Netherlands
ACM Press.
Newstrom, J. W., & Scannell, E. E. (1998). The big book of team building
games : trust-building activities, team spirit exercises, and other fun things
to do. New York; London: McGraw-Hill.
Nunamaker, J. F., Dennis, A. R., Valacich, J. S., Vogel, D., & George, J. F.
(1991). Electronic meeting systems Commun. ACM 34 (7), 40-61
Nunamaker, J. F., Dennis, A. R., Valacich, J. S., Vogel, D., & George, J. F.
(1991). Electronic meeting systems Commun. ACM 34 (7), 40-61.
O'Brien, L. (1989). Will the fax boom go bust? Telephony, 217(14), 38-46.
250
at the Workshop on Managing Tacit Knowledge at ECSCW 2001, Bonn,
Germany.
O'Hara, K., Churchill, E., Perry, M., Russell, D., & Streitz, N. (2002).
Public, Community and Situated Displays: Design, use and interaction
around shared information displays. Workshop at the Computer Supported
Cooperative Work Conference 2002 (CSCW'02), New Orleans.
O'Hara, K., Perry, M., & Lewis, S. (2003). Situated Web Signs and the
ordering of social action. In K. O’Hara, M. Perry, E. Churchill & D. Russell
(Eds.), Public and Situated Displays. Social and Interactional Aspects of
Shared Display Technologies (pp. 233-260). London: Kluwer.
O'Neill, E., Woodgate, D., & Kostakos, V. (2004). Easing the wait in the
emergency room: building a theory of public information systems In
Proceedings of the 2004 conference on Designing interactive systems:
processes, practices, methods, and techniques (pp. 17-25). Cambridge, MA,
USA ACM Press.
251
Paradiso, J., Leo, C. K., Checka, N., & Hsiao, K. (2002). Passive Acoustic
Knock Tracking for Interactive Windows. In Proceedings of the Human
Factors in Computing Systems Conference (p. 732-733), Minneapolis, USA.
Paulos, E., & Goodman, E. (2004). The familiar stranger: anxiety, comfort,
and play in public places In Proceedings of the 2004 conference on Human
factors in computing systems (pp. 223-230). Vienna, Austria ACM Press.
Pedersen, E. R., McCall, K., Moran, T. P., & Halasz, F. G. (1993). Tivoli:
an electronic whiteboard for informal workgroup meetings In Proceedings
of the SIGCHI conference on Human Factors in Computing Systems (pp.
391-398). Amsterdam, The Netherlands ACM Press.
Prante, T., Stenzel, R., Röcker, C., Streitz, N., & Magerkurth, C. (2004).
Ambient agoras: InfoRiver, SIAM, Hello.Wall In Extended abstracts of the
2004 conference on Human factors and computing systems (pp. 763-764).
Vienna, Austria ACM Press.
252
Putnam, R. D. (2000). Bowling alone : the collapse and revival of American
community. New York; London: Simon & Schuster.
Rajani, R., & Perry, M. (2000). Medical Work as Practised: implications for
the design of telemedicine and medical information systems. Paper
presented at the SIHI (Southern Institute for Health Informatics) 3rd Annual
Conference, University of Portsmouth, UK.
Rodden, T., Rogers, Y., Halloran, J., & Taylor, I. (2003). Designing novel
interactional workspaces to support face to face consultations In
Proceedings of the conference on Human factors in computing systems (pp.
57-64). Ft. Lauderdale, Florida, USA ACM Press.
253
Rogers, E. M. (1962). Diffusion of innovations. New York: Free Press of
Glencoe; London : Macmillan.
Rogers, Y., & Bellotti, V. (1997). Grounding blue-sky research: how can
ethnography help? Interactions 4 (3), 58-63.
Rogers, Y., Hazlewood, W., Blevis, E., & Lim, Y.-K. (2004). Finger talk:
collaborative decision-making using talk and fingertip interaction around a
tabletop display In Extended abstracts of the 2004 conference on Human
factors and computing systems (pp. 1271-1274). Vienna, Austria ACM
Press.
Rudström, A., & Höök, K. (2003). Designing a Mobile Social Service for a
Mall: User experiences of Kista Galleria (No. T2003:26): SICS.
254
Russell, D., & Sue, A. (2003). Large interactive public displays: Use
patterns, support patterns, community patterns. In K. O’Hara, M. Perry, E.
Churchill & D. Russell (Eds.), Public and Situated Displays. Social and
Interactional Aspects of Shared Display Technologies (pp. 3-17). London:
Kluwer.
Shen, C., Lesh, N. B., Vernier, F., Forlines, C., & Frost, J. (2002). Sharing
and building digital group histories In Proceedings of the 2002 ACM
conference on Computer supported cooperative work (pp. 324-333). New
Orleans, Louisiana, USA ACM Press.
255
Smith, J. H. (2002). The Architectures of Trust - Supporting Cooperation in
the Computer-Supported Community. Unpublished MA Thesis, University
of Copenhagen.
Smith, R. B., O'Shea, T., O'Malley, C., Scanlon, E., & Taylor, J. (1991).
Preliminary experiments with a distributed, multi-media, problem solving
environment In Studies in computer supported cooperative work: theory,
practice and design (pp. 31-48): North-Holland Publishing Co. .
Stefik, M., Foster, G., Bobrow, D. G., Kahn, K., Lanning, S., & Suchman,
L. (1987). Beyond the chalkboard: computer support for collaboration and
problem solving in meetings Commun. ACM 30 (1), 32-47.
Stewart, J., Bederson, B. B., & Druin, A. (1999). Single display groupware:
a model for co-present collaboration In Proceedings of the SIGCHI
conference on Human factors in computing systems: the CHI is the limit
(pp. 286-293). Pittsburgh, Pennsylvania, United States ACM Press
Streitz, N., Geisler, J., Holmer, T., Konomi, S., Muller-Tomfelde, C.,
Reischl, W., et al. (1999). i-LAND: an interactive landscape for creativity
and innovation. In Proceedings of the SIGCHI conference on Human
factors in computing systems: the CHI is the limit (pp. 120-127). Pittsburgh,
Pennsylvania, United States: ACM Press.
256
Paper presented at the 2nd BCS HCI Group Workshop on Culture and HCI:
Bridging Cultural and Digital Divides, London.
Tani, M., Horita, M., Yamaashi, K., Tanikoshi, K., & Futakawa, M. (1994).
Courtyard: integrating shared overview on a large screen and per-user detail
on individual screens In Conference companion on Human factors in
computing systems (pp. 202). Boston, Massachusetts, United States ACM
Press.
Tatar, D. G., Foster, G., & Bobrow, D. G. (1991). Design for conversation:
lessons from Cognoter. International Journal of Man-Machine Studies,
34(2), 185-209.
257
Twidale, M. B., Nichols, D. M., Smith, G., & Trevor, J. (1995). Supporting
collaborative learning during information searching In The first
international conference on Computer support for collaborative learning
(pp. 367-370). Indiana Univ., Bloomington, Indiana, United States
Lawrence Erlbaum Associates, Inc.
Veinott, E. S., Olson, J., Olson, G. M., & Fu, X. (1999). Video helps remote
work: speakers who need to negotiate common ground benefit from seeing
each other In Proceedings of the SIGCHI conference on Human factors in
computing systems: the CHI is the limit (pp. 302-309). Pittsburgh,
Pennsylvania, United States ACM Press.
Voida, S., Mynatt, E. D., MacIntyre, B., & Corso, G. M. (2002). Integrating
Virtual and Physical Context to Support Knowledge Workers IEEE
Pervasive Computing 1 (3), 73-79.
Want, R., Pering, T., Danneels, G., Kumar, M., Sundar, M., & Light, J.
(2002). The Personal Server: Changing the Way We Think about
Ubiquitous Computing In Proceedings of the 4th international conference
on Ubiquitous Computing (pp. 194-209). Göteborg, Sweden Springer-
Verlag.
Wasserman, S., & Faust, K. (1994). Social network analysis : methods and
applications. Cambridge: Cambridge University Press.
258
Wellman, B., & Berkowitz, S. D. (1988). Social structures : a network
approach. Greenwich, Conn.; London: JAI, 1997.
Whittaker , S., & Schwarz, H. (1995). Back to the future: pen and paper
technology supports complex group coordination. In Proceedings of CHI’95
(pp. 495-502) NY: ACM Press.
Whittaker, S., & Schwarz, H. (1999). Meetings of the Board: The Impact of
Scheduling Medium on Long Term Group Coordination in Software
Development Comput. Supported Coop. Work 8 (3), 175-205.
259
Appendix 1
260
A1.1 Semi-structured interview
A1.1.1 Background
Name:
Gender:
Age
2) What did you know about Opinionizer before you had a go on it? How
did you learn about it?
261
3) Did you watch others using it first before having a go yourself? Did it
look easy or complicated to use?
5) (For non-users) Why didn’t you try Opinionizer? Were you aware of its
existence? Did you know what it was? Did you watch others using it and
then feel it was not for you?
8) What, in your opinion, was the one best and the one worst aspect of
Opinionizer?
262
Appendix 2
263
A2.1 Participant Consent Forms & Disclaimer
The form on the following page is the ‘in loco parentis’ consent form. This
was for the head teacher to sign, as a legal guardian of the students while at
the school.
264
IN LOCO PARENTIS – DYNAMO STUDY CONSENT FORM
By signing below you are indicating that you are an ‘in loco parentis’ guardian
of the students at Blatchington Mill Sixth Form college, and that you
understand and agree with the following terms and conditions:
[1] The Dynamo researchers will be recording video in the 6th form common
room of students using or carrying out activities near the Dynamo system.
[2] Images or segments form these videos could be used by the researchers in
the reporting of the study to the academic community. This might be in the
form of academic papers, conference presentations and project website content.
[4] The researchers will not use content from this video in any other public
non-academic forum unless the explicit consent of the people involved has first
been obtained.
[5] The researchers undertake to store the video in appropriate locked facilities
where access will be limited to those within the project.
[6] The students and teachers can choose to withdraw from this study at any
time without penalty.
[7] The users of the Dynamo system will each sign a disclaimer form, which
states that they take full personal responsibility for any actions when using the
Dynamo system. It also states that any information or files displayed or shared
through Dynamo is the responsibility of the individual participants themselves,
and disclaims the universities of Sussex and Nottingham from responsibility.
If you have any questions about this study, please contact Harry Brignull
[01273 877221, harrybr@sussex.ac.uk]. If you have any other concerns about
the study or the way it is being run, please contact Dr. Geraldine Fitzpatrick:
01273 678982, geraldin@sussex.ac.uk
I consent to the conduct of the Dynamo according to the above terms on behalf
of the senior college students and teachers.
265
DATE:
266
A2.1.2 Participant Consent form
267
Blatchington Mill Informed Consent Form
By signing below, I confirm hereby that I am participating voluntarily in
this technology field study, and that I understand the following terms &
conditions:
(1) I may be video taped while using or nearby the Dynamo system.
(2) These images could be used by the researchers when they report on this
study to the academic community, e.g., in academic papers and on the
project web page. My real name will not be used in any reporting of this
work.
(3) These images will not be used in any other public non-academic forum
without your explicit consent.
(4) I can choose to withdraw from the study at any time without penalty.
(5) I have read and signed a copy of the Dynamo Disclaimer document,
which states that I take full personal responsibility for any actions I carry
out when using the Dynamo system.
If you have any questions about this study, please contact Harry Brignull
[01273 877221, harrybr@sussex.ac.uk]. If you have any other concerns about
the study or the way it is being run, please contact Dr. Geraldine Fitzpatrick:
01273 678982, geraldin@sussex.ac.uk
268
269
A2.1.3 Disclaimer Form
The form on the following page is the legal disclaimer, which details how
that when signed, the participants take full responsibility for their actions
while using Dynamo, such as the sharing of copyrighted materials and the
posting of anti-social or obscene materials on Dynamo.
270
DYNAMO LEGAL DISCLAIMER
Please sign at the bottom to indicate that you have read this document and understand it.
The Universities of Sussex and Nottingham do not condone activities and actions that
breach the rights of copyright owners. It is your responsibility to obey all laws governing
copyright materials and property. The Universities of Sussex and Nottingham respect
copyright and other laws.
Be warned in the event a user of Dynamo fails to comply with laws governing copyrighted
property such user may be exposed to criminal or civil liability which may include possible
fines or imprisonment.
By signing this disclaimer you are indicating that you take full personal responsibility for
your actions when using or acting in relation to the Dynamo system. You are also agreeing
to indemnify and hold the Universities of Sussex and Nottingham and its agents and
employees harmless from any claim or demand, including reasonable solicitors fees made
by any third party due to or arising out of your use of the Dynamo system.
The files, information and opinions displayed on the Dynamo system are the product of the
participants, each of whom has signed this document, taking full personal responsibility for
their actions. The Universities of Sussex and Nottingham disclaim all responsibility for any
files, information and opinions displayed on the Dynamo system.
By signing this disclaimer, you are indicating that you understand that it is your own
responsibility to protect yourself from computer viruses or other malware. The Universities
of Sussex and Nottingham disclaim all responsibility for any damages or losses (including,
without limitation, financial loss, damages for loss in business projects, loss of profits or
other consequential losses) arising in contract, tort or otherwise from the use of or inability
to use the Dynamo system or any material appearing on the Dynamo system, or from any
action or decision taken as a result of using the Dynamo system or any such material. The
Dynamo system contains hyperlinks to external websites. The Universities of Sussex and
Nottingham is not responsible for and has no control over the content of such sites.
Information on the Dynamo system, or available via hypertext link from the Dynamo
system, is made available without responsibility on the part of The Universities of Sussex
and Nottingham. The Universities of Sussex and Nottingham disclaim all responsibility and
liability (including for negligence) in relation to information on or accessible from the
Dynamo system.
Signature: ……..………..………..
Date: ……..………..………..
271
A2.2 Preliminary Survey
The preliminary survey was given out during the preliminary observational
study carried out prior to the deployment of Dynamo. It is shown on the
following pages.
272
Quick Questionnaire
273
On average, how long do you use computers for
each day?
0-1hrs 2-3hrs 3-4hrs 4-5hrs 5-6hrs 6-7hrs
274
Please state whether you regularly do each of
the following activities.
Activity YES NO
SMS messages
MMS messages
(containing images, video, sounds or ringtones)
Bluetooth-based file exchange using mobile
phones
(e.g. images, video, sound or ringtones)
Text emails
Email attachments
Photo-Sharing websites
(e.g. shutterfy.com, photobox.co.uk)
Instant messaging
(e.g. MSN messenger)
Shared drives or FTP servers
(e.g. Yahoo Briefcase, Apple iDisk)
If you use any other communiction or file sharing tools, then list them
here (e.g. Blogs, Wikis, etc)
275
Please state whether you regularly use
each of the following devices while at
college.
Device YES NO
Laptop
MP3 Walkman
USB pendrive
Floppy disk
Zip disk
Recordable CD
276
Your Personal Details
Your Name
Gender
Age
Email
(Optional - used only to send you occasional messages about
Dynamo over the next 2 weeks or so)
Courses studied
277
A2.3 Post Hoc Survey
The post hoc survey was given out at the end of the study in order to
ascertain details of participant use and opinion. It is shown on the following
pages.
278
Second Quick Questionnaire
279
When did you first use Dynamo?
Friday Monday Tuesday Wednesday Thursday Friday
5th 8th 9th 10th 11th 12th
If yes, When?
Friday Monday Tuesday Wednesday Thursday Friday
5th 8th 9th 10th 11th 12th
280
Could you give a reason for why you used
Dynamo the amount you did?
281
Could say in two sentences how you think
Dynamo effected life in the common room?
Your Name
Gender
Age
Email
(Optional - used only to send you occasional messages about
Dynamo over the next 2 weeks or so)
282
A2.4 Instruction leaflet
The leaflet shown on the following two pages was made available in a
prominent leaflet dispenser next to the Dynamo installation. Please note that
it has been scaled down to fit on this page.
283
284
285
286