Building Bridges - E-Infrastructures and Technologies for Lifelong Learning - page 248

Facilitating Interaction
and Awareness

of these technologies (e.g., group-group interac-
tion for videoconferencing vs. individual-group
interaction for webcasts) and different types of
entities (individuals, groups, etc.) must be con-
sidered and accounted for.

As the data suggest, the modified ePresence
system was effective in technically bridging the
conferencing and webcast systems, and facilitat-
ing interaction. Once conversations started (either
via voice or text), it was possible for participants
using all three participation options to interact
with the instructor. The most difficult aspect of
the system, however, proved to be supporting
effective instructor-student awareness. Aware-
ness allows instructors to monitor students for
comprehension, and also to tell when a satellite
or remote participant had a question.
While our system did provide basic aware-
ness, we were not successful in providing detailed
awareness information about students at the satel-
lite campus or remote participants. Our speakers
were able to see that there were students there,
and roughly how many there were, but wished
they could see detailed facial reactions and other
nonverbal cues. This is, in many ways, not sur-
prising given that awareness is a persistent theme
(and difficulty) in CSCW research (Schmidt, 2002)
and the efforts that other researchers in this area
have put into awareness in distributed lectures
(Chen, 2001). We present two specific lessons
for designers.
First is the type of entity. In bridging videocon-
ferencing and webcasting, our design takes a very
simple approach. All users, whether aggregations
of individuals at the satellite campus or individuals
signed in from home, are treated essentially the
same, and can be viewing either the webcast or
be included in the videoconference. This design
decision turned out to be problematic in that, for
groups, detail on the awareness display was in-
adequate, and individual identifying information
was not provided. For individuals, their images
often seemed unnecessarily large in comparison
with the entire satellite campus. This suggests that,
in a system that builds dynamic bridges between
multiple technologies, the different usage styles

Design principle 1: Visual representations of
remote participants must consider the variation
in size and scope of the remote audience entities
they represent.

In its most basic form, this principle could
be implemented by allowing for easy manual
reconfiguration of video images of others. Given
that users often do not take the time to manually
reconfigure their views, however, a more nuanced
approach could use face detection to gauge the
number of likely people in a video view, and
scale the view accordingly. These views could
also be treated and displayed differently based on
instructor or student preferences, and depending
on how many people they contained. Serif et al.
(2009), for example, describe shared room and
from-home participation modes. Modes appropri-
ate to the given setting could be pre-configured
for each application.
Second, the dynamic nature of the bridges our
system builds between a webcast and videoconfer-
ence presents a challenge in providing awareness
information. Participants in the webcast need not
always (and often will not) be live participants in
the voice and video conversation. This raises the
question of how to best represent webcast-only
(i.e., those not currently in the video and voice
conversation) participants on the awareness dis-
play so that the presenter knows they are there,
but they are nonetheless visually distinct from
those participants who are in the video and voice
conversation, and will be more active. Our data
suggest that awareness of both of these groups is
important, but scarce screen space and the need to
minimize cognitive loads on the presenter make
this a difficult design challenge. This challenge is
addressed more fully in Birnholtz, et al. (2008).