Journal of Systems & Information Technology 3(1)

CONTRIBUTION BY PARTICIPANTS IN FACE-TO-FACE

BUSINESS MEETINGS: IMPLICATIONS FOR
COLLABORATIVE TECHNOLOGY.

Robert J. McQueen
Department of Management Systems
University of Waikato
Hamilton, New Zealand

Karen Rayner
Department of Management Systems
University of Waikato
Hamilton, New Zealand

Ned Kock
Department of Computer Information Systems
Temple University
Philadelphia, USA

ABSTRACT

Face-to-face business meetings are a widely used method of group interaction,

and a rich source of data on what actually happens in group discussions. Active
participation in a meeting is usually perceived to be making an oral contribution of
some kind to the discussion. This paper describes a field study of ten face-to-face
business meetings which were videotaped and subsequently analysed. Participant
contributions were coded, and the data summarized. The mean contribution was
approximately 12 seconds and 18 words. The most common contribution type was
information giving. The highest single contributor in each meeting captured, on
average, about 30% of the available airtime, while the two highest, combined, captured
over half of the airtime. These findings are discussed within the context of requirements
for designers of collaborative technology systems to support group interpersonal
communication through the use of computing and data communication technologies.

15
Journal of Systems & Information Technology 3(1)

CONTRIBUTION BY INDIVIDUALS IN MEETING CONTEXTS

Our focus in this paper is the contribution by individuals to a face-to-face group

meeting. Contribution in this context is generally considered to be either a verbalisation
or an act, usually bounded by either silence or another’s contribution (Lebie et al. 1996;
Weingart 1997). Clark and Brennan (1991) hold that conversation has easily identifiable
entries, bodies and exits, aiding analysis considerably. We wish to link the findings of a
study of contribution in face-to-face meetings which follows with design issues of
Collaborative Technology (CT) systems for the support of group communication.

We will first present a conceptual model on contribution in group discussions, and

discuss the factors suggested, along with what has been reported in previous work on
group discussion. We then introduce the results of a field study of face-to-face group
meetings, identifying profiles of contribution, use of visual aids and dominant speakers.
The results of this study are then extrapolated into implications for the growing use of
CT supported group communication, which suggests consideration of factors such as
media type (face-to-face or electronic), media parameters (bandwidth, bi-directional
channels, interactivity), temporality (asynchronous or synchronous) and location (co-
located or remote). We conclude with the implications of this study for designers of CT
systems who wish to be aware of the factors that may influence contribution.

FACTORS INFLUENCING CONTRIBUTION IN GROUP DISCUSSIONS

To guide this paper, we suggest that contribution in a group discussion follows the
three stages below, illustrated in the conceptual model given in figure 1. We developed
these three stages with the main goal of highlighting relevant factors influencing
contribution in group discussions. It is not our goal, however, to propose these stages as
elements of a new group development model (see Tuckman, 1970).The stages we
suggest are:

1. Being “co-present” indicates the first stage of being connected with the other
members of the group and engaged, at least as a listener, in the group dialog. The
opportunity to listen to the contributions of others in the discussion usually requires
either physical presence, in the case of face-to-face meetings, or connectivity as in
the case of CT supported discussions.

2. Being motivated to contribute (rather than just passively listen) is the second stage
along the path to contribution. This stage is influenced by individual and group
constraints and incentives, and also by the issue under discussion and the cultural
environment of the discussion.

16
Journal of Systems & Information Technology 3(1)

Group constraints
& incentives
-turn sharing protocols Media
-status and hierarchy -ease of contributing
-cultural protocols -interactivity
-chairperson -reward for effort
management -suitability for task
-process blocks

Co-presence Motivation to Act of

with other contribute contribution
group members

Physical presence
possible?
-Travel, time available
-Electronic access
Issue under discussion
-knowledge about issue
-personal impact of
Individual constraints
decision
& incentives
-confidence
-experience
Cultural constraints -stimulation by other contributions
& incentives -affiliation intensity to group
-rank -alignment with group ideas
-gender -skills with media
information privacy

Figure 1 Conceptual model of the process of contribution to a group discussion

3. Undertaking the physical act of contributing, either by speaking in a face-to-face

meeting or by creating an electronic message to be contributed in a CT supported
discussion, is the third stage of our model. Completion of an act of contribution will
again be influenced by individual and group constraints and incentives, and also by
the medium of communication being used.

These stages are influenced by several factors.

17
Journal of Systems & Information Technology 3(1)

Individual constraints and incentives to contribution

Participation in a group situation is largely dependent on the nature of the
individual. Studies have shown a number of individual-level variables impact
willingness or ability to contribute in a discussion. These include: past experiences in
participation; attitudes towards contributing (Ajzen and Fishbein 1980); familiarity with
the other group members (Campion et al.1993; Guzzo and Dickson 1996; Jehn and Shah
1997); familiarity with the task or the subject matter (Littlepage et al.1997); personality
and values (Barrick et al. 1998; Yellen et al. 1995); levels of communication
apprehension, reticence, or unwillingness to communicate (McCroskey 1977); level of
influence or status (Bettenhausen 1991; Dubrovsky et al.1995; Nunamaker et al. 1991a;
Sproull and Kiesler 1991; Weisband 1994); uniqueness or non-commonality of
information held and anticipated reaction to it (Connolly et al.1990; Diehl and Stroebe
1987; Jay 1976; Nunamaker et al. 1991b); and position in a hierarchy of needs, values
or status (Goodall 1990). Dominant modes of participation in group discussions may be
a used as a mechanism to exert political influence, or exert power on others. Other
factors hampering effective communication within groups include territorial conflicts,
interpersonal conflicts, groupthink, burnout, withdrawal, domination, aggressiveness
(Goodall 1990), free riding or social loafing (Bettenhausen 1991; Diehl and Stroebe
1987).

Some individuals are less likely to contribute in a group situation, and a number of
theories have been advanced for this. Introverted, low status or low influence individuals
are less likely to participate, as are those who have had unfavourable experiences in
groups in the past. Domination by one or two (often extroverted or high status) speakers
is common, leading to negative expectations by other group members, and reduced
participation. This has negative implications for future group performance, as more
talkative individuals are judged more favourably and receive more positive feedback
than others do; thereby encouraging them to continue contributing (Cappella, 1979).

Group constraints and incentives to contribution

Even given the willingness of an individual to participate in a discussion, there
may be external obstacles hampering ability. Process losses result in inferior group
performance, given the circumstances, nature of the task and the abilities of the group
(Argote and McGrath 1993; Diehl and Stroebe 1987; Finholt and Sproull 1990; George
et al. 1990; Medsker et al.1995). They arise from misunderstandings,
miscommunications and difficulties in group coordination (Jehn and Shah 1997).
Primary among these losses is production blocking; the result of only one speaker being
able to hold the floor at a time. Others present must hold back contributions until the
speaker has finished; at the risk of forgetting what they were going to say, or having the
contribution rendered irrelevant by changes in context. Production blocking is a result of
cognitive limitations – simply put, it’s difficult to think and listen at the same time
(Argote and McGrath 1993; Diehl and Stroebe 1987; Nunamaker et al. 1991b).
Those holding unique information may be hesitant to step forward with it unless
they sense they have the ability to influence the rest of the group with that information
(Festinger 1950). This can reduce participation levels as dominant speakers are likely to
share common information early in the meeting, where activity is at its greatest

18
Journal of Systems & Information Technology 3(1)

(Dubrovsky et al. 1991; Jay 1976). Others may hold back information due to a conflict
between individual and group goals (Bettenhausen 1991).

The more familiar a participant is with the other members of the group, the less
reticent he or she will be to share information. Fear of negative reactions is lessened, and
the group atmosphere tends to be perceived as less threatening. Friendship groups and
formal work groups with a history of interaction display different characteristics
(Campion et al. 1993; Jehn and Shah 1997), but information is more likely to be shared
when other members of the group are known.

Similarly, the more familiar an individual is with the subject under discussion or
the task being performed, the more likely he or she will actively participate. Individuals
are more likely to share information that is held in common by the other group members
than to share information only they hold. ‘The fear of negative evaluations from other
group members prevents subjects…from presenting their more original ideas’ (Diehl and
Stroebe 1987, p498). This can hamper group outcomes but is often an aid to the process.
Other factors include the size and composition of the group (Campion et al. 1993;
Finholt and Sproull 1990; George et al. 1990; Jay 1976; McKinlay et al. 1994;
Nunamaker et al. 1991b), the physical nature of the group or surroundings (Finholt and
Sproull 1990; Hirokawa and Gouran 1989; Weingart 1997; Weisband 1994).

Group size can have a marked effect on group performance. While most research
based in sociology and psychology has typically concentrated on the ‘smallest real social
group’ (McKinlay et al. 1994, p159) of 3 members, groups in organisations tend to have
higher memberships. Studies into business meetings have used group sizes ranging from
3 to 15 (Alavi 1993) with relatively consistent contribution rates across the groups
(Nunamaker et al. 1991a). Jay (1976) argues that between 4 and 7 members is ideal, 10
is tolerable, and 12 is the outside. Other studies have found that the larger the group, the
less opportunity each individual has to make a substantial contribution.

Weingart (1997) points out that the least active member of a 3 person group
contributes about 23% of the total, compared to the third most active member in a 10
person group at 10%, and the least active in a 10 person group at 3%. Group size is one
of the major contributors to process losses in group meetings. As the size of the group
increases, there is more likely to be one or two dominant speakers, more social loafing,
less satisfaction with the group process and increased coordination needs (Bettenhausen
1991; Campion et al. 1993; Finholt and Sproull 1990; George et al. 1990).

Cultural and Issue factors

The culture of the organization in terms of expected norms of behaviour and
communication is likely to have an impact on contribution in face-to-face meetings. The
cultural backgrounds of the participants, as individuals may also have an impact on
contribution processes (McQueen and Clark, 1999). The work of Gudykunst et. al.
(1996), based on Hofstede’s (1980) cultural dimensions, suggests that the cultural
individualism-collectivism dimension has a direct effect on communication, and that
collectivist cultures often prefer high context communication, such as face-to-face.
According to Hall (1976), in these high context environments, most of the information is

19
Journal of Systems & Information Technology 3(1)

either in the physical context or internalised in the person, while very little is in the
coded, explicit, transmitted part of the message. The issue under discussion, if
personally important to a participant, may motivate contribution, or if of little personal
consequence, is likely to dissuade that participant from contributing.

The media factor

The media used for a group discussion is likely to influence the number, duration,
and content of contributions. Media-centered theories of group communication place
particular emphasis on the fit between task (i.e. group discussion characteristics) and
technology as a determinant of communication process and outcomes. That is, the main
focus of these theories is the communication medium. Arguably, the best example of
media-centered theory is media richness theory (Daft and Lengel, 1986; Daft et al.,
1987; Lengel and Daft, 1988), which argues that rational individuals predictably favor
the use of specific communication media to accomplish certain tasks. According to
media richness theory, the "richness" of different communication media is static and
varies according to a richness scale that features FtF meetings at the top and printed
documents at the bottom, with most CT systems somewhere in between (Lee, 1994;
Markus, 1994). The "richness" of a medium, according to media richness theory, is
predominantly determined by the feedback immediacy that it provides and its ability to
convey non-verbal cues.

A key hypothesis of media richness theory is that rich media are more appropriate
to support "equivocal" communication (which is likely to occur in complex tasks) than
lean media, and that aggregate data about rational individual media choices would
consistently support this hypothesis. Yet, Carlson and Davis' s (1998) study shows that
individuals may chose to use a lean medium for communication primarily due to the
ease of contributing afforded by the medium. A corollary, proposed by Kock (1998), is
that individuals who choose to use leaner media for equivocal communication will adapt
their behavior in order to compensate for inherent media constraints by producing fewer,
longer (in number of words) and more elaborate (in terms of content) contributions.

THE FIELD STUDY ENVIRONMENT AND METHODS

The organization studied is a semi-autonomous government regional research

centre, employing approximately 400 people at the location studied. Permission was
sought, and obtained, from the organisation to automatically videotape a number of
business meetings over a four week period for the purpose of better understanding
whether proposed CT supported communication support tools for asynchronous group
discussions might be of benefit. A ceiling mounted surveillance type video camera
(exposed) and sound-activated video cassette recorder (hidden) were installed in each of
two meeting rooms, with respective seating capacities of 25 and 75 people. Meeting
room users were advised through the organization newsletter of the presence and
purpose of the video cameras.

20
Journal of Systems & Information Technology 3(1)

The data was coded at an off-site location by stop-start viewing of the recording
while keying information into a database on a microcomputer. A simple program was
created which created a file record, with timestamp, when the spacebar key was pressed
on a computer keyboard. A first pass of the tape was viewed, and a technician pressed
the space bar at the start of every speaking contribution for every participant. A
contribution is defined here as that duration of time between the start and end of a
contiguous segment of speech from a contributing participant in the meeting. This file of
sequential timestamps was then loaded into a database program, and additional
information was added in a second pass through the video tape, such as speaker number,
rate of speaking, function of the comment (i.e. opinion giving), linkages to previous
speakers, and use of gestures, emotion or visual aids by the speaker. The occasional
periods of silence that occurred between speakers were also recorded. Two additional
data fields were derived from the entered data: elapsed time of each contribution, and
total words spoken by the speaker in that contribution.

OBSERVATIONS

In the following, contribution refers to a segment of speech or silence, bounded in

turn either by silence or by a segment of speech from another person. Participant refers
to a person at the meeting who made at least one contribution. There may have been
other non-contributing attendees at the meeting.

Meeting summary data

For each meeting coded, the number of participants, number of contributions,
elapsed time of meeting (hours:minutes), and total number of words are shown below.

Meeting Participants Contributions Time Words

1 4 445 1:31 8921
2 9 377 2:51 *
3 6 995 2:24 15730
4 8 1119 2:29 17261
5 6 914 2:20 15528
6 8 382 0:50 6335
7 13 490 2:30 18025
8 13 113 1:46 *
9 10 227 1:00 *
10 7 311 0:32 3582
* not coded for these meetings

Table 1 Meeting Summary Data

The total number of contributions coded was 5373. Mean elapsed time per
contribution over all 5373 contributions was 12.2 seconds. Mean number of words per

21
Journal of Systems & Information Technology 3(1)

contribution, for those meetings where word counts were coded, was 18.3. Of the 5373
contributions, 566 were periods of silence or pauses.

Use of gestures
Gestures used by participants when speaking were coded into eight categories.
Most contributions had no gestures (92.4%).

Other non-verbal communication

Pauses (for effect) were used by participants in 2.5% of all speaking contributions.
No pauses were coded for 97.5%. Emotion was shown in only 5 of 4807 contributions.

Contribution function summary

All speaking contributions (4807) were tabulated by function, as coded on the
apparent verbal intent. For each function category, the mean number of words, and mean
seconds per contribution are shown below.

Function Contributions Mean words per Mean seconds per

(percent) contribution contribution
Words Seconds
info giving 55.8 20 14
query 14.0 9 6
opinion giving 11.9 16 15
decision making 5.9 24 12
informal talk 4.3 10 6
support 3.8 7 4
directional 3.0 31 19
consensus making 1.1 28 16
brainstorming 0.2 17 9

Table 2 Contribution Function

Use of visual aids

Use of non-verbal aids by each speaker is given below.

22
Journal of Systems & Information Technology 3(1)

Function Contributions Mean words per Mean seconds per

(percent) contribution contribution
no aids 80.1 16 11
handout 16.7 25 13
whiteboard 2.5 20 20
overhead trans. 0.7 21 123
35 mm slides <0.1 * *
* consisted of one contribution, 889 seconds in duration

Table 3 Visual Aids

Interactivity - linkage to other contributions

88.2% of contributions were linked to the immediately previous contribution. A
further 3.2% were linked to one of the previous five contributions, and 0.5% were linked
to earlier contributions. 8% were not linked to a previous contribution.

Dominant speakers
Meeting dominance can be represented by how frequently a participant "takes
over" by speaking, or the percent of overall "air time" used. Dominance can therefore be
represented by the percent of contributions by a participant, or the percent of total time
used. The following table shows in columns 3 and 4 the percentage of speaking
contributions and the percentage of total time by dominant (largest number of
contributions) speaker in each of the observed meetings. Columns 5 and 6 show the
equivalent percentages for the sum of the two most dominant speakers in the meeting.
Note the number of participants in the meeting in column 2.

---------top participant------ ---top two participants-

Meeting Participants Trans. Time Trans. Time
% % % %
1 4 40.5 48.2 72.3 81.3
2 9 18.3 22.3 35.4 52.0
3 6 30.1 29.3 52.4 55.2
4 8 27.6 29.6 47.1 52.4
5 6 28.9 31.2 54.1 58.4
6 8 24.9 37.5 42.4 50.1
7 13 30.0 48.9 51.8 68.2
8 13 21.9 24.2 41.7 49.4
9 10 22.0 21.4 39.5 41.6
10 7 26.9 20.2 51.9 38.8

Means 31.3 54.7

Table 4 Dominant Speakers

23
Journal of Systems & Information Technology 3(1)

DISCUSSION

Group size
Much of the sociological small group research has been done with groups of 3 and
4 (Bales, 1970). This study observed group sizes from four to thirteen, with the median
size being six, perhaps more typical of group meetings in organizational settings.

The larger group sizes were of particular interest in this study in examining the
potential insights for support of larger groups using CT. It has been reported (Hiltz,
1988) that the main advantages of electronic group communication over face-to-face
meetings were more likely to be observed in large (>10) rather than small (3-5) groups.

Speaking rates
A main concern of potential users of CT systems is that it seems to be much more
inhibiting to free-flowing discussion to be required to type rather than talk (Mayer,
1985). The data recorded by this study on face-to-face speaking rates (words per minute)
indicated an average rate of 113 words per minute. Current technology for CT support
of group communication is heavily keyboard based. The estimated data in the table
below may be of some help in putting this speaking rate in perspective.

words per minute

Speaking rate (this study) 113
Skilled touch typist 90
Touch typist 40
Two finger typist 20
Hunt and peck typist 10

Table 5 Speaking & Typing Rates

Additionally, a study by Kock (1998) of CT supported business meetings where

the subjects interacted asynchronously suggests an average contribution rate of 6 words
per minute. This indicates that many fewer words are likely to be generated in an
electronic meeting than in a face-to-face meeting. This can be due to the speed and skills
required to type the words in or, as argued by Kock (1998), to behavior aimed at
compensating for the limitations of a leaner medium.

Characteristics of contributions
Each change of speaker in a face-to-face meeting generates a new contribution.
This contribution concept would compare to the distinct typed text messages or
comments entered by participants in an CT supported communication system. A number
of interesting data were collected about the characteristics of these contribution
segments in the observed face-to-face meetings.

The observed mean duration of contribution length for each meeting analysed
ranged from 6 seconds to 57 seconds, with the mean of all contributions from all

24
Journal of Systems & Information Technology 3(1)

meetings being about 12 seconds. This indicates that the nature of each spoken
contribution is typically one of a relatively short duration.

The number of words per contribution ranged from means of 12 to 37 for each of
the meetings observed, with an overall mean of 18 words. When compared to typed lines
of text, this represents typically one to three lines of typed words. This perhaps offsets to
some degree the concern over limited typing speed, for even the slowest hunt and peck
typist could type the equivalent of a spoken 12 second, 18 word comment in about two
minutes.

If the mean 12 second comment is taken, then an exchange of 30 contributions

might take six minutes in a face-to-face meeting. This can be compared to the time
duration of an exchange of comments in a CT, with perhaps 10 to 30 contributions per
week in similar sized conferences being typical. This may indicate that discussions
requiring a high level of interaction over short periods of time may be less suitable to
CT systems than discussions where the interaction rate required is lower, or can be
spread over a longer period of time.

Silent periods (coded when no participant was speaking) amounted to under 11%
of the contributions and about 10% of the total meeting time.

Function of contribution
The function of each contribution was coded into one of nine categories. Over all
of the 5373 contributions recorded, the dominant function category was information
giving (56%). Querying (14%) and opinion giving (12%) were the only other functions
above 10%. Decision making (6%) was in the bottom group of six other functions,
which together comprised the remaining 18% of the number of contributions.

This is encouraging data for the developers of group communication systems. It

seems likely that current keyboard based technology can readily and effectively support
both information and opinion giving functions, and in addition provide a major
enhancement in being able to go back and "replay" significant contributions of this type
later in the discussion.

Voting mechanisms would appear to be needed only infrequently in electronic

group communication systems.

Distribution of contribution rates

General observation of the data in Table 4 indicates that regardless of group size,
the two most frequent contributors will usually make a combined total of at least 40% of
the number of comments (contributions) made in the meeting, and will occupy an even
higher percentage of total meeting time. In terms of time, the most frequent single
contributor will usually be speaking at least 25% of total meeting time. It is interesting
to note that the most dominant single speakers (48 percent of time) occurred in meeting
sizes at opposite ends of the size scale, namely 4 and 13 participants.

25
Journal of Systems & Information Technology 3(1)

The more reticent half of all meeting participants typically contributed a group
total of under 25% of the speaking time in meetings with less than ten participants, and
under 15% of the time in meetings larger than ten. In essence, the lower half were left to
divide up the scraps of time left over by the dominant speakers.

From these data, it is clear that meeting participation rates are far from uniform
across all participants. In fact, most meetings could be characterized as platforms for
one or two dominant speakers to pass their information and opinion on to the generally
non-contributing majority.

Similar patterns of contribution (dominant participants) have been observed in

electronic group communication (Hiltz 1988).

Listening, reading and speaking rates

The typical speaking rates reported above were 113 words per minute. Listening
comprehension rates up to 275 words per minute are feasible (Foulke & Sticht 1969),
but in the context of a meeting, the listening rate is in lockstep with and can be no faster
than the speaking rate.

The comprehension of written words for an average adult reader, with adequate
comprehension, will vary between 250 and 350 words per minute (Rozin & Gleitman
1977). Therefore, it appears that at least a doubling of "listening" speed for meeting
participants could occur if speech was converted into text and read, and thus uncoupled
from the talking speed of the speaker.

WILL CT SUPPORTED COMMUNICATION IMPROVE OPPORTUNITIES

FOR CONTRIBUTION?

Much has been written about improving meeting processes, to facilitate equitable
contribution and effective use of time. However, the ideal meeting may not be as simple
as one where each of n participants share 1/n of the meeting time. While decision-making
is often seen as the primary justification for meeting in groups, activities such as
informal, non-agenda information sharing may be one of the key beneficial outcomes of
the meeting.

There is a need to distinguish between communication that contributes positively

to the process, and communication that either does not add to the process, or detracts
from it. Festinger (1950) makes the distinction between consummatory (based on
emotion) and instrumental communication, while others have separated contributions
into categories for analysis. Bales (1970) categorised interaction into a number of
categories based on whether the individual seems friendly, dramatizes, agrees, gives a
suggestion, gives an opinion, gives information, asks for information, asks for an
opinion, asks for a suggestion, disagrees, shows tension, or seems unfriendly. Negative
contributions can interrupt the group process and discourage some members from
participating.

26
Journal of Systems & Information Technology 3(1)

The literature suggests some of the claimed benefits of CT supported group

discussions over face-to-face meetings. These include the increased ability of low-status
or reticent individuals to make a contribution to the discussion, and reduction in process
losses due to speaker domination or production blocking. Other important improvements
are the generation of a wider range and number of ideas, the ability to capture a ‘group
memory’, and the support of asynchronous communication.

Much of the research performed on the use of Electronic Meeting Systems (EMS),
Group Decision Support Systems (GDSS), Group Support Systems (GSS) and other
computer-assisted meeting technologies has indicated that the reduced social cues
afforded by anonymity encourage more equal and active participation (Chun and Park
1998). Domination by a single participant is less likely, and with synchronous systems
there is less need to wait for a speaker to finish before another is able to contribute.
Asynchronous systems avoid this difficulty to some extent by allowing participants to
work at their own pace on areas of interest. Status effects on participation and advocacy
in groups were found to be reduced in groups communicating by e-mail rather than face
to face, and more individuals appeared willing to make the first move (Dubrovsky et al.
1995).

A major feature of face to face meetings is the use of verbal, paraverbal and non-
verbal contextual clues, including intonation, facial expressions, glances and hand
gestures (Lebie et al. 1996; McKinlay et al. 1994; Viller 1991) to qualify and enhance
contributions. These clues are generally considered to aid comprehension by other
participants in a discussion but may not actually be interpreted as valid contributions in
and of themselves. While these clues are missing in CT supported communication, there
is little evidence to demonstrate whether this has a marked impact on group
performance. Textual cues in the form of font variations may go some way towards
simulating these enhancements, as may the provision of shared workspaces or video
interaction.

IMPLICATIONS FOR CT DESIGNERS

Primary functions required

Designers of group communication systems must be needs driven, rather than
feature driven. The electronic medium for group communication and the systems to
exploit it have been in existence for over twenty years (Rice 1987). Others (Holland,
1984) have published guides for how to make effective use of face-to-face business
meetings. The present study has highlighted some aspects of face-to-face business
meetings that deserve consideration when the functionality and capabilities of new CT
systems are being designed.
Dominant speakers
The data in this study indicate that the top two participants in any group meeting
will contribute half or more of the contributions in the meeting. CT designers therefore
need to provide the tools to enable these dominant participants to easily contribute to the
discussion in the volume they desire. For any given group, these dominant speakers will

27
Journal of Systems & Information Technology 3(1)

be the heaviest users of the system, especially in the text entry area, and features such as
a choice of editors, and micro to mainframe text movement facilities will be crucial. The
future incorporation of speech recognition, which will eliminate the keyboard bottleneck
disincentive, will likely be very important for this class of users.
Listeners
People who contribute relatively little, in aggregate, to the meeting discussion
make up over half of the meeting participants. If CT systems encourage larger numbers
of peripherally involved people to join discussions, then these additional people are
likely to be only listeners, or low volume contributors. Rather than developing CT
systems assuming equality of participation, designers should provide "power listening"
features for this majority of participants.
Information exchange
Information exchange was the dominant function of the meetings studied. Text-
based electronic communication media can effectively support information exchange
(Keisler et al.., 1984), but CT designers must provide additional capabilities to enhance
this function, such as multimedia (image, speech, graphic) object handling abilities. An
implication of high levels of information exchange is a requirement for subsequent
indexing, searching and archiving of the information elements contributed to the group
discussion. This capability must provide simple but powerful tools to access the vast
amounts of data that are likely to be captured in these systems in the future.
Visual aids
Handouts and whiteboard were used in a small but not insignificant number of the
contributions observed. There are many ideas that are within the reach of today' s
technology that could be used to provide these shared workspaces, such as large format,
high resolution graphic displays or integrated fax interfaces. However, if incorporating
these capabilities, designers must avoid making the basic workstation so complex and
expensive that the basic requirement of information exchange for the majority of usage
is overpowered by the technology for a less frequent usage.
Voting and decision support
Voting was seldom used in the observed meetings. It appeared decisions, when
formally required, were reached instead by compromise and group consensus. While a
large body of research has investigated group decision support systems (Gallupe et al.
1988), the data from the present study appears to indicate that specialized decision
support functions may be of minor importance in the context of the larger requirements
of group communication.
Interactivity and linkages
The majority of contribution linkages observed in the face-to-face meetings were
to the immediately preceding contribution, and the structure of the discussion was
essentially serially, rather than hierarchically, cross-linked. Therefore, complex linking
features, which may slow down the rate of contributions by asking for linkage
information, may be a hindrance rather than an aid to system use.
Attaining asynchronous interactivity
A significant observation from this study was the very high number of relatively
short contributions, and the high percent of contributions that were linked to the

28
Journal of Systems & Information Technology 3(1)

previous contribution. The combination of these two observations leads, by implication,

to the large number of "turn-sharing" contributions that could be exchanged in the
relatively short meeting time. Using the mean contribution times, and allowing for about
90% speaking contributions, and 90% linkage to previous contribution, this would imply
about 240 "linked" and interactive contributions per hour of meeting time. CT system
design has attempted to exploit the asynchronous advantages of the communication
medium, and it has been claimed that typed contributions tend to be longer and with a
higher information content. However, the offsetting disadvantage is that a much lower
number of interactive contributions are likely over a much longer period of time (say
several weeks), greatly reducing the interactivity of CT discussions. Designers need to
address this shortcoming, and provide capability and encouragement for much more
frequent sign-on and contribution. One area which deserves attention is the high time
overhead required to get into the CT system (turn on machine, network connection, host
login, etc.) for what may be a small number of waiting comments. Continuous
connection to network services (perhaps as a background task on a workstation) may
deserve some investigation to solve this serious problem.

User interface design

The user interface is the link between the system capabilities and the user's ability
to make use of them. Potential CT users have characteristics which may span a broad
spectrum, including computer literacy, typing skills, frequency of use, and motivation
(McQueen & Sheffield, 1990). CT user interface designs may target particular
populations of users or task models to achieve their goals. Beside the more general
aspects of user interface design, the observations of this study may be helpful to the CT
user interface designer in the following areas.
Large number of small contributions
Most contributions were short in duration (12 seconds) and low in the number of
words (18). Where typed text is to be input by speakers, the input method (text grabber)
must be simple. For example, word wrap should be an important requirement to relieve
contributors from the worry of end of line spillovers. Another example might be an
automatic spell checker/corrector. On the other hand, features such as block move or
replace may be infrequently used, but if incorporated, may disadvantageously add to the
cognitive load of the user of the text grabber.
Typing versus talking
Digitized voice is becoming more widely used in voice mail systems, and should
be considered for incorporation into CT system design as soon as possible. Reasonable
quality digitized speech requires data rates of a minimum of 1000 bytes per second, so
there are some serious implications for storage of large amounts of data. A more usable
form would be text derived from speech, but economical speech recognition devices,
with large speaker independent vocabularies are not yet commercially available.
However, the future widespread use of CT systems will not likely occur until cheap,
good speech recognition is readily available. Talking rates of 113 words per minute
cannot be compared to hunt and peck typing rates of 10 words per minute. CT designers
need to begin working with speech storage and recognition technologies as soon as
possible.

29
Journal of Systems & Information Technology 3(1)

Non-text objects
Users of CT systems will demand the capability to input, store and retrieve non-
text objects, such as hand written notes, drawings, charts, and other paper based items.
The huge popularity of fax machines supports the need to be able to move easily
between paper and electronic versions of documents. Video images (for example, a
small "picturephone" image of a comment' s contributor) are also going to be desirable.
The user interface designer's task is to provide easy access to these objects without
context switches. Some kind of windowing, multimedia workstation may be a common
future CT system access mechanism.
Portability and access
Business participants in meetings are typically very mobile. Cellular telephones
attest to the current desire of a wide range of business people to be constantly available
for communication. Present CT system design is based on a desktop based workstation
connecting to a server through a network. Future systems must look beyond these
restrictions, to provide a group communication tool that is totally portable.

SUMMARY

This paper has described the results of a field study which videotaped and
analysed ten business meetings. The organization studied frequently uses business
meetings as a group communication method, but the data and conclusions derived may
not be fully generalizable to all business meetings. The data from the analysis has been
used to highlight potential design priorities and user requirements which should be
considered by designers of CT systems. In particular, the important role that information
exchange took in the observed meetings is of interest. As well, the high degree of
interactivity of the meeting participants (short duration, large number of contributions) is
something that will be difficult to replicate in a CT system. Finally, the role that both
dominant speakers, and low participation "listeners" perform in a business meeting is
important to recognize.

REFERENCES

Ajzen, I and Fishbein, M (1980) Understanding Attitudes and Predicting Social

Behavior
Alavi, M. (1993) “An Assessment of Electronic meeting Systems in a Corporate
Setting”. Information and Management; 25: 175-182
Argote, L and McGrath, J E (1993) “Group Processes in Organizations: Continuity and
Change”. International Review of Industrial and Organizational Psychology; 8:
333-389
Bales, Robert F. Interaction process analysis, in Readings on the Sociology of Small
Groups, Mills, Theodore M. & Rosenberg, Stan (eds), Prentice-Hall: New Jersey,
1970.

30
Journal of Systems & Information Technology 3(1)

Barrick, M R., Stewart, G L., Neubert M J and Mount, M K (1998) “Relating Member
Ability and Personality to Work-Team Processes and Team Effectiveness”.
Journal of Applied Psychology; 83(3): 377-391
Bettenhausen, K L (1991) “Five Years of Groups Research: What We Have Learned
and What Needs to be Addressed”. Journal of Management; 17(2): 345-381
Campion, M A., Medsker, G J and Higgs, A C (1993) “Relations Between Work Group
Characteristics and Effectiveness: Implications for Designing Effective Work
Groups”. Personnel Psychology; 46: 823-850
Cappella, J N (1979) “Talk-Silence Sequences in Informal Coversations”. Human
Communication Research; 6(1): 3-17
Carlson, P.J. and Davis, G.B. (1998), An Investigation of Media Selection Among
Directors and Managers: From "Self" to "Other" Orientation, MIS Quarterly,
V.22, No.3, pp. 335-362.
Chun, K J and Park, H K (1998) “Examining the Conflicting Results of GDSS
Research”. Information and Management; 33: 313-325
Clark, H H and Brennan, S E (1991) “Grounding in Communication”. Perspectives on
Socially Shared Cognition
Connolly, T., Jessup, L M and Valacich, J S (1990) “Effects of Anonymity and
Evaluative Tone on Idea Generation in Computer-Mediated Groups”
Management Science; 36(6): 689-703
Daft, R.L. and Lengel, R.H. (1986), Organizational Information Requirements, Media
Richness and Structural Design, Management Science, V.32, No.5, pp. 554-571.
Daft, R.L., Lengel, R.H. and Trevino, L.K. (1987), Message Equivocality, Media
Selection, and Manager Performance: Implications for Information Systems, MIS
Quarterly, V.11, No.3, pp. 355-366.
Diehl, M and Stroebe, W (1987) “Productivity Loss in Brainstorming Groups: Towards
the Solution of a Riddle” Journal of Personality and Social Psychology; 53(3):
497-509
Dubrovsky, V J., Kiesler S and Sethna, B N (1991) “The Equalization Phenomenon:
Sastus Effects in Computer-Mediated and Face-to-Face Decision-Making
Groups”. Human-Computer Interaction; 6: 119-146
Festinger, L (1950) “Informal Social Communication” Psych Review; 57(5): 271-282
Finholt, T and Sproull, L S (1990) “Electronic Groups at Work”. Organization Science;
1(1)
Foulke, E. & Sticht, G.T. Review of research on the intelligibility and compression of
accelerated speech. Psychological Bulletin (72), 1969, pp. 50-62.
Gallupe,R.B, DeSanctis, G. and Dickson, G.W. Computer-based support for group
problem finding: an experimental investigation. MIS Quarterly (12:2), June
1988, pp. 277-296.
George, J F., Easton, G K., Nunamaker, J F., and Northcraft, G B. (1990) “A Study of
Collaborative Group Work With and Without Computer-Based Support”.
Information Systems Research; 1(4): 394-415
Goodall, H L (1990) Small Group Communication in Organizations. Dubuque, IA:
W.M.C Brown Publications
Gudykunst, W.B, Matsumoto, Y., Ting-Toomey, S., Nishida, T., Kim, K., and Heyman,
S. ‘The Influence of Cultural Individualism-Collectivism, Self Construals, and

31
Journal of Systems & Information Technology 3(1)

Individual Values on Communication Styles across Cultures’, Human

Communication Research, Vol. 22, No. 4, pp. 510-543, 1996.
Guzzo, R A and Dickson, M W (!996) “Teams in Organizations: Recent Research on
Performance and Effectiveness”. Annual Review of Psychology; 47: 307-338
Hall, E. T. Beyond Culture, Doubleday, New York, 1976.
Hiltz, Starr Roxanne, Productivity Enhancement from Computer Mediated
Communication: A Systems Contingency Approach. Communications of the
ACM (31:12), 1988, pp. 1438-1454.
Hirokawa, R Y and Gouran, D S (1989) “Facilitation of Group Communication: A
Critique of Prior Research and an Agenda for Future Research”. Management
Communication Quarterly; 3(1): 71-92
Hofstede, G. Culture's Consequences: International Differences in Work-related
Values, Beverly Hills, CA: Sage, 1980.
Holland, Gary. Running a Business Meeting. Dell Publishing Co: New York, 1984.
Jay, A (1976) “How to Run a Meeting”. Harvard Business Review; 54(2): 43-57
Jehn, K A and Shah. P P (1997) “Interpersonal Relationships and Task Performance: An
Examination of Mediating Processes in Friendship and Acquaintance Groups”.
Journal of Personality and Social Psychology; 72(4): 775-790
Kock, N. (1998) “Can Communication Medium Limitations Foster Better Group
Outcomes? An Action Research Study”. Information & Management, 34(5): 295-
305
Lebie, L., Rhoades J A., and McGrath, J E (1996) “Interaction Process in Computer-
Mediated and Face-to-Face Groups”. Computer Supported Cooperative Work; 4:
127-152
Lee, A.S. (1994), Electronic Mail as a Medium for Rich Communication: An Empirical
Investigation Using Hermeneutic Interpretation, MIS Quarterly, V.18, No.2, pp.
143-157.
Lengel, R.H. and Daft, R.L. (1988), The Selection of Communication Media as an
Executive Skill, Academy of Management Executive, V.2, No.3, pp. 225-232.
Littlepage, G., Robison, W and Reddington, K (1997) “Effects of Task Experience and
Group Experience on Group Performance, Member Ability, and Recognition of
Expertise”. Organizational Behavior and Human Decision Processes; 69(2):
133-147
Markus, M.L. (1994), Electronic Mail as the Medium of Managerial Choice,
Organization Science, V.5, No.4, pp. 502-527.
Mayer, Alastair J. User friendliness at 300 baud - the CoSy approach to the human
interface. Computer Conferencing and Electronic Mail Conference Proceedings,
University of Guelph, Guelph Canada, 1985, pp. 23-34.
McCroskey, J C (1977) “Oral Communication Apprehension: A Summary of Recent
Theory and Research”. Human Communication Research; 4: 78-96
McKinlay, A., Procter R., Masting O., Woodburn R., and Arnott, J. (1994) “Studies of
Turn-taking in Computer-Mediated Communication”. Interacting with
Computers; 6(2): 151-171
McQueen Robert J. and Clark, Vanessa N. (1999). Cultural Influences in Contribution
to Electronic Group Communication Discussions . Research Report 99-04,
Department of Management Systems, The University of Waikato, Hamilton, New
Zealand.

32
Journal of Systems & Information Technology 3(1)

McQueen, Robert J. and Sheffield, J., Business meetings and computer conferencing:
requirements for the user interface. Proceedings of the Third Guelph Symposium
on Computer Mediated Communications, University of Guelph, Guelph Canada,
1990, pp. 255-268.
Medsker, L., Tan, M and Turban, E (1995) “Knowledge Acquisition from Multiple
Experts: Problems and Issues”. Expert Systems with Applications; 9(1): 35-40
Nunamaker, J F., Dennis, A R., Valacich, J S and Vogel, D R (1991a) “Information
Technology for Negotiating Groups: Generating Options for Mutual Gain”.
Management Science; 37(10): 1325-1346
Nunamaker, J F., Dennis, A R., Valacich, J S and Vogel, D R and George, J F (1991b)
“Electronic Meeting Systems to Support Group Work”. Communications of the
ACM; 34(7) July: 40-61
Rice, R.E. Computer-Mediated Communication and Organizational Innovation. Journal
of Communication (37:4), Autumn 1987, pp. 65-94.
Rozin, Paul & Gleitman, Lila R. The structure and acquisition of reading II: the reading
process and the acquisition of the alphabetic principle, in Toward a Psychology
of Reading, A.S. Reber & D.L. Scarborough (eds). Erlbaum: Hillsdale, 1977.
Sproull L and Kiesler S (1991) “Increasing Personal Connections” Connections: New
Ways of Working in the Networked Organization. Sproull, L and Kiesler, S. eds;
MIT Press: 79-101
Tuckman, B.W. (1970), Developmental Sequence in Small Groups, Group Processes,
P.B. Smith (Ed), Penguin Books, Middlesex, UK, pp. 322-351.
Viller, S (1991) “The Group Facilitator: A CSCW Perspective”. Proceedings of he
Second European Conference on Computer-Supported Cooperative Work;
September 25-27
Weingart, L R (1997) “How Did They Do That? The Ways and Means of Studying
Group Process”. Research in Organizational Behavior; 19: 189-239
Weisband, S (1994) “Overcoming Social Awareness in Computer-Supported Groups:
Does Anonymity Really Help?”. Computer Supported Cooperative Work; 2: 285-
297
Yellen, R E., Winniford, M., and Sanford, C C. (1995) “Extraversion and Introversion
in Electronically-Supported Meetings”. Information and Management; 28: 63-74

33