***** ARCHEE *************** EJC/REC Vol. 3, No. 2, 1993 ***

USING COMPUTER MEDIATED COMMUNICATION IN AN EDUCATIONAL
CONTEXT:  EDUCATIONAL OUTCOMES AND PEDAGOGICAL LESSONS OF
COMPUTER CONFERENCING


Ray Archee
University of Western Sydney, Nepean


        Abstract.  Research into computer mediated
     communication (CMC) most often has an effects
     focus within a communication, a management, or an
     information context.  CMC has seldom been reported
     being used as an instructional tool.  Using a case
     study approach, this paper reports on a series of
     computer conferencing (CC) projects used at two
     different universities where CC was used in order
     to teach aspects of small group communication on
     the one hand, and writing/rhetoric on the other.
     The main educational outcomes of these projects
     are highlighted, in contrast to the lessons learnt
     by the author.

                    Introduction

     Research into CC has had a short but distinguished
record over the last 14 or so years (for excellent reviews
see Rice, 1980 and Steinfield, 1986).  The reader is
probably aware of at least some of the earlier research
findings in this area.

     Beginning with the seminal book, _The Network Nation_
(1978), Hiltz and Turoff described trials with an early CC
system looking particularly at determinants of use and
appropriate uses; Kerr and Hiltz (1981) followed this up by
providing a complete summary of system features; Rice (1982)
has investigated group processes, and organisational
productivity effects (Rice, 1984); Kiesler, Siegel, and
McGuire (1984) have looked at social psychological effects;
and affect (Kiesler, Zubrow, & Moses, 1985).

     In more recent times the technology of CC has been
developed by operations management professionals for
decision making purposes.  Pinsonneault and Kraemer (1990)
describe two different but related kinds of systems, Group
Decision Support Systems (GDSS) and Group Communication
Support Systems (i.e.  CC), which have different kinds of
effects upon groups and the decision process.

     At this minute the technology has been taken out of the
hands of specialists by the marketing minds of several
multi-national software companies (The Lotus Corporation
being the first).  "Groupware" is now one of the trendiest
kinds of PC application and many companies are jumping on
the bandwagon to establish market share (Dancer, 1992).

     But what of the educational use of CC?  In terms of the
instructional use of CC, there are very few reports.  The
literature on technology, especially CMC, is in some ways
enamoured with the technical, with the effects/advantages of
the technology taking precedence over the real educational
benefits of CMC.

     Smeltzer (1986) in analysing management students'
reactions to a range of electronically mediated
communication (audio, video and computer conferencing) used
stimulation, reinforcement and participation as major
variables.  He found that students were comfortable with the
lack of non-verbal feedback and thus rated "teleteaching"
just as stimulating as face to face teaching.  Lack of
positive reinforcement was not a concern for students, with
greater cohesion amongst students making up for any
shortfalls.  In terms of participation, a minority of
students reported that the electronic format adversely
affected their ability to ask questions and to participate
with the professor.  However all the students believed that
the group cohesion was greater due to the teleteaching.

     Durham (1990) used a CC in order improve humanities
students' abilities at some of the basics of academic
writing: rhetorical stance, the construction of the reader
and the development of an individual's authorial voice.
Durham found that CC was useful in providing "an immediate
and intense writing environment" which developed students'
sense of the reader.  Students negotiated their own
discourse etiquette and because of the freedom that
anonymity provided, found their own authorial voices which
enhanced group cohesion.  A number of non-verbal, textual
icons are also described which added to the students'
general acceptance of the technology.  (Durham was, and
still is, a colleague of mine and her CC design was one
which I partly replicated in my first CC at UTS in 1990.)

     Gatenby (1992) used CC in order to teach Organisational
Communication to management students and found that there
were three main educational benefits.  The first benefit was
the experiential learning which students gained from the use
of a completely novel technology.  A second, was the way in
which CC encouraged students to discuss various aspects of
the course.  The third finding was the way in which CC
allowed meta-communicational analysis of the students' own
classroom interaction.  Gatenby suggests the need for
further research into CC which might focus upon one of the
following: individual participation; personality and
non-verbal cues; choice of language register; perception of
CC via discourse; and use of humour.

     This paper reports on student CMC assignments which
were conducted at two Australian universities, the
University of Technology, Sydney (UTS) and the University of
Western Sydney, Nepean (UWS), over the period 1990 to 1992.
Because of the nature of the projects as educational tools,
not research projects, data collection was somewhat
unsystematic and anecdotal, especially with the first two
CCs I ran.

     Basically, a CC yields a great deal of raw data;
analysing this data is quite problematic, however.
Moreover, the sheer number of uncontrolled variables is a
real concern: prior experience, technological apprehension,
ethnic background, attitudes to writing, user interface,
typing proficiency, and the like, serve to confound student
learning and reactions to CC.

     At UTS in 1990, 75 students in the second-year course,
"Small Group Communication" were required to participate in
an electronic Bulletin Board System (BBS) and a Computer
Conference.  Students were undertaking information science,
business, teacher education, and leisure/tourism degrees.
The CMC assignments were conducted on a Prime minicomputer
and were accessible through a number of student terminal
rooms at the university.

     At UWS in 1991, 144 students of the first-year course,
"Written and Oral Presentation 2" were required to use the
'Nepean CC System'.  These students were undertaking
computer science and industrial maths degrees, and the
assignment was conducted on a Nixdorf unix minicomputer
again accessible through a number of terminal rooms and PC
labs throughout the university's Department of Computer
Science.

     At UWS in 1992, 241 students of the first-year course,
"Reasoned Writing" were required to participate in two
computer conferences.  These students were undertaking
applied communication, liberal arts and Asian studies
degrees.  The assignment was this time conducted on an
Appletalk network with MacPlus microcomputers serving as
terminals in two of the Faculty's computer laboratories .

     In terms of software, at UTS [1990], a computer staff
member customised the Prime database application,
INFORMATION by providing it with a front end whereby users
could choose to either read, write or email individuals.  At
UWS [1991] the application used was a Unix shell script,
simply named 'CC System' which was extant on the Nixdorf
computer system.  Both of these CCs were menu-driven
interfaces in which students chose to read from or write to,
a common area, or email each other as individuals.  At UWS
[1992] the application was the Macintosh electronic mail
programme, QuickMail which the author had customised to run
as a computer conferencing system by allotting groups of
students to the one Username.

                      The Assignments

     There were three kinds of student assignments:

Assignment 1

     To contribute to a bulletin board system (BBS) which
had six "topics" set up.  The topics were as follows:

    a) Technology
    b) UTS Amalgamation Effects
    c) 1980's: Greed is Good; 1990's: ?
    d) Immigration
    e) The Australian Identity
    f) Is Higher Education Merely a Mealticket?

Assignment 2

     To form and interact on a randomly selected closed CC
group in order to make a group consensus decision on a
"desert island" type values clarification exercise which
involved students choosing one fictitious manager (out of
four) to retrench because of the recession.  The managers
were all male, of different ages, of different ethnic
backgrounds, and all had various professional "defects" such
as arrogance, chauvinism, and rumoured alcoholism.  The
ethnic races represented were Australian, Vietnamese,
Polish, and Aboriginal.

Assignment 3

     To form and interact on a randomly-selected closed CC
group in order to make a group consensus decision on the
following question:

    "Should it be law that HIV-positive AIDS carriers must
     disclose their illness whenever they go to the doctor,
     the dentist, going for a licence or even applying for a
     job."

     Assignment 3 was completed by all CC participants (UTS
[1990], UWS [1991] and UWS [1992]); Assignment 1, modelled
upon Durham's (1990) exercise was only used with the UTS
[1990] CC; Assignment 2 was only used with the UWS [1992]
CC.

     The reason for the use of two tasks with the UTS [1990]
and UWS [1992] was that these students rapidly became tired
of discussing the same question week after week.  Whilst
slow participants (the minority) would have probably
preferred one task, the enthusiastic groups rapidly reached
either consensus or total polarisation - either way was
immobilising for these students.  So an extra task was added
to capitalise on the group momentum.

     Usually existing alongside the CC was also the
opportunity for individuals to communicate using electronic
mail.  Thus during the CC period students usually had two
ways of communicating with others - the CC and email.

                         Anonymity

     One interesting feature of the design of the UTS [1990]
system was the decision to make everyone anonymous (as
Durham, 1990 did).  The first time a user logged on to the
system they were asked to nominate a pseudonym by which they
would be known for the entire period.  Some of the
interesting pseudonyms included Avocado, Bignose, Sexy Phil
and Michaelangelo.  (One student took it upon himself to use
my real name as his pseudonym!  He was most apologetic when
I queried him on this.)  The pseudonyms were useful in that
age, sex and status were completely masked from each other
from the outset.  Whilst the participants were aware that I
had the master list by which I could identify all
pseudonyms, much interest was created by this
'guess-the-sender' game.

     In the UWS [1991] CC users were pretty well obvious by
having to use their group number and the first five letters
of their surname in order to login - 52Willi was a student
with the name of Williams who was in Group Number 52.

     In the UWS [1992] CC because of the nature of the
original Electronic Mail software, all the users in the one
group had the same Username, so it was up to individuals to
identify themselves in the Subject line of their "article".
Some students realised very quickly that they did not have
to include their name making it impossible to work out the
sender of a particular piece of mail.  They also worked out
that the software had a Public Notice Board function which
they anonymously used to post "junk mail" about various
parties, and to write swear words.  This realisation led to
some onerous CC "graffiti" which at its worst was highly
insulting to a particular individual.

     This fact is entirely in keeping with the findings of
several authors who talk about effects of anonymity with CMC
(e.g.  Kiesler et al, 1984; Hiltz, Turoff, & Johnson, 1985;
Jessup, Connolly, & Galegher, 1990).  The only general
consensus is that anonymity leads to more messages being
sent.  The negative effect of "flaming" or using language
which is much stronger than usual, is a common finding in
the CMC research literature.  However the use of student
participants in many such studies must surely lead one to
question the generalizability of such findings.  Students
constitute a section of the population who are unique in
many respects.  My observation is that students, especially
males, tend to swear a lot, and thus doing the same thing on
a CC is not unusual if the participant knows that he/she can
get away with it.  It is also customary teaching practice,
at least with young children, to let students play with the
technology whether it be a musical instrument or a computer.
This play aspect of learning a new activity might simply
translate at the student level into the use of previously
taboo words and activities in order to "test the water."

                   Educational Outcomes

Novelty and Excitement

     Educational outcomes varied with each of the different
courses being taught.  But what was common to all of the CCs
was the expression by students that the CC was an exciting
and memorable activity for them.  Course evaluations
confirmed this anecdotal evidence.  Gatenby (1992) also
found that the "opportunity to try out a new technology" was
paramount to students' evaluations of the CC.  The
text-based format of a CC is a hybrid of speech and writing,
displaying features of both (and some of its own).  The
opportunity for students to experience this new asynchronous
medium of communication is challenging for brand-new users,
but becomes rewarding when they master the technology and
succeed in "publishing" their statements and opinions.  The
novelty feature alone makes CC a worthy inclusion in any
course which can support it.

Activities on the Bulletin Board System

     The number of "articles" posted to the six bulletin
boards at UTS [1990] over a period of four weeks were as
follows:

a) Technology                                 100
b) UTS Amalgamation Effects                    55
c) 1980's: Greed is Good; 1990's: ?            88
d) Immigration                                120
e) The Australian Identity                    159
f) Is Higher Education Merely a Mealticket?    66
                                              ___
                                       TOTAL  588

     The BBS proved popular as the statistics show.  So
popular in fact, that students by the second week had
trouble reading all the new messages posted to the BBS.  One
problem with the software was that articles were not date
stamped nor presented according to date.  Thus students had
no way of determining which were the latest entries to a
bulletin board except by memory.  As the number of articles
increased students had to choose from several screenfuls of
articles which eventually led bulletin board burnout with
students perhaps reading only a handful of articles at one
sitting.

     Students were at a loss at the beginning of the CC as
to how to contribute to the system.  Some mature students
decided that they would go to the library and prepare their
"articles" and then load them on after they had a finished
product.  Others of course, simply typed aloud.  What was
missing was a model of what a CC article should look like.
Failing such a specimen the students were at a loss as to
the length, the register, and the format.  What eventually
occurred was that students who began writing a formal style
of article began to "loosen up" the longer the CC went for.

     After an initial burst of enthusiasm, students began to
realise that they could not read every single article which
was posted to the BBS.  Burnout set in, with students
reading only anything which took their fancy or nothing at
all, and usually on the first screenful of messages.  After
two weeks, students were complaining that the BBS was "not
going anywhere", all people were doing was asserting some
opinions and moving on to another topic.  There was no group
cohesion since there were no consistent group members.

     In contrast to Durham (1990), individuals did not stick
to the one topic but selected from several on which to
write.  Durham (1990) found that most students kept to one
or two topics and because of this formed de facto problem
solving groups with regular membership.  This led to much
greater group cohesion and the formation of a discourse
etiquette.

     At UTS [1990] apart from turn-taking, a small degree of
persuasion, and the novelty of being able to access other
participants' ideas the general feedback after two weeks of
use was "Do we have to keep going with this?  Can't we have
a real task to perform?"

     One problem with the assignments was the fact that with
all of these topics, the author took it upon himself to kick
the discussion off with an anonymous, fairly controversial
statement.  In hindsight this may not have been a good idea.
Weisband (1989) suggests that first advocates in a
discussion are more likely to gain acceptance of their
ideas.  In this case the first message could have set the
agenda for the rest of the discussion.  Kickstarting the
topic may have in fact pushed the discussion into areas
which would not have been explored.  Two examples of this
were 1) making the "Australian Identity" related to sexual
differences, and 2) providing somewhat racist comments to
the "Immigration" topic in order to engender discussion.
These topics were the most popular and kept their original
"themes" throughout the duration of the assignment.

Decision Making

     Decision making using a CC would seem to be the natural
task of a CC group, and therefore it was expected that
students would benefit from a CC in terms of experience in
making better group decisions.  Researchers are however,
divided as to whether the lack of social status and
non-verbal communication actually helps or hinders decision
making and decision quality.  Rice (1984) points to two
"schools" of CMC decision making: the 'cool' school which
emphasises low conflict situations, where lack of status,
charisma, prejudice and reasoned, unimpassioned arguments
lead to better decisions; and the 'warm' school which
asserts that it is exactly conflict, human relations and
non-verbal communication which lead to quality decisions.
It was my experience with CC decision making, that any type
of consensus decision was quite difficult to achieve -
giving support for the 'warm' school hypothesis.

     After my brief encounter with open CC groups, with all
the subsequent CCs, students were assigned to random, closed
CC decision making groups.  The UTS [1990] CC provides a
near blueprint for the rest of the later CC decision making
groups.

     At UTS students were assigned to five random groups,
and three in-class groups of 9 or 10 members.  Because of
administrative problems the three in-class groups (from the
same workshops) only used the conferencing system once and
so can not be taken into account.  There were thus five CC
groups' decision transcripts which could be perused for the
presence or absence of consensus.

     The ability of a CC to facilitate consensus is illusory
in light of the following results.

     GROUP 1: found the mechanics of making a consensus
decision on an asynchronous computer conference difficult to
achieve.  One member of the group decided that it would be a
good idea to have members vote individually.  So she sent
several group members the same mail message.  This did not
entirely work since members kept bringing up new issues.
Then a few members tried to summarise conclusions for the
entire group but were given no feedback.  The main
impression gained from reading the transcripts is the
frustration which people felt when their summarising,
concluding gestures were not responded to.  No consensus.

     GROUP 2: looks as if it managed to come to the
conclusion that HIV carriers should be made to carry some
form identification.  This conclusion was based on the last
few days of the computer conference.  However looking
further back in time, one finds that opponents of an HIV law
never agreed to the final say by the last three conferees.
This could be a similar situation to face to face consensus
decisions where at the end of the day the decision is made
by the members still standing.  No consensus.

     GROUP 3: looks as if it reached the conclusion that it
should be law that HIV carriers must disclose their
condition to high risk members of the general public.  Again
a few members of the group affirmed this decision whilst
previous opponents make no comment.  No consensus.

     GROUP 4: had five members who formed a caucus by
affirming each other's idea, that being that the medical
profession had the right to know if someone was HIV
positive.  Unfortunately, whilst these five members were
slapping each other on the back in coming to their
consensus, three members of the group were totally isolated
and ignored.  These members' views were entirely contrary to
the majority - but no mention is made of these in the
summing up by the stronger majority.  It was as if the
majority caucus decided that it would only address its own
constituency and totally ignore outside influences.  The
minority group on the other hand protested loudly in the
beginning of the conference but were silent by the end.  No
consensus.

     GROUP 5: similarly appeared to reach the decision that
medical testing of four groups of people was necessary - all
hospital patients, suspected rapists, all new immigrants,
and all persons acting in a threatening manner with
supposedly infected needles.  This time the decision was
made by a member who summarised the group discussion, asked
"who else agrees?" and did not get any replies.  No
consensus.

     Given this problem with CC decision making it was
hypothesised that perhaps the group size was too large for
members to reach agreement.  Group size has for a long time
been viewed as a significant variable in small group
research (Hare, 1976).  Thus at UWS [1991], group size was
experimentally varied in order to compare the ease of
decision making.  Groups of 3, 5, 7, 11, 15 and 21 were
created but to no real effect.

     Decision making on the AIDS topic followed the same
basic pattern no matter what the size of the group - a few
enthusiastic members contributing most of the entries
usually with one member becoming a task leader, a few
members abstaining from contributing until the last few days
of the CC and because of this considerable frustration,
summarising statements by one or more members, but no stated
agreement by all group members, just as before.  There were
three groups whose leaders realised that voting needed to be
instituted in order for consensus to be reached.  Of these
groups (a 5- member, a 7-member, and an 11-member team) two
successfully reached the required consensus decision via
voting.  The most successful group was the 5-member team
whose leader emailed a table with every member's name on it
which asked all members to make a decision by checking a box
and returning the table to the group.  The unsuccessful
voting group, as before, totally ignored the request and
kept arguing.

     At UWS [1992] again group size was experimentally
varied to facilitate decision making and also to facilitate
administration, students needed to be allotted to groups
which would allow the 32 Usernames to be utilised fully.
Groups of 5, 7, 9, and 11 were formed.

     Assignment 2 was designed to suit a supposed
'convergent' style of group decision making.  By limiting
the number of alternatives, (one out of four) it was thought
that conferees could more easily make a decision which all
could agree on.  Nothing could have been further from the
truth - groups were again polarised, usually around one of
two managers, one because he was older, and the other
because he was arrogant and seen as a chauvinist.  Again,
voting was employed by sensible task leaders and again, not
everyone returned a vote, making the ballot invalid.  On the
student evaluations (N=151) 49.7% of conferees thought that
their group had reached consensus.  Inspection of group
transcripts reveal that groups were in fact either
deadlocked or that a few members did not contribute in order
to confirm such unanimity.

Collaborative Writing

     Research in the collaborative writing area is well
developed but mainly confined to traditional writing and
composition courses.  Collaborative writing using electronic
networks is beginning to be researched, but seldom in
relation to CC.  New research includes Thompson (1990) who
argued that electronic bulletin boards are good places for
collaborative writing to take place; and Hartman et al.
(1991) who found electronic mail networks to have very
positive effects on teacher-student and student-student
interactions.

     But a CC is also a perfect example of a collaborative
writing environment.  If the transcript of the conference is
viewed as the written text then all conferees contribute
collaboratively to the major task, i.e. the production of
the text.  It is however the assignment of special tasks, a
feature of much workplace and classroom collaboration, that
is usually missing in CC writing.  Most computer conferees
contribute to the central goal in a more or less equitable
way.  The strategy of using teams or committees of group
members to focus on task sub-goals is not viable since CC is
predominantly used asynchronously obviating the advantages
small committees which usually meet together at the same
time.

     Although the use of sub-groups is inhibited by CC,
interaction between all conferees tends to increase
markedly.  After each new 'draft' at a certain topic, all
group members can easily read, edit if necessary, and
respond to any member's contribution.  The ease of this task
on a personal computer or computer terminal far outweighs
any of the more traditional methods which include
photocopying, distribution, red pencil and retyping.  Indeed
the speed by which conferees can respond to each other's
work is problematic insofar as it creates a very large
amount of new material which all members are obliged to read
and assimilate.  Once the transcript gets longer than about
25 messages or one screenful, conferees reported that they
no longer read every new message, preferring instead to
select work from particular fellow writers.  Another
reported benefit of computer networks is the that less able
writers can communicate more via electronic means than via
traditional pencil and paper (Hartman et al., 1991).

     However there are also costs to communicating
electronically - conflicts of ideas would seem to be much
more common in CC groups than in more traditional
collaborative writing groups.  The immediacy of being able
read and comment upon others' work bestows unusually high
critical power upon both acerbic and easygoing conferees.
Hence agreeing on a decision can become a fruitless exercise
in many CC groups.  And because of the asynchronous nature
of CC, there is little pressure, if any, for conferees to
conform to group norms as is the case in face to face
meetings.  Thus individual conferees may be very reluctant
to change their point of view even after discovering that
their ideas conflict with the majority of other members.
There is a always the danger of CC writers never
collaborating unless some control is exercised by a
supervisor.

Rhetoric

     Most students who participated in the CCs had not ever
experienced such an activity prior to university.  The ones
who did have such experience (a handful) were those
mature-age students who had previously worked in
organisations which employed such technologies.  Therefore
the exact nature of what constituted a CC "article" was
unknown to most conferees, and became a matter of trial and
error whilst using the system.

     Since CC messages do not constitute a genre, and most
students had never participated in decision-making groups, a
number of questions needed to be negotiated by students in
order for them to fully participate.  Such rhetorical issues
as the length of a message (long or short), the style of a
message (formal or informal), the purpose of a message
(persuasion or contribution), who their audience should be
(another member or the whole group), the kind of support for
assertions (opinion or referenced), how to reach consensus
(voting or negotiation) all needed to be answered by
conferees in order to write a CC "article".

     The author was asked on many occasions what constituted
a message, and the reply given was "any piece of relevant
information, from a single sentence to several paragraphs."
It was intentionally left to the conferees themselves to
determine the answers to the above issues since it was
viewed that students must be allowed to discover for
themselves their natural writing processes and personae.

The Creation of Discourse and Community

     Above all other considerations, CC permits the creation
of discourse and as by-product, creates a community of
fellow participants.  Whilst the nomenclature and design of
CMC systems tend to mimic real life mail, conferences, and
bulletin boards, there is no real life equivalent of CC.  It
is a unique medium which encourages interactive and reactive
writing which could not otherwise be accomplished.  This
uniqueness also creates a sense of camaraderie which when
taken together with the computer interface and the rules
which govern the whole electronic communication process,
forms a sense of identity, purpose and community.

     Feedback became an important ingredient to whether or
not a conferee wrote interactively or just expressed their
own ideas.  Those members who did not receive responses more
than likely expressed irritation at the group's tardiness.
Ideally members would take their turn by writing an article
which was read by the next member who would either agree or
disagree with that person's ideas.  This article in turn
would then be read by a third person and so on.  Whilst this
did occur to some extent, many conferees simply ignored the
transcript, especially when it got rather long, and simply
wrote their own ideas regardless of the other members in the
group.  This highly individualistic ethos mitigated against
group cohesion and prevented the group from achieving
anything like group consensus.  Student evaluations of the
UWS [1992] course also bear this observation out (only 14.7%
agreed that the CC had promoted group cohesion).

     Conferees used a variety of discourse strategies in
order to make their points.  Members constructed arguments
which they used to express their own point of view.  Issues
of fact, of value, especially moral value and of policy were
the most common kinds of discourse content, apart from
socio-emotional statements.  Members would typically state
their position on these highly emotive subjects, and then
try to back up their position by using concrete examples of
how their conclusion would work in the real world.  Not all
conferees however provided the same quantity of evidence.

     Depending upon the group and the previous articles,
some members tried to convince others of the error of their
ways by providing counter arguments to an already-stated
position.  This met with varying degrees of success since
most participants were already advocates of one decision,
i.e. most people took a stand on these emotive issues, very
few people decided to "sit in the fence".  Whether this is
an artifact of the actual question or the medium is
uncertain.  Suffice to say that the technology by its
"cyberspatiality" and its alienness encourages people to be
much more direct in their writing and arguing - the online
nature also encouraging plain English, and rather short
articles.

Electronic Mail at UTS [1990]

     The electronic mail facility was not used by many
students to actually reach any decisions in the computer
conference.  Early on in the semester the sheer number of
email messages rivalled the number of articles posted to the
BBS.  In total there were 411 messages left on the system,
of which the majority were no longer than six lines long,
and predominantly social in origin.  Very few people
realised the potential to connect with like-minded group
members using the email.  Only one person used email in
order to get feedback on her own ideas about HIV carriers.

     Email was used in a number of ways on the system.
Inspection of the message transcripts reveal several genres
of message, including getting to know you pleasantries, try
to guess my real identity messages, feedback from the BBS,
simple social exchanges of news and events, and somewhat
risque, Minitel lookalike 'message roses'.  Approximately
10% of the email messages had some form of sexual content.
This varied from innuendo and flirtation to statements like
"This is Boob here and I love yours".  Indeed when I
announced to students (Teacher Education) that the computer
conference was about to end several students enquired
whether I was also removing their email access.  When I
asked why they were so anxious, one student replied that
there several electronic affairs going on over the system,
and that these affairs were not quite over.

                    Pedagogical Lessons

User Friendliness is Useful

     The behaviour of students who completed assignment 2
was quite different to the other CC's for the main reason
that these students used Macintosh microcomputers as
terminals for typing.  The Macintosh has a reputation for
user friendliness and the CC was no exception.  The author
had previously spent several hours customising the QuickMail
electronic mail programme so that it could use its 35
Usernames with the 244 participants.  Training students in
the use of the Macintosh took two hours by which time most
students could use the floppy disk system and could send a
piece of mail to their own group.  Unlike previous CCs where
students were constantly needing assistance the Macintosh
hardware and software was a joy to use.

     Typically a non-Macintosh based system has two parts,
the CC part which managed the transcript allowing users to
choose to read or contribute entries, and an editor, whereby
users could write and edit their articles.  It is the editor
that is typically the problem for most new students.  A
graphically oriented environment such as the Macintosh,
facilitates writing more so than any other one factor.
Students, many of whom had no knowledge of computers, were
up and running within two hours of starting.  The Mac-CC
students reported that the computer platform had been very
easy to use and thus should be a prime consideration to
anyone contemplating running a CC for the first time.

Access to Technical Help is Important

     Unless you are a programmer, most academics are going
to need to be on more than just speaking terms with
technical staff who can assist students with problems.
There is no such thing as a trouble-free CC program
(although the QuickMail CC came close), and therefore
glitches will occur which will be extremely irritating to
students.  At the beginning, I was on very good terms with
the programmer who wrote the front end to the UTS [1990] CC.
I subsequently had to ask for his help at least twice a week
for the next four weeks.  This assistance was given
begrudgingly by the end of the CC, since he had expected the
CC to just run itself.  Problems constantly arose which
demanded his time and attention.  One should expect the
unexpected with students and CC technology.  And so should
your technical assistants.

All Students are Not the Same

     Obviously all students are not the same, and the CC
articles show significant differences with respect to other
variables such as discipline of study and ethnic background.
In summary, the more Anglo-Saxon and the more Humanities
oriented, the more a student wrote and argued.  Students
who came from other ethnic backgrounds, as a whole, wrote
less than their Anglo-Saxon counterparts.  Computer science
students and Industrial Maths students wrote less than their
Humanities/Liberal Arts counterparts, by a factor of three.
Even though all students were required to participate in the
CCs, it was the mostly the Computer Science/Industrial Maths
students who did not bother until right at the end of the
CC.

Face to Face Groups are Better at Decision Making

     One of the main advantages of real-life groups over
computer conferences is that face to face decisions can be
made so that every member of the group has access to
immediate verbal and non-verbal feedback.  A consensus
decision would thus be a conclusion reached by all of the
group members at around about the same time.  Lack of
conflict and new issues would confirm this consensus.

     One of the most obvious difficulties with the computer
conferences was the fact that group members could not easily
obtain confirmation of their proposals from all other
members. i.e. the medium mitigated against immediate
feedback, and so group summarisers/leaders were lost as to
how well their ideas were received.  Thus when such leaders
finished their summaries they would often end with comments
like 'Does anybody agree with me?' or 'Who agrees?' and then
attempt to obtain support from all other members in writing,
unfortunately to no avail, the majority of times.

     The so-called decision making phase of a group is
weakened, or may be entirely missing when spread out over
several days or weeks, and so individuals may not realise
that the group is actually in the process of coming to a
conclusion and thus carry on with their own highly personal
monologues.  Disdain of members who contribute infrequently,
or who continue to "blow their own trumpets" could explain
why a minority voice might be totally ignored whilst a
caucus may reach their own decisions.

     Lack of any non-verbal communication, and feedback
meant that members are probably uncertain of the group
reaction to their various points of view.  Thus whether to
press more strongly or lighten up on their arguments is
usually a guessing game.  Lack of non-verbal communication
also mitigates against the recognition by conferees that the
group might have entered the next stage of development such
as a performing stage.

Does Consensus Really Exist?

     The concept we all know as "consensus" has been
disputed in recent times by researchers who ask whether in
fact, group meetings reach a negotiated agreement or a
voting agreement instead (Philips & Wood, 1984).  It is my
experience that face to face group meetings usually reach
decisions by either a vote, or by abstinence when the
decision is being tabled.  Whether this is really consensus
is a matter of definition and conjecture.  Thus it is
entirely possible that the CC groups in fact mimicked face
to face groups, and that neither kinds of groups exhibit
consensus.  If this is the case then CC should not be seen
as a poor runner-up to face-to-face decision making, but
rather a distinctive alternative with several benefits.

     CC can be seen to have several advantages over
traditional media.  It allows members total freedom when it
comes to personal expression, free from the inhibitions of
shyness, dominating personalities, or simply the inability
to think and talk confidently in front of a group of people.
Writing per se gives one time to think, reflect and most
importantly, to revise one's proposals.  And the ability to
choose which computer conference articles to read, and in
which order seemingly endows conferees with power, thus
promoting self-esteem and making individuals feel like
worthwhile participants of a group.  This is a marked
contrast to the face to face situation where members can not
choose who will speak and in which order, and where usually
one or more members opt out of the discussion.

                        Conclusions

     This report is limited by a lack of detailed analyses
of the three CC transripts.  As I have already said, CCs
produce a great deal raw data (over 2000 entries with the
Mac CC), which the researcher is at a loss with how to
handle.  Ordinary database applications only allow storage
and text string retrieval.  One useful technique has been to
use a qualitative research programme such as NUDIST (see
Richards & Richards, 1991) in order to index and code every
CC entry.  Results using this kind of analysis are pending.

     The whole experience with all three CCs was both
exciting, and frustrating.  Exciting because they were
innovative and very well received by the majority of
students and the faculty staff involved.  When students use
their own free time in order to willingly work in front of a
monitor then something is working right.

     Frustrating because whenever problems occurred with the
non-Macintosh CCs, my only recourse was to ask the
designer/programmer of the system to repair the problem, and
trouble could arise every day.

     The system designers who enthusiastically began
projects ended up disgruntled and incredulous, begrudgingly
mending glitches in the program.  My advice to anyone
embarking on implementing computer-mediated communication is
to learn as much as possible about the programme you will be
using, be prepared for problems and make sure you have very
sympathetic support for your project.

     In terms of future research, it has come to my
attention that there is an alternative to CC on the Internet
itself, and this is the synchronous "chat" programme called
the Internet Relay Chat or IRC.  Although used for mainly
social purposes by a large student population at present,
IRC would make a very useful comparison to CC so that the
real-time effects can be differentiated from the
delayed-time ones of CC.  Is it the lack of immediate
feedback that inhibits conferees from reaching consensus?
Does synchronous CMC have any advantages over asynchronous
CMC?  Here is one direction which CMC researchers have yet
to travel.


                        References

Dancer, H. (1992).  Groupware: Defining the need to know.
     Australian Personal Computing, 13(10), 132-152.

Durham, M. (1990).  Computer conferencing, students'
     rhetorical stance and the demands of academic
     discourse. Journal of Computer Assisted Learning,
     6, 265-272.

Gatenby, B. (1992).  Computer conferencing in an
     organisational communication class.  In B. Ticeheurst
     (Ed.) Communication Australia: A Search for Meaning in
     Changing Times.  Griffin Publications, London, and The
     Australian Communication Association, CRIA, Canberra.

Hare, A.P.  (1976).  The handbook of small group
     communication.  London: The Free Press.

Hartman, K., & others (1991).  Patterns of social
     interaction and learning to write: Some effects of
     network technologies. Written Communication, 8,
     79-113.

Hiltz S.R., & Turoff, M. (1978).  The network nation: Human
     communication via computer.  Reading, MA: Addison-
     Wesley.

Hiltz S.R., Turoff, M., & Johnson, K. (1985).
     Disinhibition, deindividuation, and group process in
     computerized conferences: A field experiment comparing
     pen name and real name conferences.  Paper presented at
     International Communication Association 34th Annual
     Conference, San Francisco, CA.

Jessup, L.M., Connolly, T., & Galegher, J. (1990).  The
     effects of anonymity on GDSS group process with an
     idea-generating task.  MIS Quarterly, 14(3), 313-321.

Kerr, E.B., & Hiltz, S.R.  (1981).  Cognitive impacts of
     Computer-mediated communication systems upon
     individuals.  In L. F. Lunin, M. Henderson, Madeline,
     and H. Wooster (Eds.) The Information Community: An
     Alliance for Progress: Proceedings of the American
     Society for Information Science 44th Annual Meeting
     (pp. 232-234).  White Plains, NY: Knowledge Industry
     Publications.

Kiesler, S., Siegel, J., & McGuire, T.W.  (1984).  Social
     psychological aspects of computer-mediated
     communication.  American Psychologist, 39, 1123-1134.

Kiesler, S., Zubrow, D., & Moses, A.M.  (1985).  Affect in
     computer-mediated communication: an experiment in
     synchronous terminal-to-terminal discussion.  Human
     Computer Interaction, 1, 77-104.

Pinsonneault, A., & Kraemer, K.L.  (1990).  The effects of
     electronic meetings on group processes and outcomes: An
     assessment of the empirical research.  European Journal
     of Operational Research, 46, 143-161.

Philips, G.M., & Wood, J.T.  (1984).  (Eds.)  Emergent
     issues in human decision making.  Carbondale, IL:
     Southern Illinois Press.

Rice, R.E.  (1980).  The impacts of computer-mediated
     organisational and interpersonal communication.  In
     Martha E. Williams (Ed.) Annual Review of Information
     Science and Technology Vol 15, (pp. 221-249).  White
     Plains, NY:  Knowledge Industry Publications.

Rice, R. (1982).  Communication networking in computer
     conferencing systems: A longitudinal study of group
     roles and system structure.  In M. Burgoon (Ed.)
     Communication Yearbook 6 (pp. 925-944).  Beverly
     Hills, CA: Sage Publications.

Rice, R. (1984).  Mediated communication.  In R. Rice and
     Associates (Eds.) The New Media (pp. 133-136).
     Beverly Hills, CA: Sage Publications.

Richards, T., & Richards, L. (1991).  Computing in
     qualitative analysis, a healthy development?
     Qualitative Health Research, 1(2), (May), 234-262.

Smeltzer, L. (1986).  An analysis of receivers' reactions to
     electronically mediated communication.  Journal of
     Business Communication, 23(4), 37-54.

Steinfield, C.W.  (1986).  Computer-mediated communication
     systems.  In Martha E. Williams (Ed.) Annual Review of
     Information Science and Technology Vol 21, (pp. 167-
     202).  White Plains, NY:  Knowledge Industry
     Publications.

Thompson, D. (1990).  Electronic bulletin boards: A timeless
     place for collaborative writing projects. Computers
     and Composition, 7(3), 43-53.

Weisband, S.P.  (1989).  Discussion, advocacy and computer-
     mediated communication effects in group decision
     making.  Unpublished Doctoral Thesis, Carnegie Mellon
     University, Pittsburgh, Pennsylvaia.
------------------------------------------------------------
                      Copyright 1993
   Communication Institute for Online Scholarship, Inc.