A MULTIDIMENSIONAL MODEL TO ANALYZE SOCIAL AND

TECHNICAL FACTORS IN COMPUTER-MEDIATED

COMMUNICATION

Monique Janneck

Department of Psychology, University of Hamburg, Von-Melle-Park 11, 20146 Hamburg, Germany

Keywords: Computer-mediated communication, Computer-supported cooperative learning, Computer-supported

Cooperative work.

Abstract: This paper proposes a multidimensional model to analyze problems in computer-mediated communication

(CMC), which can serve as a framework to integrate existing CMC approaches and also offers guidelines

for the selection, the design, and the social and organizational integration of CMC tools. The specific

strength of the model is its clear distinction between social and technical factors influencing computer-

mediated communication. A case study of groupware use is presented to demonstrate the usefulness of the

model to analyze difficulties in CMC settings and decide whether to address a certain problem on a design

level or a personal, social, or organizational level.

1 INTRODUCTION AND

BACKGROUND

Numerous theories of computer-mediated

communication (CMC) provide explanations of

why communication processes succeed or fail.

However, there is a lack of comprehensive models

integrating different theoretical approaches.

Furthermore, CMC theories predominantly focus on

analyzing computer-mediated communication from

a (mainly) psychological perspective, and less on the

design of CMC tools from a software engineering

perspective. This paper proposes a multidimensional

model of computer-mediated communication, which

can serve as a framework to integrate existing CMC

approaches. Furthermore, implications for the

selection, the design, and the integration of CMC

tools into existing social and organizational

structures can be drawn from the model.

The following paragraphs very briefly introduce

some influential CMC theories.

For example, so-called “cues filtered out”

approaches emphasize that computer-mediated

communication is often text-based, and therefore

nonverbal communication signs as well as social

cues such as age, sex, ethnics, socioeconomic status,

appearance etc. are lost or at least difficult to

convey. While some authors (especially in early

CMC research) take a deficit-oriented view of CMC

as “impoverished” form of communication (e.g.

Herrmann 1993), others emphasize that filtering out

social cues might result in a more equal and honest

interaction which is less inhibited e.g. by prejudices

or social status (e.g. Sproull & Kiesler 1991, Hian et

al. 2004, McKenna & Bargh 2000, Whitty & Gavin

2001). However, disinhibition and anonymity might

also result in aggression, hostility and harassment

(e.g. Burnett & Burkle 2004).

Media choice theories emphasize that different

media are appropriate for different communication

tasks and needs. Therefore, successful

communication depends on an adequate media

choice. For example, the Media Richness Theory

(Daft & Lengel 1986) classifies different media

according to their capability of conveying complex

information: Richer media such as video

conferencing or telephone should be chosen for

highly complex or ambiguous communication tasks,

while less rich media such as e-mail are more

appropriate and efficient for simpler tasks.

Both cues-filtered-out approaches and media

choice theories indicate that computer-mediated

communication might be more anonymous and less

personal than face-to-face interaction. Empirical

studies often show that CMC is more task-oriented.

CMC groups perform worse on tasks involving high

levels of socio-emotional interaction, especially

521

Janneck M.

A MULTIDIMENSIONAL MODEL TO ANALYZE SOCIAL AND TECHNICAL FACTORS IN COMPUTER-MEDIATED COMMUNICATION.

DOI: 10.5220/0001834905210526

In Proceedings of the Fifth International Conference on Web Information Systems and Technologies (WEBIST 2009), page

ISBN: 978-989-8111-81-4

under time pressure (e.g. Bordia 1997, Hian et al.

2004, Burnett 2000, Burnett & Burkle 2004, Birnie

& Horvath 2002).

The different approaches highlight different

aspects of CMC and possible causes for breakdowns

and problems. However, when problems arise in

computer-mediated communication situations, the

specific causes are often difficult to figure out.

Especially, it is often hard to distinguish between

social and technical problems. As a result, people

might try to solve social problems technically, and

vice versa–a phenomenon that can also be observed

in research on computer-supported communication

and cooperation.

The multidimensional model presented here

helps to analyze difficulties in computer-mediated

communication by making a clear distinction

between characteristics and behavior of the human

communication partners on the one hand, and the

media features and “behavior” on the other hand.

The model is presented in the next section, followed

by a case study illustrating its usefulness for analysis

and design.

2 A MULTIDIMENSIONAL

MODEL OF BREAKDOWNS IN

COMPUTER-MEDIATED

COMMUNICATION

The multidimensional model defines five factors

influencing communication (see figure 1).

Figure 1: A multi-dimensional model of factors

influencing computer-mediated communication.

A. Human Communication Problems:

Naturally, just like in face-to-face interaction,

problems and misunderstandings might occur in

computer-mediated communication, which have

social or personal reasons and have nothing to do

with the communication media used. They are of

little interest for developers of CMC tools because

they cannot be influenced or reduced by media

design. Nevertheless, the literature describes

repeated attempts to solve or avoid social problems

through technology, for example, when trying to

compensate a lack of trust among interaction

partners by sophisticated access controls (e.g. Sikkel

1997, Sohlenkamp et al. 2000). Notwithstanding,

understanding and distinguishing human

communication problems from truly CMC-related

problems is crucial to identify and address actual

design challenges and also to devise accompanying

measures to support and moderate media use.

B. Individual Characteristics:

The individual characteristics and experiences of the

people involved in the interaction will obviously

influence communication in a decisive way. As this

is also true for face-to-face interaction, CMC is

additionally influenced by the communicators’

technical skills (e.g. using browsers or e-mail),

media competencies (e.g. making appropriate media

choices), or mental models, i.e. users’ conceptions

and beliefs regarding configuration and functioning

of a software system which influence software use.

C. Technical Errors and Failures:

Data loss, e.g. due to software bugs or network

problems, is a source of problems which can be

compared to transmission errors as described by the

‘classical’ mathematical theory of communication

by Shannon & Weaver (1949). Technical errors may

result in a partial or complete message loss. Data

loss is especially problematic if it goes unnoticed by

sender and recipient of the information, e.g. if an e-

mail is lost without an error prompt. Web

applications might provide insufficient feedback

when information is displayed differently to several

recipients, e.g. because of individual customizations

or access rights: As a result, the sender might not

know if and how his message was displayed.

Technical problems also include system

incompatibilities (e.g. different operating systems,

data formats etc.). Technical failures are often hard

to understand or avoid, even for experienced users.

D. Media Characteristics:

As described by media choice theories, difficulties

may arise if CMC tools are not suitable for the

underlying communication needs: Media

characteristics influence communication success.

For example, scheduling an appointment with

several participants will likely be much more

WEBIST 2009 - 5th International Conference on Web Information Systems and Technologies

522

complicated and time-consuming when

asynchronous technology such as e-mail is used,

compared to i.e. telephone or chat, because

immediate feedback or enquiries are not possible.

For the design of communication technology this

means that it should be appropriate for the task: All-

embracing tools catering for diverse communication

needs might have to be viewed critical. For example,

it is questionable whether technology can convey

high as well as low levels of social presence.

E. Technical Communication Problems:

Technical communication problems might arise in

situations when the technology itself assumes the

role of a communicator, or communication processes

are automated by the technology, respectively.

Maass & Oberquelle (1992) call this an “agent

perspective” in software development: The computer

assumes the role of an equal partner who is able to

interpret and react to user behavior in an adequate

way. They criticize that such system behavior is

seldom transparent for users. Furthermore, an

anthropomorphic presentation of technology might

raise unrealistic expectations among users regarding

the computer’s actual communication skills.

In computer-mediated communication there is a

broad spectrum of autonomous communication acts

reaching from simple automatisms to complex

interpretations. One example is inferring users’

emotional states by interpreting certain emotional

keywords in the texts (Perry & Donath 2004).

Another example is the e-mail recovery function

offered by some e-mail clients, allowing users to call

back (still unread) e-mails under certain

circumstances. However, recipients using different

e-mail clients usually receive a strange automatic

reply, offering neither an excuse nor request for

confidentiality etc. like one would expect from a

human sender. Still, by offering such a function the

notion of a competent communicator is evoked, who

is able to handle a possibly embarrassing or

awkward situation. The following anecdote might

illustrate this: A colleague receiving such an e-mail

reading “Mrs. X wants to call back this e-mail”

wrote back, amused: “But Mr. Y doesn’t want to

give that e-mail back!”

3 PRACTICAL IMPLICATIONS:

A CASE STUDY

The following sections illustrate how the

multidimensional model can be used to analyze

CMC problems and derive design implications,

using a case study of an educational groupware

system, which was evaluated by means of an online

survey with more than 1500 users.

3.1 Software and use Context

The software CommSy is a web-based groupware

system to support communication and coordination

in working and learning groups, comparable to e.g.

BSCW (e..g. Klöckner 2002), phpBB (e.g. Stefanov

et al. 2005), or Moodle (e.g. Cole & Foster 2007). It

supports communication (for example with news

and discussion forums) and the exchange of working

materials (with e.g. file uploads and online

documents) and also offers project management

functions (e.g. a shared calendar and to-do lists).

The groupware consists of so-called workspaces:

– Project Workspaces are designed for use in

closed groups of approximately 10 to 30

members (e.g. student groups). To support group

activities, the software offers groupware

functionalities, such as announcement of news or

events, discussion forums, personal homepages

for the presentation of members to the group and

materials that can be written in a cooperative

way, collected and classified by the users.

– Community Workspaces incorporate project

workspaces into a larger structure supporting not

only small groups, but a community of users

(e.g. all members of a school or university) over

a longer period of time, similar to an intranet

structure.

CommSy was developed for use in educational

settings, both secondary and undergraduate

education. It has been used in a variety of teaching

fields, including history, languages, economics, and

informatics. Furthermore, the groupware is also used

in the public and private sector, for example in

public administration or business networks. CommSy

is Open Source software distributed under the GNU

General Public License. A description of the

software, the development process and its use

contexts is given e.g. by Bleek & Finck (2005),

Janneck et al. (2006), Pape et al. (2002).

3.2 Analyzing Problems in

Computer-mediated

Communication

Throughout the last five years, the use of the

groupware in secondary and post-secondary

education has been regularly evaluated by means of

an online survey with several thousand respondents

(Janneck 2007). In this paper, we specifically refer

to the latest survey, which was conducted in the

A MULTIDIMENSIONAL MODEL TO ANALYZE SOCIAL AND TECHNICAL FACTORS IN

COMPUTER-MEDIATED COMMUNICATION

523

spring of 2008. 1538 users participated in the

survey, mainly university students and teachers

(63%). 32% were secondary school students or

teachers, 5% from other use contexts (e.g. business

networks). 64% were female, 36% male.

The survey covered usage preferences and

characteristics such as frequency and most-used

features, usefulness, usage problems, and usability

measures. In this study, we focus on usage

problems. Users were asked to report all problems

they had encountered while using the software. They

were able choose from a list of possible usage

problems, which were known to occur from prior

evaluations, as well as add and describe all other

problems they had encountered. A total of 1120

problems were reported (multiple answers were

possible). The responses were then categorized

according to the five dimensions described above.

Figure 2 shows the distribution.

Figure 2: Distribution of usage problems regarding the

five levels of the model.

Most frequently, users complained about a lack

of interest and active participation by other users

(22% of all mentions), problems regarding

collaborative structuring and editing content (13%),

and to a lesser extent, an insufficient introduction to

use (8%) and a lack of incentives for use (4%).

These problems can be attributed to an insufficient

moderation of use and, thus, located on the

dimension of Human Communication Problems (A)–

the largest category with a total of 47% of mentions.

Technical Errors and Failures (C) constitute the

second largest category with a total of 25% of

mentions. This is made up by software

bugs/incompatibilities, especially with the browser

(14%), and long response times (11%).

Media Characteristics (D) make up a

comparable share of mentions (23%), summing up

functions not suitable for the use context (e.g. lack

of synchronous communication tools). Furthermore,

there were deficits in supporting awareness (see

section 3.3).

Problems arising from Individual Characteristics

(B) were hardly mentioned (5%), which might be

due to the self-report data. Some users mentioned

difficulties regarding the metaphors and terms used

in the software. For example, some users tried to use

the so-called “clipboard” (which is used to copy

posts within the groupware workspace) to save files

to their desktop, inferring from the usage of the

familiar Windows clipboard. Furthermore,

especially teachers reported that some participants

were reluctant to participate in computer-mediated

communication.

Technical Communication Problems (E) were

irrelevant in the case study, because the groupware

offers hardly any automatic or interpretative

functions.

This problem analysis is exemplary and shows

an individual profile, which cannot be generalized to

other software tools or even other user groups of the

same system. Nevertheless, it is an interesting result

that the bulk of usage problems that were reported

can be attributed to personal and social factors, and

cannot be tackled by the software design. Instead,

they would have to be addressed by measures to

better support and supervise software use. Second

come seemingly “trivial” technical problems.

Problems requiring conceptual changes in the

software design only come in third. Thus, a strategy

relying primarily on changes to the software design

and functionalities would presumably be rather

ineffective in this particular case.

3.3 Implications for Shaping CMC

Tools and Processes

After using the multidimensional model to analyze

problems in CMC settings, the following section

highlights how it can be used to draw conclusions

for the design of CMC tools and computer-mediated

interaction processes. The model helps to distinguish

whether to address a certain problem on a design

level or a personal, social, or organizational level.

A. Human Communication Problems:

Clarify whether the difficulties that can be observed

are actually due to the technology or rather due to

difficulties in social interaction processes, which

can’t be addressed by the software design.

Æ In the case study, the most frequent problems

could be traced back to deficits regarding the

didactical integration of the tool into the course

WEBIST 2009 - 5th International Conference on Web Information Systems and Technologies

524

design. Those problems might be addressed by

providing better occasions for use or agreeing upon

common rules for usage (e.g. frequency of use).

B. Individual Characteristics:

The current state of knowledge and the experiences

of the (prospective) user group should be analyzed

and taken into account when selecting or designing

CMC media (e.g., simple tools might be more

appropriate for an inexperienced user group).

Training and support measures must be adapted to

users’ experience and needs.

In the case study, teachers who observe

reluctance among students to communicate online

might try to establish a playful approach to CMC,

e.g. by letting students communicate informally

before using the tool in class.

C. Technical Errors and Failures:

Of course, data loss due to technical errors must be

reduced to a minimum. Furthermore, especially with

communication technology it is important to provide

transparent error handling processes, allowing the

user to understand what problems occurred and

whether the message transmission was impaired.

Regarding the case study, users should receive

information about frequent browser or software

incompatibilities.

D. Media Characteristics:

Check whether the CMC tool/software is appropriate

for the users’ purposes and tasks and whether the

tool has a coherent overall design (rather than

integrating contradictory functions or concepts of

use).

In the case study, the analysis revealed a lack of

awareness (e.g. Gutwin & Greenberg 1999, Dourish

& Bellotti 1992) functions (e.g. a “who is online”

display or access statistics), which can be addressed

by the software design (Janneck 2007, 2009).

E. Technical Communication Problems:

Check if and to what extent the CMC tools influence

the human communication process (e.g. by

automatic functions) and whether this is actually

necessary and appropriate in the respective context.

Make sure users get transparent feedback if (and

how) the communication content or process is

altered by the software.

In the case study there is no need for action in

this regard.

4 CONCLUSIONS

This paper presented a multidimensional model to

analyze problems in computer-mediated

communication (CMC) regarding to five levels of

potential difficulties and demonstrated its use for the

analysis of CMC situations by means of a case

study. Likewise, practical implications for the design

of CMC tools and the social and organizational

frame were exemplarily shown.

Those implications and recommendations

support and refine existing design guidelines such as

the Dialogue Principles established in the DIN EN

ISO 9241-110 (2006) or guidelines referring more

specifically to the design of communication and

cooperation software (e.g. Dieberger et al. 2000,

Erickson 2003), as well as basic psychological

principles of human-computer interaction such as

the use of metaphors in design. The specific

contribution of the model presented here is to

distinguish clearly between technical and social

factors influencing CMC.

Methodically, empirical and explorative

evaluation approaches (such as surveys, interviews,

or usability tests) should be chosen to grasp the

users’ subjective views and experiences. For that

reason, evaluations shouldn’t rely solely on expert or

inspection methods (e.g. usability walkthroughs or

heuristic evaluations).

From a research perspective the model helps to

clarify and understand the different factors

influencing CMC and the success or failure of

interaction processes. Nevertheless, it has to be

noted clearly that the model cannot be seen as a full-

flegded ‘theory’. It serves as a frame, which can

integrate existing CMC and, generally,

communication theories on the different levels.

Furthermore, it adds to those mainly psychological

analysis models the perspective of systems and

sociotechnical design.

For further refinement and substantiation of the

model, more empirical work is needed. Thus, a next

step will be to study other CMC tools, settings, and

user groups and compare them regarding the

distribution of problem dimensions to see if the

model continues to serve as a useful analytic tool or

whether new categories emerge from the empirical

material. Furthermore, potential interactions and

interdependencies between the different levels

described by the model will have to be investigated

and discussed.

A MULTIDIMENSIONAL MODEL TO ANALYZE SOCIAL AND TECHNICAL FACTORS IN

COMPUTER-MEDIATED COMMUNICATION

525

REFERENCES

Birnie, S. A., Horvath, P., 2002. Psychological Predictors

of Internet Social Communication. Journal of

Computer-Mediated Commmunication 7, 4.

Bleek, W.-G., Finck, M., 2005.Ensuring Transparency -

Migrating a Closed Software Development to an Open

Source Software Project.In IRIS28: Proceedings of

the 28. Conference on Information Systems Research

in Scandinavia, 6-9 August 2005, Kristiansand,

Norway.

Bordia, P., 1997. Face-to-face versus computer-mediated

communication. A synthesis of experimental literature.

Journal of Business Communications 34, 99-120.

Burnett, G., 2000. Information Exchange in Virtual

Communities: A Typology. Information Research: An

International Electronic Journal 5, 4.

Burnett, G., Buerkle, H., 2004. Information Exchange in

Virtual Communities: A Comparative Study. Journal

of Computer-Mediated Commmunication 9, 2.

Cole, J., Foster, H., 2007. Using Moodle. O’Reilly.

Daft, R. L., Lengel, R. H., 1986. Organizational

Information Requirements, Media Richness and

Structural Design. Management Science 32, 554-571.

Dieberger, A., Dourish, P., K. Hoek, K., Resnick, P.,

Wexelblat, A., 2000. Social Navigation: Techniques

for Building More Usable Systems. Interactions, 7, 6,

36-45.

DIN EN ISO 9241-110, 2006. Ergonomics of human-

system interaction–Part 110: Dialogue principles.

Beuth.

Dourish, P., Bellotti, V., 1992. Awareness and

coordination in shared workspaces. In Proceedings of

the 1992 ACM Conference on Computer-Supported

Cooperative Work, ACM Press, 107-114.

Erickson, T., 2003. Designing Visualizations of Social

Activity: Six Claims. In Conference on Human

Factors in Computing Systems–CHI ’03, ACM Press,

846-847.

Gutwin, C., Greenberg, S., 1999. A Framework of

Awareness for Small Groups in Shared-Workspace.

Technical Report, Department of Computer Science,

University of Saskatchewan, Canada.

Herrmann, T., 1993. Loss of Situative Context and its

Relevance for Computer-Mediated Communication

and Cooperation. In Proceedings of the IFIP WG9.1

Working Conference on NetWORKing, North-Holland,

87-96.

Hian, L. B., Chuan, S. L., Trevor, T. M. K., Detenber, B.

H., 2004. Getting to Know You: Exploring the

Development of Relational Intimacy in Computer-

mediated Communication. Journal of Computer-

Mediated Commmunication 9, 3.

Janneck, M., 2007. Quadratische Kommunikation im Netz:

Gruppeninteraktion und das Design von CSCL-

Systemen [Group interaction and the design of CSCL

systems]. Lohmar: Eul.

Janneck, M., 2009. Designing for social awareness of

cooperative activities. In: Proceedings of WEBIST

2009, Lisboa, Portugal, March 2009, INSTICC Press.

Janneck, M., Obendorf, H., Finck, M., Janneck, Mi., 2006.

Tying Collaborative Knowledge Nets. In: Hassanaly,

P., Herrmann, T., Kunau, G., Zacklad, M. (eds),

Cooperative Systems Design–COOP 2006,

Supplement to the Proceedings, 57-65.

Klöckner, K., 2002. BSCW. Cooperation support for

distributed workgroups. In Proceedings of the 10th

Euromicro Workshop on Parallel, Distributed and

Network-based Processing, 277-282.

Maaß, S., Oberquelle, H., 1992. Perspectives and

Metaphors of Human-Computer-Interaction. In Floyd,

C., Züllighoven, H., Budde, R. & Keil-Slawik, R.

(eds): Software Development and Reality

Construction, Springer, 233-251.

McKenna, K. Y. A., Bargh, J. A., 2000. Plan 9 from

cyberspace: The implications of the Internet for

personality. Personality and Social Psychology Review

42, 1, 57–75.

Pape, B., Bleek, W.-G., Jackewitz, I., Janneck, Mi.,

2002.Software requirements for project-based

learning - CommSy as an exemplary solution.In

Proceedings of the 35th Hawaii International

Conference on System Sciences 2002.

Perry, E., Donath, J., 2004) Anthropomorphic

Visualizations: A New Approach for Depicting

Participants in Online Spaces. In Conference on

Human Factors in Computing Systems–CHI ’04, ACM

Press, 1115-1118.

Shannon, C. E., Weaver, W., 1949. The mathematical

theory of communication. University of Illinois Press.

Sikkel, K., 1997. A Group-based Authorization Model for

Cooperative Systems. In Proc. of ECSCW'97: Fifth

European Conference on Computer Supported

Cooperative Work, Kluwer Academic Publishers, 345-

360.

Sohlenkamp, M., Prinz, W., Fuchs, L., 2000. PoliAwac -

Design and Evaluation of an Awareness Enhanced

Groupware Client. AI and Society–Special Issue on

CSCW, 14, 1, 31-47.

Sproull, L., Kiesler, S., 1999. Connections: New Ways of

Working in the Networked Organization. MIT Press.

Stefanov, S., Rogers, J., Lothar, M., 2005. Building Online

Communities with Phpbb 2. Packt Pub.

Whitty, M.T., Gavin, J., 2001. Age/sex/location:

uncovering the social cues in the development of

online relationships. In CyberPsychology & Behavior

45, 623–630.

WEBIST 2009 - 5th International Conference on Web Information Systems and Technologies

526