Complete Site Outline
Write to me:
"The future is likely to belong to those organizations
that never stop asking, `How can we better organize and manage ourselves?'"
(Galbraith and Lawler, 1993.)
What determine managers' choice of communication media? Why do
managers prefer to use one communication media over another? Within
the domain of information system and communication research there
exists a substantial body of theories that explain manager's media
choice. The purpose of this chapter is to review the theories on
media choice and to identify the deficiencies of these theories.
The early theories on media choice tend to be from the rational
and individual school of thought. The first theory is the social
presence theory of Short, Williams and Christie (1976). This theory
emphasises the psychological aspect of using communication media:
media choice hinges upon the ability of the media to convey the
nature of the relationship between the communicators. In this regard,
communication media can be described as warm, personal, sensitive
A few years later, Daft and Wington (1979) proposed a language
variety theory to explain media choice. It evolved from the idea
that certain media, such as painting or music, are capable of conveying
a broader range of ideas, meanings, emotions compared to mathematics.
Such media have a higher language variety. This led to the suggestion
that language variety needs to be matched with the communication
task. Equivocal and complex social tasks are said to require a medium
with high language variety.
Obviously, this is not directly applicable to manager's choice
as they do not use painting or music as a mean of communication,
but this notion lays the foundation for media richness theory. This
is where the discussion will turn to next.
The previous two theories are fragmented and non-comprehensive.
With media richness theory a much fuller picture of media choice
emerged. Daft and Lengel (1984, 1986) were the first to propose
the media richness theory. The media richness theory is a general theory
that explains managerial behaviours as well as organizational design. It became one of the most popular and widely studied
models of media choice because of its well structured and intuitive
Media richness theory establishes a framework that ranks communication
media along a continuum in terms of their "richness". In this context,
"richness" denotes the capacity of the media to: (i) carry large
volume of data, and (ii) convey meaning.
More specifically media richness refers to the ability of the media
to change human understanding, overcome different conceptual frames
of reference, or clarify ambiguous issues in a timely manner (Daft
and Lengel 1984, 1986). Thus, where the mode of communication provides
new substantial understanding it is considered "rich"; otherwise,
it is "lean".
Four criteria (the original criteria) were used by Daft and Lengel
(1984) to classify communication media along the continuum. These
are summed up in Table 2.1.
These criteria denote the qualities of rich media; their attributes
impact upon human understanding and frame of reference (Daft and
Lengel 1984, 1986). Consequently, communication media possessing
more features of the criteria would rank higher on the richness
scale compared to one possessing less. For example, using these
criteria, oral media (eg. face-to-face and telephone) are believed
to be richer than written media because they provide opportunities
for immediate feedback and can have multiple cues in a natural language
tailored to the circumstances. Typically, synchronous media (ie.
with immediate feedback) are considered richer than asynchronous
media (ie. involve delay in the communication process). So telephone
was ranked lower than face-to-face because it can transmit fewer
cues (ie. cannot see the other person). An example of the ranking
is illustrated in Table 2.2.
Daft and Lengel (1984) presented a contingency framework that
matches two concepts: (i) the nature of the information task, and
(ii) the characteristics of communication media as represented by
the media richness scale. Drawing from the work of Weick (1979)
and Galbraith (1973), Daft and Lengel distinguished two information
tasks carried out by organisations: (a) interpretation of external
environment, and (b) coordination of internal organizational activities.
Recognising the complexity of the social system that constitutes
an organisation, both aspects are argued to be required in an information
processing structure to handle the transfer of a vast volume of
information and to significantly reduce ambiguity.
Despite the importance of both aspects the media choice literature
has somehow gravitated towards the latter.
Ambiguity reduction is seen as the defining characteristic of a
rich media. It is relevant to both information tasks. To successfully
interpret the external environment, for example in formulating a
strategic plan, organisation members need to establish a shared
view of events, because external data are often unstructured, ambiguous
and sometimes downright confusing. So Daft and Lengel (1984) argued
that an organisation information processing system needs to reduce
ambiguity, or more precisely, equivocality. Formally, "equivocality"
is defined as:
"[T]he extend to which data are unclear and
suggest multiple interpretations about the environment...[such that
it] is reduced through shared observation and discussion until a
common grammar and course of action can be agreed on"(Daft and Weick,
1984, at p291).
In other words, where a matter is equivocal managers would be
confused and disagreed with one another; they would not know how
to address the matter. This is because they lack mutual understanding
about the matter, and until they can come to an understanding they
would not be able to proceed any further. In this regard, equivocality
is very different to "uncertainty", which is caused by an insufficiency
of information to perform the task. This can be easily cured by
any additional information regardless of the medium of communication
used. Equivocality, by contrast, requires that a shared meaning
or understanding be reached between the parties.
Given that rich media are supposed to facilitate understanding
it is argued by Daft and Lengel as the ideal media to handle equivocal
task. Thus the greater the equivocality in a communication task,
the richer the media that is required. This is akin to a contingency
type logic claiming effectively that a particular communication
task needs to be matched with a particular communication media to
enable efficient information processing within the organisation.
This approach is summed up neatly by this definition of the richness
"a process by which individual managers rationally
attempt to match the characteristics of the communication media
at their disposal to the requirements of their communication tasks
in order to achieve personal and organizational effectiveness"
This relationship is illustrated in Figure 2.1.
This logic is generally consistent with managers' usage pattern
of communication media in practice: managers do seem to choose media
by matching it to the richness of the media. Research such as that
of Mintzberg (1972) did find that managers preferred to use face-to-face
means of communication rather than the more formal written reports
because such formal reports were said to be untimely and inadequate
for the non-routine and ambiguous tasks of managers. Under the richness
scale, face-to-face communication is the richest medium.
In sum, as a theory media richness can be regarded as both prescriptive
and descriptive. Media richness theory is a prescriptive model in
that it posits that organizational effectiveness required a match
between information processing requirement (eg. ambiguity reduction)
and communication media (eg. face-to-face, writing, etc): Daft and
Lengel (1984, 1986). It is descriptive insofar that it explains
and provides reasons for the choice of media in different circumstances
(Daft et al, 1987; Daft and Lengel, 1990; Trevino et al, 1990)
The original criteria are based on the traditional mode of intra-organizational
communication. In particular, it emphasises the strength of face-to-face
communication. By using face-to-face communication as a benchmark
for comparison (Culnan and Markus, 1987), the presumption is that
face-to-face is the optimal means of communication. The fact that
managers generally preferred to communicate face-to-face provides
some support for this presumption.
Relying on this presumption, a persistent view in the literature
is that an electronic mode of communication is not a rich medium
because it lacks the qualities that are deemed to be high in richness.
Email, for example, is very low down the scale in Table
2.2 because as a text based medium it has no social presence
and has limited ability to transmit any cues. This in turn leads
to the general claim that where managers perform equivocal tasks
they would rely much less on modern communication media (eg. Trevino
et al, 1987; Daft et al, 1990).
However, the original Daft and Lengel criteria were not design
with modern communication media in mind. Indeed, in a study of interactive
media, such as email, the notion of richness has been shown not
to be an inherent property of the medium (Lee, 1994). Consequently,
it is highly questionable that the notion of `richness', at least
as based on the original criteria, can provide a fair basis of comparison
between traditional and modern communication media. In seeing face-to-face
as a de facto standard or an optimum media, the results are
naturally biased against modern communication media.
Modern communication technologies have qualities not found in
traditional communication media. This leads some to argue that modern
communication medium, such as email, is in fact much richer than
was the case using the original criteria (Markus, 1994; Sproull,
1991). Sproull (1991) was the first to provide an updated definition of richness
by incorporating the value-added features of modern communication
technologies, especially email and groupware. Subsequently, Valacich
et al (1993) identified another related quality. The four addition
criteria are outlined in Table 2.3.
Anyhow,it can be argued that the focus of media richness theory
is still on face-to-face. In fact, it is to be doubted whether "richness"
is an appropriate notion to assess modern communication technologies.
The notion of richness just does not seem to cover the qualities
possessed by modern communication technologies; these qualities
go beyond `richness' and the notion of communication task equivocality.
The research papers studying media richness are summarised in Table
There are numerous studies providing empirical support for media
richness theory. The primary claim of the theory that managers will
choose rich media (as defined under the original criteria) in situation
where the communication task is high in ambiguity has been established
in many studies (Daft et al, 1987; Russ et al, 1990; Trevino et
al, 1990; and Whitfield et al , 1996).
Complementing such survey style studies is an investigation by
Trevino et al (1987) in which the managers are interviewed as to
the incidents and rationale for choosing particular communication
media. The responses are overwhelmingly that fact-to-face communication
(a rich medium) is preferred in situations where the messages to
be conveyed are ambiguous.
While there is a solid body of research supporting the position
of media richness theory, an equally large number of research studies
provide confounding evidence. Indeed, empirical support for media
richness theory is particularly weak in relation to modern communication
media like email or voice-mail (Markus, 1994; Markus et al, 1992;
Rice and Shook, 1990).
It is no surprise that research giving conflicting evidence tends
to be of more recent origin. With the incorporation of more modern
communication technologies in the research design, studies are only
recently in a position to expose the predisposition of the media
richness theory towards traditional communication media. The increased
prevalence of technology usage in organisations in general may provide
In a meta study by Rice and Shook (1990), the prediction of media
richness theory was found not to hold. The inadequacy of the theory
really became apparent in a study on email usage by Markus (1994).
Markus (1994) conducted a fairly comprehensive study involving a
survey of 504 managers and an interview with 29 personnel (whose
positions range from chairman to administrative assistant), and
collection of archival data. The study found that managers do not
regard email to be particularly rich. This is consistent with the
richness model under the original criteria, but is inconsistent
with the updated criteria. Yet the most curious finding is that
mangers used email substantially more than the theory predicted.
This study effectively weakens the media richness theory: the theory
was not found to provide an adequate explanation and the updated
criteria were also found to be incapable of capturing managers'
perception of richness.
Why? Using evidence from the interviews, Markus concluded that
social processes provide a superior explanation of media choice.
And this position is consistent with other studies: Schmitz and
Fulk (1991), Schmitz (1987) and Steinfield and Fulk (1986).
Likewise, D'Ambra (1995) demonstrated media richness theory failed
to fully explain media choice in an Australian study focusing on
the use of voice mail in organizations.
The reliability and validity of two fundamental constructs of media
richness theory are still not clear. While there are plenty of papers
examining the predictive and explanatory power of the media richness
theory, and there are solid empirical justifications for the link
between equivocality and media richness (Daft et al, 1987; Trevino
et al, 1987; and Russ et al, 1990); the underlying construct of
media richness and task equivocality have received limited attention
in the literature. Only recently did a paper by D'Ambra (1995) investigate
In most studies media richness is not something that is explicitly
measured. This can be seen in the summary in Table 2.4.
Often, it is simply determined by its position on the continuum
of richness derived from the set of original criteria noted above.
In other words, media richness is treated as an "invariant objective
features" of each communication media (Schmitz and Fulk, 1991).
The central problem is that these original criteria are themselves
outdated with no recognition of the very different qualities of
modern communication technologies. After investigating the matter
D'Ambra (1995) came to the conclusion that the reliability of the
media richness scale is much weaker at the "lean" end of the scale.
And at this lean end, of course, are found the modern communication
technologies. D'Ambra claimed that this highlights the inability
of the traditional media richness scale to capture the full range
of attributes or qualities of these modern communication media.
Again, the verdict is very similar in relation to task equivocality.
In most prior studies equivocality is either not explicitly measured,
or if it is, it is determined by a panel of judges and not by the
subjects of the media choice research.
Moreover, although the scales used was found to be "adequate", D'Ambra
(1995) noted that its operationalisation is "problematic".
This is not surprising because a cursory examination of the literature
on task equivocality and uncertainty reveals that although this
construct has a long theoretical history and is well theorised conceptually,
its operationalisation is through the imperfect proxy of task uncertainty.
Under media richness theory task characteristics refer primarily
to the notion of equivocality. The question that arises is why just
equivocality? A communication task can possess many other task characteristics.
For example, the task could involve different locations, or there
may be a time limit, or it may involve more than one person, etc.
These other task characteristics are further considered in Chapter
Failure to consider these task characteristics make prior studies
of media choice highly artificial. While the narrow perspective
seems to serve its purpose within the context of media richness
theory, it is inadequate to provide an understanding of the overall
media choice process. Indeed, as established subsequently in this
thesis, equivocality alone is inadequate to explain the media choice
process. This narrow focus also tends to be biased against modern
communication media. This could explain some of the curious and
conflicting results in prior studies.
The rational logic in media richness theory may provide an intuitive
foundation to explain and predict media choice, but the current
approach in measuring media richness has a critical flaw: media
richness is treated as an "invariant objective feature".
It is assumed to be the same regardless of context or person; individual
mangers are assumed to be aware of all the inherent characteristics
of each media, and make choice on the basis of these characteristics.
This is unrealistic. People have imperfect cognitive abilities;
and the perception of one person is not necessarily the same as
another. This is very different to the model assumption that people
somehow perceive and act in the same objective manner. Such
unrealistic assumptions represent the focal point of attack from
the social orientated school of thought:
"There is no simple relationship between
message and medium... The rational model of media choice fails to
suffice in its explanation of managerial communication behaviour
when managers' preferences for media are taken into accounting and
when on considers personal communication-related variables..."
In addition, attempts to somehow categorise modern communication
technologies, such as "multi-media" technology, along a continuum
of richness may be rather meaningless in that the new media technologies
are very fluid such that they encompass the characteristics of more
than one traditional mode of communication.
The questionable reliability and validity of the measures for media
richness and task equivocality plus the mixed results in the literature,
especially in relation to modern communication media do indicate
that "media richness theory does not fully explain media choice
in an organization setting."
This is consistent with the argument that the notion of media richness
is not design with modern communication media in mind. Even with
the inclusion of the qualities of modern communication technologies,
the notion still tends to favour the more traditional media.
The bias against modern communication media is amplified by the
focus on only one task characteristics. There is little doubt that
task equivocality is important and insightful in explaining mangers'
choice of communication media, but in reality there are many other
variables that impact on the requirement of each communication task.
It is submitted that the narrow focus of previous media choice research
has deprived the studies of additional explanatory power in relation
to the overall media choice process. This may even explain the mixed
finding in previous literature.
The inability of the richness theory to explain the confounding
evidence mentioned above have given rise to another body of theories
known as the "collective-level" theories that focus on the broader
collective variables like structure, environment, culture and politics. These theories are not based on rational choice alone but
are a "collective behavioural response to a socially-constructed
definition of the medium's appropriateness."
Under these theories, "richness" is an outcome of social behaviour
not the cause of social behaviour (in relation to media choice).
In other words, the social context provides a framework to assess
the different mediums by emphasising certain media characteristics
and guiding the interpretation of them. The features of social and
rational theories are compared in Table 2.5.
The most comprehensive theory in this regard is the social influence
model of technology use (Fulk et al, 1987; Fulk et al, 1990). It posited that individual media's perception is not just
a function of objective (rational) choice, but is partly a social
construction. Fulk, et al (1990) found support for the two main
limbs of this model.
Table 2.5: Social and Rational Theories Compared
* All salient to the individual.
* Still relevant; but importance decline relative to social
* Not all features are salient.
* People analyse the media features objectively;
* An objectively rational process that is based on the salient
and invariant media characteristics.
* A subjectively rational process;
* Individuals develop socially acceptable interpretations
and behaviours that is justifiable with social norms or expectation;
ie. influenced by others.
|Role of Social
* Social context plays a significant role in media choice;
* Social cues influence individual's interpretation of the
objective requirement of the communication task;
* Establish social norms that inhibit media use.
|Examples of Theories
|Social Influence Model;
First, properties like "richness" are posited to be subjective
and influence by attitudes, statements and behaviours of other people.
So unlike classical richness theory, it recognises that people's
perception does differ, that richness is not an objective function
of some inherent characteristics of the media. This is based on
the recognition that the social context can provide the norms and
expectation that shape people's perception and behaviours. For example, in an office where everyone does
not use email and make negative comments in relation to it, any
particular individual would be less likely to regard email as rich.
Related to this idea are papers that invoke situational variables
as an influence of media choice: Markus (1986), Steinfield and Fulk
(1986), and Trevino, et al (1987). There is the critical mass theory
by Markus (1987). Steinfield and Fulk (1986) and Cadwell, et al
(1995) found evidence suggesting that geographical distance, amount
of message content and time pressure (ie. urgency of message) provide
a strong influence on managers' media choice. In addition, Trevino,
et al (1987) found that communication media has symbolic meaning
associated with. Depending on the shared meaning evolved through
time by the mutual interaction of people each media will be implicitly
understood to serve a particular purpose or to be used in particular
situations. This shared social understanding dictates media choice
Also, this aspect of social construction theory is related to
institutional theory. Loosely
speaking, institutional theory claimed that the norms, values, standards,
expectations engendered by the institutional environment shape also
the norms, values, standards, expectation of the individual (Scott,
1987). For example, where it is the accepted practice of the industry
to use email for communication, a firm and the people working within
it are likely to follow the same practice such as to give them a
sense of legitimacy. On the basis of this theory, Markus (1987)
concluded that "whereas information richness theory predicts that
the uses of new media will be relatively invariant across social
units, institutional theory predicts that social definitions, and
hence also uses, of new media will vary, perhaps dramatically across
Second, the social influence model incorporates the idea that
expertise with a medium also influence media use. For example, lack
of skill will inhibit uses, or that the medium may not be perceived
to be as "rich" (as otherwise) if the user lacks the skill or training
to use it properly. In relation to email, this is supported indirectly
by Komsky (1991) which found that frequent users of email tend to
be more tolerant of system problems and to perceive the medium as
easy to use.
The critical mass theory posits that the choice to use interactive
medium (due to its unique characteristics) depends largely on whether
the interactive media is accepted by the society at large, a "universal
access medium", because until that happens people would be reluctant
to pioneer the use of the media (Markus, 1987; Huber and Daft, 1987). This pre-condition for the choice of communication media
makes intuitive sense; and is clearly related to the institutional
theory noted earlier: the more people that use a particular medium
for a particular purpose the more likely one is to use that medium
for that purpose.
A further look at the literature reveals consistency with theories
on IT maturity. This notion essentially conceptualised the extent
that the organization accept IT.
From the idea that IT maturity evolved in stages,
it is arguable that the communication media choice is likely to
vary along with it. It is reasonable to say that the extent that
IT has permeated the organization's infrastructure and culture would
impact upon people's choice of communication media. The more that
IT has permeated the organization, the more that people are likely
to use modern communication technologies. This intuitive argument
is consistent with current media choice theories:
(i) Social construction media choice theory suggests that social
context can supply the norms and expectation that influence individual
perception. Now, the higher the level of IT maturity, the more likely
that usage of modern communication technologies would be the norm
and hence reinforce the social acceptance of the medium. In other
words, where there is high IT maturity, the usage of modern communication
technologies could become an institutionalised phenomenon, thereby
leading to a culture that values and encourages the use of modern
(ii) Where there is high IT maturity, people's expertise about
modern communication media would also expect to be high by definition.
Thus, people are less likely to be inhibited from using the technology.
(iii) With the advent of IT, there are more communication media
offering benefits different to that of the traditional media. Thus,
even judging from the inherent characteristic perspective alone
media choice is likely to vary as a result.
The above discussions have greatly subsumed the adequacy of the
rational approach, especially media richness theory. However, this
thesis maintains a `loose' rational framework.
This thesis is not denying the validity of the social perspective.
Rather than viewing the social perspective as an alternative, it
is best seen as complementary. The social perspective never denies
the relevance of the objective features of media, only that it emphasises
that "richness" is more a function of social behaviours or construction.
Although the fuzziness or chaotic feel associated with the social
construction approach seems totally inconsistent with a rational
approach, the point to note is that the two perspectives are not
mutually exclusive. This argument is consistent with modern chaos
and self-organisation theory that claims there is order in chaos.
Until now the literature tends to focus either on the rational
or the social perspective. While the rational approach is more systematic
and provides a clearer structure of the media choice process, it
lacks the richness of detail offered by the social theories. Yet,
a rational approach is more practical in that it provides a framework
that can facilitate the analysis and design of organization information
processing system. This account for the continuing popularity of
individual rational choice approach (Markus, 1994).
Meanwhile, social theories may seem rather fuzzy but the process
is not irrational. The social elements can themselves be
rationalised. Take symbolic interactionism for instance, which claims
that people use media to signify something through the socially
agreed meaning of different communication media. For example, people
may want to convey a sense of authority by communicating in writing.
In this context, if managers are asked why they chose to communicate
in writing, they would respond by saying that want to show their
authority. Is this not rational?
Consequently, this thesis presents a more comprehensive framework,
in Chapter 3, that integrates the rational
and social elements in a "loose" rational framework. It is "loose"
because the framework recognises that people are not perfectly rational.
People do not function in a mechanistic manner. The social basis
of people means that their decision process is necessarily complex. Indeed, this lack of perfect rationality could
provide another explanation for the deficiency in the traditional
media richness theory. In this regard, transaction cost economics
may offer some insight.
Literature on media choice has provided many explanations for media
choice. Media richness theory, for instance, posits a match between
the task equivocality and the characteristics of the communication
channel. There are also many social theories that supplement this
rational model. Yet inadequacies of these theories are apparent
in the literature particularly in relation to modern communication
The once dominant rational approach seems to have been subsumed
by the rise of social type theories. Yes, social theories do offer
a fresh and softer perspective but another explanation may be that
the traditional rational approach is too narrow. The failure to
examine the effect of other task characteristics has greatly restricted
the ability of these theories to explain the overall media choice
It is submitted that a broader perspective needs to be adopted,
one that consolidates the different theories on media choice, whether
social or rational, and incorporates a fuller and more realistic
range of communication task characteristics. To this end a comprehensive
framework that elaborates the media choice process is developed
in the next chapter.
 This is also known as the information
 The origin of the theory can
be traced to the PhD dissertation of Bodensteiner (1970). It is
grounded upon the "information processing theory" in the field of
 Because of the limited scope
of this thesis, the organization design and effectiveness aspects
of media richness theory will not be addressed.
 More generally, Daft and Lengel
(1984) at p196, defined "richness" to mean "potential information-carrying
capacity of data." Arguably the reference to `data' is somewhat
confusing if used in relation to communication channel because data
are transfer through the channel. The point to note in relation
to Daft and Lengel's definition is that the data, as well as the
channel, can be described in terms of "richness". In this paper,
richness refers to the latter only.
 Ie. The literature tends not
to focus on the coordination aspect and the ability of the media
to transfer information within the organization. Yet, organisations
being a social system is bound to experience equivocality during
internal coordination processes. There are bound to be some disagreement.
So ambiguity reduction is equally important in relation to internal
 Markus (1994) at p505.
 Sproull, R. (1991).
 This argument is made clearer
during the Chapter 5 discussion of the findings.
 See Section
 The exceptions are: Fulk and
Ryu (1990), Trevino et al (1990) and D'Ambra (1995).
 A point further discussion
in Section 2.4.4.
 D'Ambra (1995) did get the
subjects to determine equivocality in part of his study.
 D'Ambra (1995), p177 &
 Alexander et al (1991),
p170. In this study it was found that writing apprehension can influence
managers' choice of media. See also Section
 The transaction cost economic
viewpoint presented in Chapter 3 eases this
 This would particularly
be the case with groupware (eg. Lotus Notes). This is an aspect
that is not addressed in this thesis.
 D'Ambra et al (1997), p18.
 See Romm, Pliskin and Rifkin
(1996) for a summary of these variables.
 Markus (1994), p522.
 In Fulk et al (1987) it
was referred to as the "social Information processing Model" because
of the psychological origin of the model.
 This is grounded upon the
Social Information processing theory in psychology (Salancik and
Pfeffer, 1977, 1978; Pfeffer, 1982).
 Markus (1994) referred to
this as "social definition theory". Although he focuses on institutional
theory, but he also includes under this umbrella term: (i) structuration
theory; and (ii) social construction of technology theory.
 Markus (1994), p509.
 It should be noted that
even if media acceptance lies beyond the critical mass, it does
not mean that it will be accepted automatically; it meant only that
resistance due to lack of popular support no longer exist.
 Notice that this theory
could also fit under rational model insofar that the manager is
assumed by Markus (1987) to be rational. Although it is perhaps
more accurately classified as a social construction theory in that
the theory emphasises the influence of social forces.
 This notion is also called
IT sophistication. It was first used by Churchill et al in a study
examining how mangers used computer-based IS. Churchill, Kempster
and Uretsky (1969); See Karimi et al (1996) for a quick summary
of the literature on IT maturity.
 Under Nolan's model, the
IS of a typical company evolves along six stages: initiation, contagion,
control, integration, data administration, and maturity. There is
an alternative called technology assimilation model; see McFarlan
 Indeed this point even acknowledged
in the original media richness model. See Daft and Lengel (1984).
 See Section