Вы находитесь на странице: 1из 20

The Information Society, 20: 325–344, 2004

Copyright 
c Taylor & Francis Inc.
ISSN: 0197-2243 print / 1087-6537 online
DOI: 10.1080/01972240490507974

The Concept of Information Overload: A Review


of Literature from Organization Science, Accounting,
Marketing, MIS, and Related Disciplines

Martin J. Eppler and Jeanne Mengis


Institute of Corporate Communication, University of Lugano, Lugano, Switzerland

help readers (researchers and managers alike) to recognize


Based on literature from the domains of organization science, information overload symptoms, causes, and possible
marketing, accounting, and management information systems, this countermeasures in their own work environment, as the
review article examines the theoretical basis of the information flood of potentially relevant information has become a
overload discourse and presents an overview of the main defini- ubiquitous research and business problem, from reading
tions, situations, causes, effects, and countermeasures. It analyzes relevant articles or reports, to screening e-mails or brows-
the contributions from the last 30 years to consolidate the existing ing the Internet.
research in a conceptual framework and to identify future research While this is not the first review article on the topic
directions. of information overload (see Edmunds & Morris, 2000),
it is the first one to analyze the problem of information
Keywords information explosion, information management strate- overload across various management disciplines, such as
gies, information overload, information processing, organization science, accounting, marketing, and manage-
information skills, information technology ment information systems (MIS). Other review articles on
the subject follow a discipline-based approach. Malhotra
et al. (1982) and more recently Owen (1992) focus on con-
In this article, we present a review of the literature on in- sumer research (see also Meyer, 1998); Schick et al. (1990)
formation overload in management-related academic pub- examine relevant accounting literature; and Edmunds and
lications. The main elements of our approach are literature Morris (2000), Grise and Gallupe (1999/2000), and Nelson
synopsis, analysis, and discussion (Webster & Watson, (2001) concentrate on MIS research. Our review of con-
2002). These three elements serve, in our view, the three tributions in the area of information overload is interdis-
main purposes of a literature review, namely, to provide an ciplinary because it aims to identify similarities and dif-
overview of a discourse domain (e.g., compiling the main ferences among the various management perspectives and
terms, elements, constructs, approaches and authors), to show to what extent they have discussed information over-
analyze and compare the various contributions (as well as load. We hope that by doing so, we can identify syner-
their impact), and to highlight current research deficits and gies between the different streams of information over-
future research directions. These three objectives should load research and highlight future research areas. Another
be met, with regard to the topic of information overload, benefit of an interdisciplinary literature review is that it
as a clear overview, an analysis of the major contribu- can provide a more (cross-)-validated and general col-
tions, and an identification of future research needs still lection of possible symptoms, causes, and countermea-
missing for this topic. The literature review should also sures and thus lead to a more complete understanding of
the phenomenon. This literature-based understanding can
then be used to construct testable models on information
Received 27 October 2002; accepted 20 May 2004.
overload.
We thank the four anonymous reviewers and the two editors for their A second difference of our review in relation to prior
insightful suggestions. contributions is the way that the literature is summarized
Address correspondence to Martin J. Eppler, Institute of Corporate and analyzed, as we present the results of our review
Communication, University of Lugano, Via G. Buffi, 13, 6900 Lugano, in a highly compressed and often visual format. By our
Switzerland. E-mail: Martin.Eppler@lu.unisi.ch providing various diagrammatic overviews of the reviewed

325
326 M. J. EPPLER ET AL.

literature, patterns in the development of the field can be-


come visible. The major benefit of this visual approach
is a more concise representation of the discourse domain,
which allows for easier comparisons and hopefully also
leads to a reduction of information overload for our
readers.

THE CONCEPT OF INFORMATION OVERLOAD


In ordinary language, the term “information overload”
is often used to convey the simple notion of receiving
too much information. Within the research community, FIG. 1. Information overload as the inverted U-curve.
this everyday use of the term has led to various con-
structs, synonyms, and related terms, such as cognitive
overload (Vollmann, 1991), sensory overload (Libowski, This inverted U-curve represents the first important def-
1975), communication overload (Meier, 1963), knowledge inition of information overload, which was strongly de-
overload (Hunt & Newman, 1997), and information fa- bated in the following (see Malhotra et al., 1982; Russo,
tigue syndrome (Wurman, 2001). These constructs have 1974; or McKinnon & Bruns, 1992). For an overview of the
been applied to a variety of situations, ranging from au- various ways researchers have marked the point at which
diting (Simnet, 1996), to strategizing (Sparrow, 1999), information overload occurs, see Table 2.
business consulting (Hansen & Haas, 2001), management Authors in the field of marketing define information
meetings (Grise & Gallupe, 1999/2000), and supermarket overload by comparing the volume of information supply
shopping (Jacoby et al., 1974; Friedmann, 1977), to name (e.g., the number of available brands) with the information-
but a few overload contexts (for an extended list of processing capacity of an individual. Information over-
the contexts in which information overload has been dis- load occurs when the supply exceeds the capacity. Dys-
cussed in management-related academic literature see functional consequences (such as stress or anxiety) and
Table 1). a diminished decision quality are the result. A similar
Research on information overload relevant for the realm way of conceiving the information overload phenomenon
of management has mainly been undertaken in the areas compares the individual’s information-processing capac-
of accounting (e.g., Schick et al., 1990), management in- ity (i.e., the quantity of information one can integrate into
formation systems (MIS) (initially highlighted by Ackoff, the decision-making process within a specific time pe-
1967), organization science (e.g., Galbraith, 1974; riod) with the information-processing requirements (i.e.,
Tushman & Nadler, 1978), and marketing and more spe- the amount of information one has to integrate in order
cially consumer research (e.g., Jacoby, 1984; Keller & to complete a task). This is the “classic” definition of in-
Staelin; 1987; Malhotra, 1984). The main focus of these formation overload, based on the information-processing
disciplines is how the performance (in terms of adequate view of the organization suggested by Galbraith (1974) and
decision making) of an individual varies with the amount expanded by Tushman and Nadler (1978). Following their
of information he or she is exposed to. Researchers across reasoning, information overload can be explained via the
various disciplines have found that the performance (i.e., following formula: information processing requirements
the quality of decisions or reasoning in general) of an indi- > information processing capacities. The terms “require-
vidual correlates positively with the amount of information ments” and “capacities” in this definition can be measured
he or she receives—up to a certain point. If further infor- in terms of available time. The requirements refer to a given
mation is provided beyond this point, the performance of amount of information that has to be processed within a
the individual will rapidly decline (Chewning & Harrell, certain time period. If the capacity of an individual only
1990). The information provided beyond this point will no allows a smaller amount of information to be processed
longer be integrated into the decision-making process and in the available time slot, then information overload is
information overload will be the result (O’Reilly, 1980). the consequence. Tushman and Nadler define informa-
The burden of a heavy information load will confuse the in- tion processing in this context as the “gathering, inter-
dividual, affect his or her ability to set priorities, and make preting, and synthesis of information in the context of or-
prior information harder to recall (Schick et al., 1990). ganizational decision making” (Tushman & Nadler, 1978,
Figure 1 provides a schematic version of this discovery. It p. 614). Many variations of this definition exist. Schick
is generally referred to as the inverted U-curve, following et al. (1990) also stress the time factor as the most
the initial work of Schroder Driver, and Streufert (Schroder important issue regarding the information overload prob-
et al., 1967). lem. Interestingly, this discussion includes the Schroder
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 327

TABLE 1
Information overload situations
Context/overload situation References

Information retrieval, organization, • Searching on the Internet Berghel, 1997


and analysis processes • Screening medical information Bawden, 2001
• Financial distress analysis Chewning and Harrell, 1990
• Evaluating the variety of product Herbig and Kramer, 1994
functions
• Analysis activities (strategic Meyer, 1998
portfolio, environmental, new
product analysis, service decisions)
• Investment analysis Tuttle and Burton, 1999
• Library management Meier, 1963
Decision processes • Managerial decisions in general Ackoff, 1967; Iselin, 1993
• Management (project, strategic, Chervany and Dickson, 1974; Haksever and
production management) Fisher, 1996; Meyer, 1998; Sparrow, 1999
• Supermarkets (choice of product) Friedmann, 1977; Jacoby et al., 1974
• Bankruptcy prediction process Casey, 1980; Iselin, 1993
• Capital budgeting process Swain and Haka, 2000
• Welfare assistance (decisions about O’Reilly, 1980
type and amount)
• Innovation choice Herbig and Kramer, 1994
• Price setting Meyer, 1998
• Advertising media selection Meyer, 1998
• Strategy development Sparrow, 1999
• Physician’s decision making Hunt and Newman, 1997
• Financial decision making Iselin, 1988; Revsine, 1970
• Brand choice (consumer decision Jacoby et al., 1974, 1987; Malhotra, 1982;
making) Owen, 1992; Scammon, 1977; Wilkie, 1974
• Aviation O’Reilly, 1980
Communication processes • Meetings Schick et al., 1990
• Telephone conversations Schick et al., 1990
• The use of groupware applications Schultze and Vandenbosch 1998
• Bulletin board systems (BBS) Hiltz and Turoff, 1985
• Face-to-face discussions Sparrow, 1999
• Telephone-company services Griffeth et al., 1988
• Electronic meetings Grise and Gallupe, 1999, 2000
• Idea organization Grise and Gallupe, 1999, 2000
• E-mail Bawden, 2001; Speier et al., 1999; Denning,
1982
• Management consulting Hansen and Haas, 2001
• City interactions Milgram, 1970
• Disclosure law, contract complexity, Grether et al., 1986
legal disclaimers

et al. (1967) view that information load and processing 1987), not only the amount of information and the avail-
capacity are not independent, since the former can influ- able processing time (i.e., the quantitative dimension),
ence the latter—that is, high information load can increase but also the characteristics of information (i.e., the qual-
one’s processing capacity up to a certain point (see also itative dimension) are seen as major overload elements.
Schultze & Vandenbosch, 1998). In other studies (Iselin, Keller and Staelin refer to the overall quality or “useful-
1993; Keller & Staelin, 1987; Owen, 1992; Schneider, ness of the available . . . information” (1987, p. 202), while
328 M. J. EPPLER ET AL.

TABLE 2
Definitions of information overload
Definitions Components/dimensions References

The decision maker is considered to have • Inverted U-curve: relationship between Chewning and Harrell (1990),
experienced information overload at the point amount of information provided and Cook (1993), Griffeth et al.
where the amount of information actually amount of information integrated by (1988), Schroder et al.
integrated into the decision begins to decline. decision maker (1967),
Beyond this point, the individual’s decisions • Information utilization Swain and Haka (2000)
reflect a lesser utilization of the available
information.
Information overload occurs when the volume of • Volume of information supply Jacoby et al. (1974),
the information supply exceeds the limited (information items versus - chunks) Malhotra (1982),
human information processing capacity. • Information processing capacity Meyer (1998)
Dysfunctional effects such as stress and • Dysfunctional consequences
confusion are the result.
Information overload occurs when the • Information-processing capacity Galbraith (1974),
information-processing requirements • Information-processing requirements Tushman and Nadler (1978)
(information needed to complete a task) exceed
the information-processing capacity (the
quantity of information one can integrate into the
decision-making process).
Information overload occurs when the • Time demands of information processing; Schick, et al. (1990),
information-processing demands on time to available time versus invested time Tuttle and Burton (1999)
perform interactions and internal calculations • Number of interactions (with
exceed the supply or capacity of time available subordinates, colleagues, superiors)
for such processing. • Internal calculations (i.e., thinking time)
Information overload occurs when the • Information-processing requirements Keller and Staelin (1987),
information-processing requirements exceed the • Information-processing capacity Schneider (1987),
information-processing capacity. Not only is the • Quantitative and qualitative dimensions of Owen (1992), Iselin (1993)
amount of information (quantitative aspect) that information (multidimensional approach)
has to be integrated crucial but also the
characteristics (qualitative aspect) of
information.
Information overload occurs when the decision • Subjective component: opinion, job and Abdel-Khalik (1973), Iselin
maker estimates he or she has to handle more communication satisfaction (1993), O’Reilly (1980),
information than he or she can efficiently use. • Situational factors and personal factors Haksever and Fisher (1996)
Amount of reading matter ingested exceeds • Subjective cause component: energy Wurman (1990), Wurman
amount of energy available for digestion; the • Symptom: stress, overstimulation (2001), Shenk (1997)
surplus accumulates and is converted by stress • Subjective effect: information overload
and overstimulation into the unhealthy state anxiety
known as information overload anxiety.

Schneider (1987) distinguishes various information attri- proach are O’Reilly (1980), Haksever and Fisher (1996),
butes, such as the level of novelty, ambiguity, uncertainty, and Lesca and Lesca (1995). In this “subjective” view
intensity, or complexity. These information characteristics of overload, the feelings of stress, confusion, pressure,
or quality attributes can either contribute to overload or anxiety, and low motivation are the crucial factors that
reduce it. signal the occurrence of information overload. Empiri-
Beyond these approaches that try to conceptualize and cal research that follows this subjective view of the over-
measure the phenomenon of overload objectively, there load phenomenon typically employs interviews or survey
are others that conceive overload on the basis of sub- methods (such as Haksever & Fisher, 1996) as opposed to
jective experience. Authors who have followed this ap- experiments.
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 329

This brief overview of the most frequently used defini- ber of 97 considered articles. The 71 articles that were
tions and the contexts within which they were developed eliminated discussed information overload in very specific
delineates the intellectual territory which is examined in contexts that are quite different from those of today’s orga-
this literature review. Having described the background of nizations. They discussed information overload in contexts
the overload concept, we next briefly outline the method- such as library and bibliographic research, documentation
ology used to analyze the relevant literature. of large-scale engineering designs, and students conduct-
ing research for their essays). Figure 2 reveals that the
METHODOLOGY greatest number of retained articles comes from the mar-
keting domain, followed by articles within organization
To screen the relevant articles within the literature on infor- science, then accounting, and finally MIS had the smallest
mation overload, we used the electronic database provided number of articles on the subject.
by EBSCOhost and limited our research to the articles One limitation of this methodology is that some arti-
included by the Business Premier Source. This database cles that have dealt with the issue, but have used labels
provides full-text access to 3000 journals, of which more other than the four terms we used as keywords, are not
than 1000 are peer reviewed. EBSCOhost enabled us to taken into consideration (i.e., labels such as data smog,
search in the title or abstract of an article with the fol- information fatigue/overkill/overabundance/breakdown/
lowing keywords: information overload, information load, explosion/deluge/flood/stress/plethora, document tsun-
cognitive overload, and cognitive load, which resulted in ami, sensory overload, etc.; see Eppler, 1998, for these
a total number of 548 retrieved articles. In order to reduce and other labels). These different terms, however, have
this large number to a more relevant subset, we introduced not achieved wide acceptance within the scientific com-
further criteria (see Figure 2), which were: first, a publi- munity and hence do not represent core contributions to
cation date after 1970 (when computers started to be used this scientific debate. Another limitation of our approach
more extensively in the workplace); second, that the ar- is that contributions on information overload that discuss
ticle is peer reviewed (which resulted in 205 remaining the phenomenon from other perspectives (such as psy-
articles); third, that information overload is a dominant chology, health care, and mass communication) are not
and systematically addressed subject in the article and not extensively addressed. Examples of such important contri-
just mentioned once or twice (resulting in a total of 168 butions include Miller’s “The magical number seven plus
articles); and finally, that the article approaches the sub- or minus two” (Miller, 1956) and Simon’s seminal “Infor-
ject within the context of one of the four areas of interest, mation processing models of cognition” (Simon, 1979), to
namely, accounting, marketing, MIS, and organization sci- name but two crucial contributions. In order to moderate
ence (with regard to the article’s topics and its publication this limitation, we used an additional inclusion criterion:
journal). This selection procedure has led to a total num- If a publication was cited in more than two other overload
articles, we screened it to see whether information over-
load was indeed a major topic of the publication, and if
that was the case, we included it in our analysis. We pro-
ceeded likewise for books that have been cited frequently
in relevant journal articles (such as Wurman, 1990, 2001;
Shenk, 1997).

A CONCEPTUAL FRAMEWORK FOR


INFORMATION OVERLOAD RESEARCH
In order to provide a more complete (and less fragmented)
picture of the research conducted on information over-
load, the following framework lays out the most impor-
tant topic clusters in the literature and underscores their
relationships. These topic clusters are the main causes
of information overload, the symptoms, and appropriate
countermeasures for mitigating the problem.
It is important to note that the framework depicted
in Figure 3 is not based on a linear logic of causes and ef-
fects, but instead emphasizes a system of circular,
interdependent relationships. Correspondingly, it stresses
FIG. 2. Selection criteria and article base. the fact that any countermeasure that is aimed at a specific
330 M. J. EPPLER ET AL.

FIG. 3. A conceptual framework to structure research on information overload.

overload cause can have significant side effects on other cessing, or communicating information, the tasks or pro-
causes. Although this fact is frequently acknowledged in cesses that need to be completed by a person, team, or
literature (e.g., Bawden, 2001), it has scarcely been ex- organization, the organizational design (i.e., the formal
plored empirically (for an exception, see Evaristo, 1993). and informal work structures), and the information tech-
In particular, the effect of certain (new) information tech- nology that is used (and how it is used) in a company.
nology applications on the quality of information (see Usually information overload emerges not because of one
Wang et al., 1998), on the motivation of the individual, of these factors but because of a mix of all five causes.
and on task parameters has been neglected. Also, a con- All five causes influence the two fundamental variables of
certed effort needs to be made to employ research meth- information overload: the information processing capac-
ods that can capture contextual factors (such as industry ity (IPC)—which is for example influenced by personal
characteristics, the firm’s development stage, and the staff characteristics—and the information processing require-
structure) that are of critical importance for the occurrence ments (IPR)—which are often determined by the nature of
of overload. In general, research that provides “deep con- the task or process. We discuss these five causes and their
text” is missing, as most information overload research is influence on IPC and IPR briefly in the next paragraphs.
experimental, survey based, or purely conceptual. An important factor influencing the occurrence of
This framework also highlights the fact that there can- information overload is the organizational design of
not be a definitive solution for information overload. There a company (Galbraith, 1974; Tushman & Nadler, 1978).
will always be a need for a continuous cycle of improve- Changes in the organizational design, for instance, due to
ment and refinement. We discuss the main elements (the disintermediation or centralization (Schneider, 1987) or
causes, symptoms and countermeasures) of the framework because of a move to interdisciplinary teams (Bawden,
and the relevant literature in the subsequent sections. At the 2001), can lead to greater IPRs because they create the
end of this section, we also demonstrate how this concep- need for more intensive communication and coordination.
tual framework can be converted into empirically testable On the other hand, better coordination through standards,
models. common procedures, rules, or dedicated coordination cen-
ters (Galbraith, 1974) can reduce the IPR and positively
influence the IPC (Galbraith, 1974; Schick et al., 1990;
Causes of Information Overload
Tushman & Nadler, 1978; for other organizational
The main reasons for information overload at organiza- design elements that influence information overload see
tional and interpersonal levels can be related to five con- Schneider, 1987).
structs, as shown in Figure 3. These inductively generated After organizational design, the next important factor is
constructs are the information itself (its quantity, frequ- the nature of information itself. Schneider (1987) stresses
ency, intensity, and quality), the person receiving, pro- the fact that it is not only the amount of information that
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 331

determines information overload, but also the specific impact on an individual’s IPC and IPR. Pushing selected
characteristics of information (also see Sparrow, 1998). pieces of information to specific groups reduces on the
Such characteristics are the level of uncertainty associ- one hand their information retrieval time, but increases
ated with information and the level of ambiguity, novelty, on the other the amount of potentially useless information
complexity, and intensity (Schneider, 1987). Simpson and that a person has to deal with (Edmunds & Morris, 2000).
Prusak (1995) argue that modifying the quality of infor- In addition, it causes more frequent interruptions (Speier
mation can have great effects on the likelihood of infor- et al., 1999). Information technology can thus potentially
mation overload. Improving the quality (e.g., conciseness, increase the individual’s IPC while at the same time in-
consistency, comprehensibility, etc.) of information can creasing the IPR.
improve the information-processing capacity of the indi- A complete list of the specific overload causes that are
vidual, as he or she is able to use high-quality informa- mentioned in the relevant literature can be found in Table 3.
tion more quickly and better than ill-structured, unclear It is structured according to the five categories discussed
information. earlier.
The person and his or her attitude, qualification, and ex- Having reviewed the major causes of information over-
perience are another important factor. While earlier studies load and their impact on IPC and IPR, we can now look at
simply state that a person’s capacity to process information their effects or observable symptoms.
is limited (Jacoby et al., 1974; Galbraith, 1974; Malhotra,
1982; Simon, 1979; Tushman & Nadler, 1978), more re-
Symptoms of Information Overload
cent studies include specific limiting factors such as per-
sonal skills (Owen, 1992), the level of experience (Swain One of the first researchers to examine the effects of over-
& Haka, 2000), and the motivation of a person (Muller, load was the American psychologist Stanley Milgram
1984). Personal traits thus directly affect IPC. (1970), who analyzed signal overload for people living
Another important factor is the tasks and processes that in large cities. In his study, he identified six common reac-
need to be completed with the help of information. The tions to the constant exposure to heavy information load,
less a process is based on reoccurring routines (Tushman which are allocation of less time to each input, disregard
& Nadler, 1975) and the more complex it is in terms of the of low-priority inputs, redrawing of boundaries in some
configuration of its steps (Bawden, 2001; Grise & Gallupe, social transactions to shift the burden of overload to the
1999, 2000), the higher is the information load and the other party of the exchange, reduction of inputs by filtering
greater is the time pressure on the individual (Schick et al., devices, refusal of communication reception (via unlisted
1990). The combination of these two factors that increase telephone numbers, unfriendly facial expressions, etc.),
the IPR can lead to information overload. Information and finally creation of specialized institutions to absorb
overload is especially likely if the process is frequently in- inputs that would otherwise swamp the individual (see
terrupted and the concentration of the individual suffers as also Weick, 1970, for this point).
a consequence (Speier et al., 1999). Information overload In the organizational context, frequently described
is also more likely if managers face an ever greater num- symptoms of information overload on the individual level
ber of parallel projects or tasks that they have to manage are a general lack of perspective (Schick et al., 1990),
(see Wurman, 2001). In this way, complex tasks or pro- cognitive strain and stress (Malhotra, 1982; Schick et al.,
cesses directly increase the IPR. This fact is aggravated by 1990), a greater tolerance of error (Sparrow, 1999), lower
a reduced IPC as a result of frequent context switching or job satisfaction (Jacoby, 1984), and the inability to use
distraction. information to make a decision (Bawden, 2001)—the so-
Finally, information technology and its use and misuse called paralysis by analysis. Many other symptoms noted
are a major reason why information overload has become by different researchers are listed in Table 4.
a critical issue in many organizations in the 1980s and The big question with regard to effects of information
1990s. The development and deployment of new infor- overload is whether and how it impacts decision accuracy,
mation and communication technologies, such as the In- decision time, and general performance. While research re-
ternet, intranets, and extranets, but especially e-mail, are sults have often been contradictory, especially among the
universally seen as one major cause of information over- groundbreaking studies in marketing (the inconsistencies
load (Bawden, 2001). There are, however, also arguments were in part due to methodological problems; see Jacoby
in favor of e-mail. Edmunds and Morris (2000), for ex- et al., 1974; Malhotra et al., 1982; Muller, 1984), there is
ample, stress advantages like the fact that e-mail is an wide consensus today that heavy information load can af-
asynchronous form of communication and is less likely to fect the performance of an individual negatively (whether
interrupt the normal work flow. Closely related to the prob- measured in terms of accuracy or speed). When infor-
lem of e-mail overload is the discussion of pull versus push mation supply exceeds the information-processing capac-
technologies and whether they have a positive or negative ity, a person has difficulties in identifying the relevant
332 M. J. EPPLER ET AL.

TABLE 3
Causes of information overload
Causes of information overload References

Personal factors • Limitations in the individual human Herbig and Kramer, 1994
information-processing capacity
• Decision scope and resulting documentation needs Kock, 2001
• Motivation, attitude, satisfaction Muller, 1984
• Personal traits (experience, skills, ideology, age) Owen, 1992; Hiltz and Turoff, 1985;
Muller, 1984; Schneider, 1987; Swain
and Haka, 2000
• Personal situation (time of the day, noise, temperature, Owen, 1992; O’Reilly, 1980
amount of sleep)
• Senders screen outgoing information insufficiently Van Zandt, 2001
• Users of computers adapt their way of interacting with Maes, 1994
computers too slowly with respect to the technological
development
• Social communication barriers break down Schultze and Vandenbosch, 1998
Information characteristics • Number of items of information rises Bawden, 2001; Herbig and Kramer, 1994;
Jacoby et al., 1974; Jacoby 1977, 1984;
Malhotra, 1982
• Uncertainty of information (info needed vs. info Schneider, 1987; Tushman and Nadler,
available) 1978
• Diversity of information and number of alternatives Bawden, 2001; Iselin, 1988; Schroder
increase et al., 1967
• Ambiguity of information Schneider, 1987; Sparrow, 1999
• Novelty of information Schneider, 1987
• Complexity of information Schneider, 1987
• Intensity of information Schneider, 1987
• Dimensions of information increase Schroder et al., 1967
• Information quality, value, half-life Sparrow, 1998, 1999
• Overabundance of irrelevant information Ackoff, 1967
Task and process parameters • Tasks are less routine Tushman and Nadler, 1975
• Complexity of tasks and task interdependencies Tushman and Nadler, 1975
• Time pressure Schick et al., 1990
• Task interruptions for complex tasks Speier et al., 1999
• Too many, too detailed standards (in accounting) Schick et al., 1990
• Simultaneous input of information into the process Grise and Gallupe, 1999, 2000
• Innovations evolve rapidly—shortened life cycle Herbig and Kramer, 1994
• Interdisciplinary work Bawden, 2001
Organizational design • Collaborative work Wilson, 1996
• Centralization (bottlenecks) or disintermediation Schneider, 1987
(information searching is done by end users rather than
by information professionals)
• Accumulation of information to demonstrate power Edmunds and Morris, 2000
• Group heterogeneity Grise and Gallupe, 1999
• New information and communication technologies (e.g., Bawden, 2001; Schultze and Vandenbosch,
groupware) 1998; Speier et al., 1999
Information technology • Push systems Bawden, 2001
• E-mails Bawden, 2001
• Intranet, extranet, Internet Bawden, 2001
• Rise in number of television channels Edmunds and Morris, 2000
• Various distribution channels for the same content Edmunds and Morris, 2000
• Vast storage capacity of the systems Schultze and Vandenbosch, 1998
• Low duplication costs Schultze and Vandenbosch, 1998
• Speed of access Schultze and Vandenbosch, 1998
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 333

TABLE 4
Symptoms or effects of information overload

Symptoms References

Limited information search and • Search strategies through information sets Swain and Haka, 2000
retrieval strategies become less systematic (this is less true for more
experienced searchers)
• Limited search directions Cook, 1993
• Move from compensatory search patterns to Cook, 1993
noncompensatory search patterns
• Identification and selection of relevant Jacoby, 1977; Schneider, 1987
information becomes increasingly difficult
• Difficulties to reach target groups (sender Herbig and Kramer, 1994
perspective)
Arbitrary information analysis • Overlapping and inconsistent information Eppler, 1998
and organization categories
• Ignore information and be highly selective Bawden, 2001; Edmunds and Morris,
(omission) 2000; Herbig and Kramer, 1994; Hiltz
and Turoff, 1985; Sparrow, 1999
• Loss of control over information Bawden, 2001; Wurman, 1990
• Lack of critical evaluation (become too Shenk, 1997; Schick et al., 1990;
credulous) and superficial analysis Schultze and Vandenbosch, 1998
• Loss of differentiation Schneider, 1987
• Relationship between details and overall Owen, 1992; Schneider, 1987
perspective is weakened and peripherical cues
get overestimated
• Higher time requirements for information Jacoby, 1984; Hiltz and Turoff, 1985
handling and time delays
• Abstraction and necessity to give meaning lead Sparrow, 1999; Walsh, 1995
to misinterpretation
Suboptimal decisions • Decision accuracy/quality lowered Malhotra, 1982; Jacoby, 1984; Hwang
and Lin, 1999
• Decision effectiveness lowered Schroder et al., 1967
• Inefficient work Bawden, 2001
• Potential paralysis and delay of decisions Bawden, 2001; Schick et al., 1990
Strenuous personal situation • Demotivation Baldacchino et al., 2002
• Satisfaction negatively affected Jacoby, 1984; Jones, 1997
• Stress, confusion, and cognitive strain Jones, 1997; Malhotra, 1982; Schick
et al., 1990
• Lacks to learn since too little time is at Sparrow, 1999
disposition
• Greater tolerance of error Sparrow, 1999
• Lack of perspective Schick et al., 1990
• Sense of loss of control leads to a breakdown in Schneider, 1987
communication
• False sense of security due to uncertainty Meyer, 1998; Jacoby, 1984; O’Reilly,
reduction (overconfidence) 1980

information (Jacoby, 1977), becomes highly selective and perspective (Schneider, 1987), needs more time to reach
ignores a large amount of information (Bawden, 2001; a decision (Jacoby, 1984), and finally does not reach a
Herbig & Kramer, 1994; Sparrow, 1999), has difficulties in decision of adequate accuracy (Malhotra, 1982). Because
identifying the relationship between details and the overall of these many potential negative effects, it is important to
334 M. J. EPPLER ET AL.

devise effective countermeasures. They should address not 1999/2000). Examples of such intelligent systems are de-
only the symptoms of information overload but also its cision support systems (DSS) that reduce a large set of
causes. In the next subsection we provide an overview of options to a manageable size (Cook, 1993).
such mechanisms. In the survey just described we can see that many au-
thors list a multitude of possible countermeasures, but that
they do not provide specific suggestions on how to com-
Countermeasures Against Information Overload
bine organizational, technological, personal, and infor-
Literature on information overload not only discusses ma- mation- and task-based improvement actions. Clearly, a
jor causes and effects, but also proposes possible effec- systematic methodology (comparable to other standard-
tive countermeasures to address the issues related to in- ized problem solving approaches) to prevent or reduce
formation overload. These countermeasures range from information overload is still missing. Such a methodol-
general suggestions concerning attitude to very specific ogy should combine insights from various disciplines to
software tools (such as filtering agents, automatic sum- provide effective countermeasures that can be adapted to
marizers, or visualization algorithms) that help to process various contexts. For example, insights from consumer
large amounts of information. A list of countermeasures research on the importance of branding for reducing in-
mentioned in the literature can be found in Table 5, which formation overload can be used for new MIS instruments
uses the same organizational schema that was used to clas- (Berghel, 1997; Jacoby et al., 1974). Such prescriptive
sify the causes, so that the two (causes and countermea- suggestions must be based on rigorous empirical research.
sures) can be directly related to one another (keeping in The next subsection outlines how our framework can be
mind possible side effects). employed to conduct such empirical research.
With regard to information itself, information overload
can be reduced if efforts are made to assure that it is of
Testable Models Derived from the Framework
high value, that it is delivered in the most convenient way
and format (Simpson & Prusak, 1995), that it is visual- The framework just presented serves primarily as an ori-
ized, compressed, and aggregated (Ackoff, 1967; Meyer, enting map for understanding and examining the overall
1998), and that signals and testimonials are used to min- scope of information load research across different dis-
imize the risks associated with information (Herbig & ciplines. However, it can also serve as a basis for future
Kramer, 1994). On the individual level, it is important to empirical research. For instance, three testable models can
provide training programs to augment the information lit- be derived from the framework.
eracy of information consumers (Bawden, 2001; Koniger The first testable model operationalizes the five cause
& Janowitz, 1995; Schick et al., 1990) and to give em- categories as independent variables that lead to (or pre-
ployees the right tools so that they can improve their time dict) information overload (the dependent variable). Each
(Bawden, 2001) and information management (Edmunds cause group consists of the individual items described in
& Morris, 2000) skills. As far as improvements for the the causes summary table (see Table 3). The data are col-
organizational design are concerned, various authors take lected via questions asked in a Likert-scale manner. In this
on conflicting positions. While earlier contributions stress way, the correlation between causes and the occurrence of
the importance of self-contained tasks and lateral relation- information overload (measured as the subjective feeling
ships (Galbraith, 1974), more recent studies see this fo- of not being able to process all relevant information in the
cus on collaborative and interdisciplinary work as a cause available time) can be measured. In addition, a question-
rather than as a countermeasure of information overload naire based on this model can be used to test whether we
(Bawden, 2001; Wilson, 1996). If the cause of informa- have allocated the individual causes to the right cause cat-
tion overload relates to process problems, several authors egory (based on goodness of fit or Cronbach alpha values).
suggest standardization of operating procedures (Bawden, The second testable model relates to the possible symp-
2001; Schick et al., 1990; Schneider, 1987), collabora- toms of information overload. The symptoms that are listed
tion with information specialists within the process teams in Table 4 can be converted into questions. Based on ques-
(Edmunds & Morris, 2000), and use of facilitators or col- tionnaire results, one can then build groups of symptoms
laborative tools (such as virtual team rooms) as “pro- through factor analysis and correlate these groups (and the
cess enablers” for cognitive support (Grise & Gallupe, individual symptoms) with the question regarding over-
1999/2000). Finally, at the level of information technol- load (e.g. “Do you feel that you suffer from information
ogy, several authors advocate the use of intelligent infor- overload?”). This can help us understand which symp-
mation management systems for fostering an easier pri- toms may be most representative of the overload phe-
oritization of information (Bawden, 2001; Meyer, 1998; nomenon and thereby validate our symptom categories. In
Schick et al., 1990) and providing quality filters (Ack- this model, the independent variables would be the identi-
off, 1967; Edmunds & Morris, 2000; Grise & Gallupe, fied symptoms, whereas the dependent variable would be
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 335

TABLE 5
Countermeasures against information overload

Countermeasures References

Personal factors • Improve personal time management skills and techniques Bawden, 2001
• Training programs to augment information literacy: Bawden, 2001; Jones, 1997;
information-processing skills such as file handling, using Schick et al., 1990; Koniger
e-mail, classification of documents, etc. and Janowitz, 1995
• Improve personal information management Edmunds and Morris, 2000
• Systematic priority setting Schick et al., 1990
• Improve the screening skills for information Van Zandt, 2001
Information characteristics • Raise general quality of information (i.e., its usefulness, Allert, 2001; Keller and Staelin,
conciseness) by defining quality standards 1987; Meglio and Kleiner,
1990; Simpson and Prusak,
1995
• Focus on creating value-added information Simpson and Prusak, 1995
• Promulgation of rules for information and communication Bawden, 2001
design (e.g., e-mail etiquette)
• Compress, aggregate, categorize, and structure information Ackoff, 1967; Grise and Gallupe,
1999/2000; Hiltz and Turoff,
1985; Iselin, 1988; Koniger and
Janowitz, 1995; Scammon,
1977
• Visualization, the use of graphs Chan, 2001; Meyer, 1998
• Formalization of language Galbraith, 1974
• Brand names for information Berghel, 1997
• Form must follow function must follow usability Herbig and Kramer, 1994
• Simplify functionalities and design of products Herbig and Kramer, 1994
• Customization of information Ansari and Mela, 2003; Berghel,
1997; Meglio and Kleiner, 1990
• Intelligent interfaces Bawden, 2001
• Determine various versions of an information with various Denning, 1982
levels of detail and elaborate additional information that
serves as summaries
• Organize text with hypertext structures or gophers Nelson, 2001
• Interlink various information types (as internal with Denton, 2001; Meglio and
external information) Kleiner, 1990
Task and process parameters • Standardize operating procedures Bawden, 2001; Schneider, 1987;
Schick et al., 1990
• Define decision models developed for specific decision Ackoff, 1967; Chewning and
processes (e.g., decision rules) Harrell, 1990
• Install an exception-reporting system Ackoff, 1967
• Allow more time for task performance Schick et al., 1990
• Schedule uninterrupted blocks of time for completing Sorohan, 1994
critical work
• Adequate selection of media for the task Schick et al., 1990
• Handle incoming information at once Sorohan, 1994
• Collaboration with information specialists within the teams Edmunds and Morris, 2000
• Bring decisions to where information exists when this Galbraith, 1974
information is qualitative and ambiguous
• Install process enablers for cognitive support Grise and Gallupe, 1999/2000
• Use simpler information-processing strategies Schick et al., 1990
(Continued on next page)
336 M. J. EPPLER ET AL.

TABLE 5
Countermeasures against information overload (Continued)

Countermeasures Reference

• Regulate the rate of information flow Grise and Gallupe, 1999, 2000
• Search procedures and strategy Ackoff, 1967; Bawden, 2001; Meyer,
1998; Olsen et al., 1998; Revsine,
1970
• Define specific, clear goals for the information in order to Baldacchino et al., 2002; Denton,
contextualize it and turn it meaningful 2001; Meglio and Kleiner 1990
• Communicate information needs to providers Meglio and Kleiner, 1990
• Provide incentives that are directly related with decisions Tuttle and Burton, 1999
in order to make decision relevant information be
processed more effectively
• Install a measurement system for information quality Denton, 2001
Organizational design • Coordination through interlinked units Tushman and Nadler, 1978
• Augment info processing capacity through chances in org. Galbraith, 1974; Schick et al., 1990;
design Tushman and Nadler, 1978
• Creation of lateral relationships (integrate roles, create Galbraith, 1974
liaisons between roles, teamwork etc.)
• Coordination by goal setting, hierarchy, and rules Galbraith, 1974
depending on frequency of exceptions (uncertainty)
• Creation of self-contained tasks (reduced division of Galbraith, 1974
labor, authority structures based on output categories) →
autonomous groups
• Reduce divergence among people (e.g., with regard to Schneider, 1987
expectations) trough socialization (e.g., frequent
face-to-face interactions)
• Install appropriate measures of performance Ackoff, 1967
• Hire additional employees Schick et al., 1990
• Create slack resources Galbraith, 1974
Information technology • Intelligent information management (prioritization) Bawden, 2001; Meyer, 1998; Schick
application et al., 1990
• Install voting structures to make users evaluate the Denning, 1982; Hiltz and Turoff, 1985
information
• Prefer push to pull technologies Edmunds and Morris, 2000; Denning,
1982; Friedmann, 1977; Herbig and
Kramer, 1994
• Facilitator support through (e-)tools Grise and Gallupe, 1999, 2000
• Decision support systems should reduce a large set of Cook, 1993
alternatives to a manageable size
• Use natural language processing systems (search with Nelson, 2001
artificial intelligence)
• Information quality filters Ackoff, 1967; Bawden, 2001;
Denning, 1982; Edmunds and
Morris, 2000; Grise and Gallupe,
1999, 2000; Hiltz and Turoff, 1985;
Jones, 1997
• Intelligent data selectors (intelligent agents) Berghel, 1997; Edmunds and Morris,
2000; Maes, 1994
• Use systems that offer various information organization Hiltz and Turoff, 1985; Sorohan, 1994
options (e.g. filing systems)
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 337

the occurrence of information overload (in the opinion of gain a deeper understanding of the information overload
the respondents). phenomenon.
The third testable model addresses possible counter-
measures against information overload. It uses the five The Publication Timeline: Overload Research
(cause) categories to ask respondents about countermea- Patterns by Discipline
sures that may or may not be in place in their organization
(and that may or may not help fight overload). Based on the The timeline diagram does not focus on particular con-
survey results, the effectiveness of these countermeasures structs, but on the authors and their impact. It is a good
(as well as their grouping) can be evaluated. The inde- visualization tool if the historic or process perspective of a
pendent variables are the already implemented counter- discourse is analyzed. We have drawn a timeline diagram
measures in a company, whereas the dependent variable is for each one of the four areas in which information over-
the occurrence of information overload for the questioned load research has primarily been conducted over the last 30
individuals. years, namely, accounting, marketing, organizational be-
The main challenge in developing these three models havior, and management information systems (MIS). The
is adequately converting the factors we have found in the objective is not to map out all the articles per field and
literature to scaleable questions that can be answered ac- show all the references to other overload articles, since the
curately (and honestly) by the respondents. resulting diagrams would get too crowded and loose clarity
The framework presented so far gives a systematic over- and insight. The goal instead is to foreground major con-
view on the major findings of scientific research on infor- tributions. To determine the “relevant” contributions, we
mation overload. The discussion on how our framework have limited ourselves to articles that were cited repeatedly
can be tested with the help of three individual models indi- by other articles. In the following subsections, we look at
cates how future studies in the field can proceed. In order to each domain timeline in detail and provide suggestions for
generate further suggestions on the future of information future research.
overload research, we next go beyond the mere descrip-
tion of the field and analyze its inherent discourse patterns. Marketing
This will enable us to see other development needs and ne-
glected areas. Information overload within marketing, or more specifi-
cally within consumer research, has become a critical issue
since the explosion in the number of brands in the early
Biases and Tendencies in the Literature
seventies. Figure 4 reveals that only a few studies have
In order to characterize the four literature domains, we been done on a conceptual level and almost all the over-
employ two visualization formats: the publication or ci- load research in marketing is of empirical nature. This may
tation timeline for the analysis of the impact of relevant lead to slower, but more rigorous, theoretical progress. For
authors in various management domains and the nature their theoretical base, the marketing researchers rely on
of their contribution, and the discipline Venn diagram for the findings of psychologists and cognitive scientists, in
the analysis of interdisciplinary research on the topic to particular on Miller’s study on our limited capacity for

FIG. 4. Timeline of publications and citations of information overload studies in the area of marketing.
338 M. J. EPPLER ET AL.

information processing (Miller, 1956). Moreover, the em- eral trend. Both articles include extensive literature re-
pirical research is, with the only exception of Muller views and contain important insights from organization
(1984), exclusively based on experiments and neglects sur- researchers as well as from MIS scholars. Similar to the
veys or case studies. case of marketing, the empirical research in accounting is
The tangled structure of the references reflects the in- based on experimental situations and not on field research
tense debate that occurred around the Jacoby et al. first in organizations. Additionally, Figure 5 shows that Casey,
study. The methodology employed by them was contested Iselin, and Abdel-Khalik are authors with a high impact
by Wilkie (1974), Scammon (1977), Malhotra (1984), and on the studies of information overload in accounting. As a
others. Jacoby and Malhotra emerge to be the gurus within tendency, the researchers who conduct empirical research
the field. The most intense period of research on informa- often refer to conceptual studies, but the latter rarely refer
tion overload was from the mid-1970s to the mid-1980s. to the former. This, however, would be crucial for generat-
The main question that preoccupied the researchers was ing consistent theoretical progress in the study of overload
whether the number of brands and their attributes (infor- in accounting. The main theme of accounting studies is the
mation load) influence product choice of consumers. Gen- impact of information load on decision quality or accuracy
erally, the research in the field of marketing focuses on for matters regarding budgeting decisions (as in Swain &
the impact of information overload on decision quality, on Haka, 2000) or predictions of bankruptcy (as in Casey,
decision time, and on the actual number of information 1980).
items that can be processed in a typical purchase situation.
As a result of this focus, the information overload litera-
ture in the marketing domain neglects vital issues such as Organization Science
skills, timing, and technological and organizational issues. What is striking in the area of organization science is
Nevertheless, the marketing discipline has legitimized the that almost all the contributions on information overload
research on information overload by experimentally doc- are conceptual articles. The few empirical papers include
umenting the inverted U-curve effect. O’Reilly (1980) and Griffeth, Carson, and Marin (1988).
These two studies work with a subjective definition of in-
formation overload and focus on the satisfaction of the per-
Accounting
son experiencing information overload. The measurement
The timeline of the contributions from the field of ac- tools they employ are questionnaires and not experiments.
counting (Figure 5) presents a similar picture as the one Figure 6 depicts a rather loosely connected structure
of marketing, insofar as the conducted research is almost of citations, in which Galbraith and Tushman and Nadler
exclusively empirical. Again, the theoretical basis is bor- have prominent positions. The main reason for this loose-
rowed from psychologists and cognitive scientists such ness of structure is that the authors refer to organization
as Schroder et al. (1967), Miller (1956), and Simon and scientists who made important contributions for general
Newell (1971). Apart from these fundamental insights organizational issues, but not specifically in information
from psychology, the research is not particularly inter- overload-related topics. These contributions are therefore
disciplinary. Schick et al. (1990) and to a smaller extent not visible in the diagram. The most intense research ac-
also Tuttle and Burton (1999) are exceptions to this gen- tivity took place in the 1990s. Possible reasons for this

FIG. 5. Timeline of publications and citations of information overload studies in the area of accounting.
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 339

FIG. 6. Timeline of publications and citations of information overload studies in the area of organization science.

heightened interest in the 1990s include the rapid propa- information overload as a starting point for their technol-
gation of the Internet and other information technologies, ogy application discussions. Information overload per se
as well as the trend toward collaborative work and flat is mostly not systematically defined, discussed, or ana-
hierarchies (Meyer, 1998). The organization science re- lyzed, but seen as a given problem that has to be resolved.
searchers are mainly interested in showing whether and Consequently, the net number of articles dealing primarily
how the information-processing capacity of an individual with information overload in the MIS field is remarkably
can be expanded through changes in the organizational de- low when compared to the total number of MIS papers that
sign and how this design influences the information pro- address the phenomenon in their title or abstract.
cessing requirements. Again, what has been said for the The major publication activity occurred in the 1990s
other research areas is also true for organization science, (except for Ackoff, 1967; Denning, 1982; and Hiltz &
namely, that the research in this domain is not highly in- Turoff, 1985) (Figure 7). In spite of this fact—and with the
terdisciplinary. This is surprising for a field that typically exception of Schultze and Vandenbosch’s article (1998)
incorporates many concepts from related social sciences. that combines insights from accounting, marketing, and
organization science—the MIS researchers do not seem to
profit enough from already existing information overload
Management Information Systems
studies outside of their field. As mentioned earlier, the
Surprisingly, MIS has not been the discipline that has dealt focus of MIS researchers has been to propose effective
with information overload in the most extensive manner. countermeasures, and not to study the root causes of the
Authors in the field of MIS mostly use the concept of problem or its contextual factors.

FIG. 7. Timeline of publications and citations of information overload studies in the area of MIS.
340 M. J. EPPLER ET AL.

Consequently, the MIS research is concentrated on con- consistent transfer from empirical to conceptual research
ceptual studies and there is an obvious missing link and vice versa. This, however, is a crucial prerequisite
between conceptual and empirical studies; the two app- for cumulative research. Another prerequisite for cumula-
roaches do not often refer to one other. One very valu- tive research is the transfer of research findings between
able exception is again the contribution of Schultze and closely related disciplines. This important issue is further
Vandenbosch (1998), which combines both a literature re- explored in the next section.
view and a survey. The article refers to conceptual papers
as well as to empirical findings from other areas outside the
The Status of Interdisciplinary Information
MIS domain. Aside from this exception, MIS researchers
Overload Research
tend to be mainly interested in finding technical solutions
for the information overload problem. Their contributions The Venn diagram depicted in Figure 8 maps the cross-
are thus interesting with regard to technology-based coun- citations between major overload articles. The inclusion
termeasures against information overload. criteria are the same ones as for the publication timelines.
From the analysis of the different time lines several con- It facilitates an examination of the interdisciplinary status
clusions can be drawn. One is that the transfer between em- of information overload research.
pirical and conceptual studies can be improved and should In general, only a few authors integrate various man-
be intensified in future research. agement perspectives to study the problem of informa-
Most of the empirical research that has been conducted tion overload. In fact, there are no intersections between
within the aforementioned disciplines is done in experi- the area of accounting and either marketing or MIS to
mental settings and hence does not rely on authentic man- our knowledge. Similarly, there is no intersection between
agement contexts. Interestingly, some research areas focus marketing and MIS. Most of the intersections (in terms
more on empirical studies and lack conceptual research, of citing and using relevant work from other domains)
which is true for accounting and marketing, while the ar- are visible within the area of organization science. Some
eas of organization science and MIS are more interested authors of the other three fields integrate findings from
in conceptual approaches. But all the four areas, except organization science. The diagram does not lay out the
to some extent the area of accounting, do not achieve a entire scope of interdisciplinary research, because it does

FIG. 8. Cross-referencing among major information overload studies.


THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 341

not show whether authors integrate perspectives of other der to keep up with the developments in their field, re-
research disciplines, such as cognitive science or psychol- searchers have to specialize and limit their research scope.
ogy. Clearly, future studies should draw more extensively Here editors and reviewers have a vital role to play. They
on existing research in other contexts (Akin, 1997). can inform researchers about related and relevant findings
that have been excluded from the researcher’s focus area.
Editors may also occasionally ask reviewers from other
CONCLUSIONS
disciplines to evaluate submitted articles so that they can
In conclusion, we discuss some of the limitations of our provide suggestions from other fields. Another approach
approach and we highlight implications of our analysis for could be to create dedicated interdisciplinary journals that
future research on information overload. encourage contributions that cross traditional disciplinary
The limitations of this review article relate to its method- boundaries (such as The Information Society).
ology and scope. In terms of methodology, the approach This push toward interdisciplinarity does not have to
we have chosen is a qualitative, inductive one with a fo- lead to an identity crisis of a discipline such as MIS
cus on surfacing the major categories in the overload dis- (Benbasat & Zmud, 2003) if the individual disciplines are
course. A bibliometric approach could have revealed more aware of their relative strengths and weaknesses regard-
detailed results regarding the impact of certain overload ing a particular research topic. As shown in our review
contributions, and a hypothesis-based method would have article, each one of the four domains has its advantages
led to a greater focus in the article. Our aim, by contrast, and drawbacks for the study of information overload. The
has been to provide a broad overview of the main dis- advantage of MIS studies on information overload, for
course elements in four business-related fields. In terms of example, is their focus on solutions and on the effects of
scope, it is clear that the four fields we reviewed are not the new information technology on the individual, the group,
only areas where information overload is a major concern. and the organization. However, it may have overlooked
Library studies, pedagogy, military studies, and entertain- some of the organizational parameters and their role in in-
ment are other domains that could have been included in creasing or decreasing information overload. Organization
the review. Our focus, however, has been on central disci- science can provide insightful suggestions on this score
plines of business-related research. Based on the reviewed (such as the effect of decentralization on the amount of in-
literature in these four fields, several directions for future formation that needs to be communicated and processed)
research can be envisioned. leading to more realistic information technology (IT) solu-
One, we need to employ alternative research methods tions. On the other hand, the organizational point of view
that can be used to study the phenomenon. Our review may lack insights on the individual’s reactions to such
shows that information overload has mainly been studied changes. This, in turn, is where the experimental studies
via experiments with the exception of a few studies that of accounting and marketing can provide helpful findings
used surveys, qualitative interviews, click-through anal- and methodologies. These insights can be combined and
ysis, document analysis, and formal modeling methods. operationalized in empirical investigations, for example,
Because of this dominant focus on experiments, we sug- regarding the effects of e-mail load and e-mail policies on
gest that other research methods should be employed in worker productivity, decision quality, and communication
order to triangulate prior findings. Such methods could behavior.
include ethnographies, action research, case studies, and Such interdisciplinary approaches to the study of over-
longitudinal studies, all of which capture more of the con- load may, however, require research projects based on
textual side of the overload problem than experiments. interdisciplinary teams that combine the talents of (for ex-
These inductive methods can then lead to more informed ample) MIS, accounting, and organization science schol-
hypotheses and refined experiments. A longitudinal ap- ars. They may reduce the potential overload for the indi-
proach could for example be used to examine the effects vidual researcher, as he or she can focus on his or her area
of prolonged overload on an organization and on the pro- of expertise while incorporating insights from other areas
cessing capacity of its employees. through other team members. The reasons why such in-
Two, we need to examine ways of increasing the amount terdisciplinary research teams are not more common are
of cross-fertilization in information overload research. As manifold and include existing research habits, assumptions
we have noted at various points in this article, truly inter- and methods, and institutional barriers, as well as com-
disciplinary approaches are not very common in informa- munication and terminology problems. Nevertheless, the
tion overload research. One of the reasons for this, and— overload problem calls for interdisciplinary approaches as
we suspect—for the lack of transfer among empirical and many of the open research questions in this field cross
conceptual studies, is that conducting interdisciplinary re- traditional disciplinary boundaries (from understanding
search increases the degree of information (over)load for individual coping behavior to designing organizational
the researchers themselves (Bawden, 2001, p. 9). In or- countermeasures).
342 M. J. EPPLER ET AL.

One such open research question relates to the interrela- Chewning, E. C., Jr., and Harrell, A. M. 1990. The effect of information
tionships highlighted in our conceptual framework, specif- load on decision makers’ cue utilization levels and decision quality
ically the reciprocal effects of technological, personal, in a financial distress decision task. Accounting, Organizations and
information-based, task-oriented, and organizational Society 15:527–542.
changes. Our framework and its derived models for sur- Cook, G. J. 1993. An empirical investigation of information search
strategies with implications for decision support system design.
vey research on causes, symptoms, and countermeasures
Decision Sciences 24:683–699.
can be used for such purposes. This can lead to a rank- Denning, P. J. 1982. Electronic junk. Communications of the ACM
ing that distinguishes high-impact causes and countermea- 25:163–168.
sures from low-impact causes and inefficient countermea- Denton, D. K. 2001. Better decisions with less information. Industrial
sures. Although the tables in this article provide a good Management 43:21–26.
overview of possible causes and countermeasures, they do Edmunds, A., and Morris, A. 2000. The problem of information over-
not yet qualify or prioritize these factors. Future research load in business organizations: A review on the literature. Interna-
should thus examine the interrelationships between the tional Journal of Information Management 20:17–28.
listed causes and countermeasures in more detail. This Eppler, M. 1998. Informative action: An analysis of management and
can lead to MIS solutions that address the problem drivers the information overload. PhD thesis, HEC Management Studies,
and root causes of the overload issue with effective coun- University of Geneva, Geneva.
Evaristo, R. 1993. An empirical investigation of the impact of informa-
termeasures.
tion characteristics and information technology on individual infor-
mation load. PhD thesis, Carlson School of Management, University
of Minnesota, Minneapolis.
REFERENCES
Friedmann, M. 1977. Consumer use of informational aids in supermar-
Abdel-Khalik, A. R. 1973. The effect of aggregating accounting reports kets. Journal of Consumer Affairs 78–155.
on the quality of the lending decision: An empirical investigation. Galbraith, J. R. 1974. Organization design: An information processing
Journal of Accounting Research Suppl. 11:104–138. view. Interfaces 3:28–36.
Ackoff, R. L. 1967. Management misinformation systems. Manage- Grether, D., Schwartz, A., and Wilde, L. 1986. The irrelevance of in-
ment Science 14:147–156. formation overload: An analysis of search and disclosure. Southern
Akin, L. K. 1997. Information overload. A multi-disciplinary explica- California Management Review 59:277–303.
tion and citation ranking within three selected disciplines: Library Griffeth, R. W., Carson, K. D., and Marin, D. B. 1988. Information
studies, psychology/psychiatry, and consumer science 1960–1996. overload: A test of an inverted U hypothesis with hourly and salaried
PhD thesis, Texas Woman’s University. employees. Academy of Management Proceedings 232–237.
Allert, J. L. 2001. A 12-step (or so) program for information junkies. Grise, M., and Gallupe, R. B. 1999/2000. Information overload: Ad-
Training & Development 55:32–37. dressing the productivity paradox in face-to-face electronic meetings.
Ansari, A., and Mela, C. F. 2003. E-Customization. Journal of Market- Journal of Management Information Systems 16(3):157–185.
ing Research 15:131–145. Haksever, A. M., and Fisher, N. 1996. A method of measuring infor-
Baldacchino, C., Armistead, C., and Parker, D. 2002. Information over- mation overload in construction project management. Proceedings
load: It’s time to face the problem. Management Services 46:18–19. CIB W89 Beijing International Conference, pp. 310–323.
Bawden, D. 2001. Information overload. Library & Information Brief- Hanka, R., and Fuka, K. 2000. Information overload and “just in time”
ings 92. http://litc.sbu.ac.uk/publications/lframe.html (accessed 10 knowledge. Electronic Library 18:279–284.
June 2002). Hansen, M. T., and Haas, M. R. 2001. Competing for attention in knowl-
Bawden, D., Holtham, C., and Courtney, N. C. 1999. Perspectives on edge markets: Electronic document dissemination in a manage-
information overload. ASLIB Proceedings 51(8):249–255. ment consulting company. Administrative Science Quarterly 46:1–
Benbasat, J., and Zmud, R. W. 2003. The identity crisis within the IS 28.
discipline: Defining and communicating the discipline’s core prop- Herbig, P. A., and Kramer, H. 1994. The effect of information overload
erties. MIS Quarterly 27(2):183–194. on the innovation choice process. Journal of Consumer Marketing
Berghel, H. 1997. Cyberspace 2000: Dealing with information over- 11:45–54.
load. Communications of the ACM 40:19–24. Hiltz, S., and Turoff, M. 1985. Structuring computer-mediated commu-
Casey, C. J., Jr. 1980. Variation in accounting information load: The nication systems to avoid information overload. Communications of
effect on loan officers’ predictions of bankruptcy. Accounting Review the ACM 28:680–689.
55:36–50. Hunt, R. E., and Newman, R. G. 1997. Medical knowledge overload:
Casey, C. J. 1982. Coping with information overload: The need for A disturbing trend for physicians. Health Care Management Review
empirical research. Cost and Management 4:31–38. 22:70–75.
Chan, S. Y. 2001. The use of graphs as decision aids in relation to Hwang, M. I., and Lin, J. W. 1999. Information dimension, informa-
information overload and managerial decision quality. Journal of tion overload and decision quality. Journal of Information Science
Information Science 6:417–426. 25(3):213–218.
Chervany, N., and Dickson, G. 1974. An experimental evaluation of Iselin, E. R. 1988. The effects of information load and information di-
information overload in a production environment. Management versity on decision quality in a structured decision task. Accounting,
Science 10:1335–1344. Organizations and Society 13:147–164.
THE CONCEPT OF INFORMATION OVERLOAD: REVIEW 343

Iselin, E. R. 1993. The effects of the information and data prop- Nelson, M. R. 2001. We have the information you want, but getting it will
erties of financial ratios and statements on managerial deci- cost you: Being held hostage by information overload. http://www.
sion quality. Journal of Business Finance & Accounting 20:249– acm.org/crossroads /xrds1-1/mnelson.html (accessed 14 June 2002).
267. O’Reilly, C. A. 1980. Individuals and information overload in organiza-
Jacoby, J. 1977. Information load and decision quality: Some contested tions: Is more necessarily better? Academy of Management Journal
issues. Journal of Marketing Research 14:569–573. 23:684–696.
Jacoby, J. 1984. Perspectives on information overload. Journal of Con- Olsen, K. A., Sochats, K. M., and Williams, J. G. 1998. Full text search-
sumer Research 10:432–436. ing and information overload. Intelligent Information and Library
Jacoby, J., Speller, D. E., and Berning, C. K. 1974. Brand choice be- Review 30:105–122.
havior as a function of information load: Replication and extension. Owen, R. S. 1992. Clarifying the simple assumption of the information
Journal of Consumer Research 1:33–43. load paradigm. Advances in Consumer Research 19:770–776.
Jones, B. 1986. Dying for information? Management Review 86:9. Payne, J. W. 1976. Task complexity and contingent processing in deci-
Keller, K. L., and Staelin, R. 1987. Effects of quality and quantity of in- sion making: An information search and protocol analysis. Organi-
formation on decision effectiveness. Journal of Consumer Research zational Behavior and Human Performance 16:366–387.
14:200–213. Revsine, L. 1970. Data expansion and conceptual structure. Accounting
Kock, N. 2000. Information overload and worker performance: A Review 45:704–712.
process-centered view. Knowledge and Process Management 4:256– Rudd, J., and Rudd, M. J. 1986. Coping with information load:
264. User strategies and implications for librarians. College & Research
Koniger, P., and Janowitz, K. 1995. Drowning in information, but thirsty Libraries, 47:315–322.
for knowledge. International Journal of Information Management Russo, J. E. 1974. More information is better: A reevaluation of Jacoby,
15:5–16. Speller, and Kohn. Journal of Consumer Research 1(3):68–72.
Lesca, H., and Lesca, E. 1995. Gestion de l’information, qualité de Scammon, D. L. 1977. ‘Information load’ and consumers. Journal of
l’information et performances de l’entreprise. Paris: Litec. Consumer Research 4:148–156.
Libby, R., and Lewis, B. L. 1982. Human information processing in Schick, A. G., Gorden, L. A., and Haka, S. 1990. Information over-
accounting: The state of the art in 1982. Accounting, Organizations load: A temporal approach. Accounting Organizations and Society
and Society 7:231–285 15:199–220.
Libowski, Z. 1975. Sensory and information inputs overload: Behav- Schneider, S. C. 1987. Information overload: Causes and consequences.
ioral effects. Comprehensive Psychiatry 16:199–221. Human Systems Management 7: 143–153.
Malhotra, N. K. 1982. Information load and consumer decision making. Schroder, H. M., Driver, M. J., and Streufert, S. 1967. Human infor-
Journal of Consumer Research 8:419–431. mation processing—Individuals and groups functioning in complex
Malhotra, N. K. 1984. Reflections on the information overload social situations. New York: Holt, Rinehart, & Winston.
paradigm in consumer decision making. Journal of Consumer Schultze, U., and Vandenbosch, B. 1998. Information overload in a
Research 10:436–441. groupware environment: Now you see it, now you don’t. Journal of
Malhotra, N. K., Jain, A. K., and Lagakos, S. W. 1982. The information Organizational Computing and Electronic Commerce 8(2):127–148.
overload controversy: An alternative viewpoint. Journal of Market- Shenk, D. 1997. Data smog. Surviving the information glut. London:
ing 46:27–37. Abacus.
Maes, P. 1994. Agents that reduce work and information overload. Simnet, R. 1996. The effect of information selection, information pro-
Communications of the ACM 37:31–40 cessing and task complexity on predictive accuracy of auditors.
McKinnon, S. M., and Bruns, W. J., Jr. 1992. The information mosaic. Accounting, Organizations and Society 21:699–719.
Boston: Harvard Business School Press. Simon, H. A. 1979. Information processing models of cognition. Annual
Meier, R. L. 1963. Communications overload: Proposals from the study Review of Psychology 30:363–396.
of a university library. Administrative Science Quarterly 7:521– Simon, H. A., and Newell, A. 1971. Human problem solving: The state
544. of the theory in 1970. American Psychologist 26:145–159.
Meglio, C. E., and Kleiner, B. E. 1990. Managing information Simpson, C. W., and Prusak, L. 1995. Troubles with information
overload. Industrial Management & Data Systems 1. http://www. overload—Moving from quantity to quality in information provision.
leporello.emeraldinisght.com (accessed 22 June 2002). International Journal of Information Management 15:413–425.
Meyer, J. 1998. Information overload in marketing management. Mar- Snowball, D. 1980. Some effects of accounting expertise and informa-
keting Intelligence & Planning, 16:200–209. tion load: An empirical study. Accounting, Organizations and Society
Meyer, R. J., and Johnson, E. J. 1989. Information overload and the 5(3):323–338.
nonrobustness of linear models: A comment on Keller and Staelin. Sorohan, E. G. 1994. Coping with overload. Training & Development
Journal of Consumer Research 15:498–503. 48:13.
Milgram, S. 1970. The experience of living in the cities. Science 13 Sparrow, P. R. 1998. Information overload. In The experience of man-
March:1461–1468. aging: A skills workbook, eds. K. Legge, C. Clegg, and S. Walsh,
Miller, J. A. 1956. The magical number seven plus or minus two: Some pp. 111–118. London: MacMillan.
limits on our capacity for processing information. Psychological Re- Sparrow, P. R. 1999. Strategy and cognition: Understanding the role
view 63:81–97. of management knowledge structures, organizational memory and
Muller, T. E. 1984. Buyer response to variations in product information information overload. Creativity and Innovation Management
load. Psychological Review 63:81–97. 8:140–149.
344 M. J. EPPLER ET AL.

Speier, C., Valacich, J. S., and Vessey, I. 1999. The influence of task Walsh, J. P. 1995. Managerial and organizational cognition: Notes
interruption on individual decision making: An information overload from a trip down memory lane. Organization Science 3:280–
perspective. Decision Sciences 30:337–359. 321.
Swain, M. R., and Haka, S. F. 2000. Effects of information load on capi- Wang, R. Lee, Pipino, Y. W., and Strong, D. 1998. Manage your infor-
tal budgeting decisions. Behavioral Research in Accounting 12:171– mation as a product. Sloan Management Review 4:95–105.
199. Webster, J., and Watson, R. T. 2002. Analyzing the past to prepare
Tushman, M. L., and Nadler, D. A. 1978. Information processing as an for the future: Writing a literature review. MIS Quarterly 26(2):xiii–
integrating concept in organizational design. Academy of Manage- xxiii.
ment Review 3:613–625. Weick, K. E. 1970. The twigging of overload. In People and informa-
Tuttle, B., and Burton, F. G. 1999. The effects of a modest incentive tion, ed. H. B. Pepinsky, pp. 67–129. New York: Pergamon.
on information overload in an investment analysis task. Accounting, Wilkie, W. L. 1974. Analysis of effects of information load. Journal of
Organizations and Society 24:673–687. Marketing Research 11:462–466.
Van Zandt, T. 2001. Information overload in a network of targeted Wilson, P. 1996. Interdisciplinary research and information overload.
communication. INSEAD Working Paper 2001/36/EPS. Available Library Trends 45:192–203.
at http://ged.insead.edu/fichiersti/inseadwp2001/2001-36.pdf Wurman, R. S. 1990. Information anxiety. What to do when information
Vollmann, T. E. 1991. Cutting the Gordian knot of misguided perfor- doesn’t tell you what you need to know. New York: Bantam Books.
mance measurement. Industrial Management & Data Systems 1:24– Wurman, R. S. 2001. Information anxiety 2. Indianapolis, IN:
26. Macmillan.