Web Systems Design and Online Consumer Behavior [Electronic resources] نسخه متنی

Theoretical Background

There are several important bodies of literature that need to be examined here to provide the
theoretical background for the research framework developed in this study. In the following
section, the theoretical background of objective complexity; perceived complexity, cognitive
complexity, and optimal complexity will be discussed.

As an important element, complexity is frequently mentioned in system design (Lawler &
Elliot, 1996), public affairs (Al-Menayes & Sun, 1993), social psychology (Satish &
Streufert, 1997), behavioral science (Goswami, 1998; Sweller, 1998), information technology
(Banker, Davis & Slaughter, 1998; Hurt & Hibbard, 1989; Jacobson, 1992), human resources
(Meyer, Shinar, & Leiser, 1997), and advertising (Anderson & Jolson, 1980; Chamblee,
Gilmore, Thomas, & Soldow, 1993; George 1998; Morrison & Dainoff, 1972).

The definition of complexity varies depending on the domain and the purpose of research. Even
within the same domain, complexity could be defined differently. For instance, to measure the
complexity of advertising copy, Morrison and Dainoff (1972) tested the effect of visual complexity
of an advertisement. The complexity is defined in terms of the geometric characteristics of the
picture. Anderson and Jolson (1980), on the other hand, measure the technique complexity of
advertising copy, while Chamblee et al. (1993) employ the measurement of lexical complexity of ad’s
copy.

Despite the abundant literature discussing the issue of complexity, little attention has been
paid to the differences between objective complexity and perceived complexity. To gap this lack of
research, I arranged the following session based on the categorization of complexity.

Objective Complexity

To most of the Internet users, the Internet itself is a stimulus that provides surfing
experiences to fulfill the users’ task, i.e., the issue of objective complexity could be approached
using both the data displayed and the task performed. Hence, the two most relevant discussions on
objective complexity to our current purpose are Berlyne’s stimulus complexity and Wood’s task
complexity.

• 
  

  Berlyne’s Stimulus Complexity: Berlyne (1958) carried out
  a series of studies on visual figures. He considered stimulus complexity using the different
  complexity forms — irregularity of arrangement, amount of material, heterogeneity of elements,
  irregularity of shape; incongruity, and incongruous juxtaposition.

  Berlyne’s theory has broad influence in lots of studies. For instance, Meyer et al. (1997)
  study complexity of displayed data using measurement based on Berlyne’s (1971, 1974) stimulus
  complexity theory. The complexity of displayed data is defined in terms of three factors: (a) the
  number of points in the display (display with fewer data points are less complex); (b) the
  configuration of the display (i.e., 12 data points may appear on one line of 12, two lines of six,
  three lines of four, or four lines of three points); and (c) the regularity of the displayed data
  (whether they are dispersed erratically).



• 
  

  Wood’s Task Complexity: Wood (1986), on the other hand,
  coins different sets of constructs to define complexity of task. He argues that task complexity,
  which “describes the relationships between task inputs, will be an important determinant of human
  performance through the demands it places on the knowledge, skills, and resources of individual
  task performers” (p. 66). Three types of task complexity are defined in his construct: component,
  coordinative, and dynamic.

Component complexity of a task is defined as a direct function of the number of distinct acts
need to be executed in the task performance and the number of distinct information cues must be
processed in the performance of those acts. Coordinative complexity refers to the nature of
relationships between task inputs and task products. The forms and strength of the relationships
between information cues, acts, and products, as well as the sequencing of inputs, are all aspects
of coordinative complexity. In addition to the static complexity of the acts and information cues
needed to perform a task, individuals must frequently adapt themselves to changes in the
cause-effect chain or means-ends hierarchy for a task or during the performance of the task. The
third dimension of task complexity, which is named dynamic complexity, is due to changes in the
states of the world which has an effect on the relationships between task inputs and products. In
dynamically complex tasks the parameter values for the relationships between task inputs and
products are non-stationary. Changes in either the set of required acts and information cues or the
relationships between inputs and products can create shifts in the knowledge or skills required for
a task (Wood, 1986). Wood’s (1986) concept of complexity has been well adopted in the decision
science field. For instance, Banker et al. (1998) adopt this multidimensional definition of
complexity (Wood, 1986) to map dimension of software complexity into component, coordinative and
dynamic dimensions.

Perceived Complexity

I have described two models of objective complexity in
the
previous section. Objective
measurement alone, however, is usually handicapped by inaccuracy. Objective measures often downplay
the distinction between objective and subjective constructs, though subjective constructs are
supposed to be more important in online consumer research due to the following reasons. (1) Web
surfers’ preference toward a Web site is theorized to be based on subjective as opposed to
objective complexity (Beach & Mitchell, 1978), and (2) other research suggests that subjective
measures are often in disagreement with their objective counterparts (Abelson & Levi, 1985;
Adelbratt & Montgomery, 1980; Wright, 1975). Introducing measures of the Web surfers’ own
perceived complexity, independent of the objective one that actually exists, has been suggested as
a way of mitigating criticisms that the intended framework is tautological (Abelson & Levi,
1985).

Cognitive science researchers have long been recognizing the importance of the perceived
complexity. For instance, Rao (1985) points out that as perceived complexity of tasks went up, so
did the amount of information acquired and the level of processing. Researchers in other fields
such as information science also allude to the importance of perceived complexity. For instance,
Davis (1989) has found that perceived ease of use of a system is one of the fundamental
determinants of system use.

The present online consumer research can be roughly categorized into two streams: the
system-centered approach versus a user-centered approach (Unz & Hesse, 1999). System-centered
research is largely concerned with objective characteristics or formal features of Web sites (Bucy,
Lang, Potter, & Grabe, 1999; Ghose & Dou, 1998; Ha & James, 1998; Stout, Villegas,
& Kim, 2001), whereas the user-centered approach tends to focus on user response to and
perception of the Web site (Chen & Wells, 1999; Eighmey, 1997; Eighmey & McCord, 1998).
These two streams both provide meaningful insights into the understanding of this new media,
although a more comprehensive understanding would have been warranted if the researchers had taken
both approaches into consideration. This research hence aims to address this important issue by
taking both of the approaches into consideration, i.e., the objective complexity will be proposed
under a system-centered approach while the perceived complexity will be proposed under a
usercentered approach. Specifically, objective complexity serves as a convenient benchmark for
e-marketers to design its Web component, while perceived complexity more reliably reflects the
consumers’ perception. As such, it is meaningful to explore the relationship between the objective
and perceived complexity as well as their individual and combined impacts on the communication
effectiveness of a Web site. In addition, considering both the objective and the perceived
complexity might help marketers to chart the optimal complexity—the congruency between Web user
characteristics (e.g., Web experience, focused attention) and Web site characteristics (e.g.,
structural components of the Web site).

Optimal Complexity

Optimal level complexity has been explicitly and implicitly researched by several previous
studies in a variety of areas. For instance, Wood (1986) suggests the possibility of a curvilinear
relationship between complexity and performance, which has important implications for the design of
tasks. He further suggests that if the goal in task design is to maximize outcomes such as
performance and job satisfaction, task should be designed which include optimal levels of
complexity for particular groups of individuals.

Russo and Leclerc (1991) found presenting too little information may not let individuals
fulfill their information needs. On the other hand, presenting too much information can hinder an
individual’s ability to efficiently comprehend and analyze information (Estelami, 1997; Russo,
1975). Previous studies have shown that focusing on improving information displays can have a
greater impact than simply supplying additional information (Bettman, 1979; Magat & Viscusi,
1992; Payne et al., 1993; Russo, 1977; Selart, Garling, & Montgomery, 1998).

Some researchers (Sieber & Lanzetta, 1964; Streufert, Suedfeld, & Driver, 1965) found
a curvilinear relationship between perceived environmental complexity and information search with
the latter declining after a certain level of the former. However, other researchers like Lussier
and Olshavsky (1974) found no evidence of a decline in searching or processing as the environment
became more and more complex. LeMay and Aronow (1977), however, found a strictly linear
relationship in a small scale study of laid out area in a suburban shopping center. Their subjects
looked longer at figures of increasing complexity than at simpler figures. Other studies which
reveal similar findings were conducted by Scammon (1977), Einhorn (1971), Streufert (1970) and
Deane, Hammond and Summers (1972).

Bettman (1979) attempts to explain this apparent contradiction by suggesting that time is the
crucial factor. A limited amount of time to make a choice forces decreases in search and level of
processing but if the consumer can devote as much time as desired, then the relationship continues
to be strictly positive.

Berlyne (1958) found in his study that the more complex stimulus receive longer inspection
from subjects. Meanwhile, he also indicated that the “more complex” figures were not tremendously
complex. Much more complex patterns than the ones he used in the study might conceivably have been
shunned rather than preferentially inspected. North and Hargreaves’ (1996) study on subjects’
responses to music in aerobic exercise and yogic relaxation classes support the prediction of
Berlyne’s theory that there should be an inverted U-relationship between ratings of liking and
complexity: with increasing complexity, liking initially increases, but then decreases
substantially.

Since communication effectiveness anchors on the users’ liking toward their experience, as
well as on the search efforts they are willing to make, in this research, I take on the tradition
that aestheticians have often asserted that an intermediate degree of complexity (“Unity in
diversity”) makes for maximum appeal, and hence, maximum communication effectiveness. I contend
that the exact degree of the optimal complexity should be charted dynamically based on the level of
objective and perceived complexity and may vary depending on personality traits. The most important
personality trait needing to be considered is consumers’ cognitive complexity.

Cognitive Complexity: Consumer as Limited Information
Processor

In the area of consumer behavior research, the
consumer is paradigmatically viewed as a limited information processor. The most prevalent
contemporary view in the consumer behavior literature is that a consumer is a “limited information
processor,” i.e., “that consumers have limited capacities to process information” (Bettman,
Johnson, & Payne, 1991, p. 57). This view is based upon three attentional theories, i.e.,
Filter theory (Broadbent, 1958; 1971), Capacity theory (Kahneman, 1973) and Resources theory (e.g.,
Navon & Gopher, 1979; Norman & Bobrow, 1975), originated from auditory and visual
perception research. The consumer behavior literature adopted these viewpoints and treated
consumers as limited information processors (De Heer, 1999).

Further, the conflict between human’s limited cognitive capacity (Miller, 1956) and the
information intensity varies among people with different levels of cognitive complexity
requirement. Cognitive complexity has been viewed as a structural variable impacting how one
construes objects from the environment. Cognitively complex individuals have been described as more
likely to seek information in order to interact in the environment (Levanthal, Singer, & Jones,
1965), more likely to seek diversity in the environment (Bieri et al., 1966), more capable of
integrating information, and more likely to be “‘promiscuous information gathers’ who have high
rates of exposure to persuasive communications, whether consistent with current belief, or not”
(Reardon, 1981, p. 128).

For information processing in a short time, Miller (1956) points out that when the amount of
input information is increased, the observer will begin to make more and more errors. Hence, there
is a limit of accuracy of an individual’s absolute judgments. He uses “Channel Capacity” to
represent the greatest amount of information that an individual can be given about the stimulus on
the basis of absolute judgments. Consumers’ cognitive complexity level decides their ultimate level
of channel capacity. Hence, the relationship between a Web site’s objective complexity and
perceived complexity as well as their individual and combined impacts on the Web site’s
communication effectiveness may very well be moderated by individuals’ cognitive complexity.