L ,
”fin, _
. .
.
L
. , L .4.-
..Eu. .. m
y
5“! -w u .l . ,
LE1? . 5a
,Aav.n.a¢.; A. 6. .3
, .n . .
l
, .
711.19.!
. \i. 21.0.
THESIS
l o O ’3
37/ ‘7/9
2..
9’. ( “
b g i
LIBRARY
Michigan State
University
This is to certify that the
thesis entitled
PERCENED COSTS AND VIRTUAL EXPERIENCE:
THE ADDED VALUE OF RICH CONTENT IN
BZC ELECTRONIC COMMERCE
presented by
Younbo Jung
has been accepted towards futfillment
' of the requirements for the
MA. degree in TELECOMMUNICATION
V Majof’Pgfis’sors Signature
April 15, 2003
Date
MSU is an Aflimatfla Arman/Equal Opportunity Minion
"F’flfl— hm“ -I
F' p7 “P* ‘ —'v '— —- .
,.v_’ —v—~_——-—v- “vfiwvv—fi- _‘,_—
‘ ”Y ..—.-‘~_ .. -— _-_.- -—- --- 'fi.~ a.— *— _‘,, w—v.’——
r’V- - - 4'
PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.
DATE DUE | DATE DUE DATE DUE
mm |
AUBEMJQFE
6/01 c:/ClRC/DateDue.965-p. 15
PERCEIVED COSTS AND VIRTUAL EXPERIENCE: THE ADDED VALUE OF
RICH CONTENT IN BZC ELECTRONIC COMMERCE
By
Younbo Jung
A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF ARTS
Department of Telecommunication
2003
ABSTRACT
PERCEIVED COSTS AND VIRTUAL EXPERIENCE: THE ADDED VALUE OF
RICH CONTENT IN B2C ELECTRONIC COMlVfERCE
By
Younbo Jung
Many studies have demonstrated the positive effects of virtual experience on
BZC e-commerce. However, providing rich content may be a double edged sword for e-
Retailers due to inevitable download delays under the current “last mile” Internet
infrastructure. Therefore, the goal of this study is to examine the relationships among
virtual experience, connection speeds, and product attributes to develop a clearer
understanding of how to balance the Web page content and loading time delays across
diflerent product categories in B2C e-commerce. An economic model for the added value
of rich content was conceptualized to help understand the relationships better.
An experimental research method was used to test the hypotheses. Although the
impacts of download delays on the added value of rich content were not clearly explained
in this study, the interpretation of the findings and suggestions for future studies were
discussed to address guidelines for the dilemma of rich content.
Copyright By
YOUNBO JUNG
2003
To My Parents
ACKNOWLEDGEMENTS
I would like to thank the fine faculty of the department of Telecommunication at
Michigan State University for assisting me in the realization of this project. Especially, I
am grateful to my advisor and committee chairperson, Dr. Charles Steinfield, for his
guidance and encouragement through this endeavor as well as my entire graduate study. I
would also like to extend my gratitude to Dr. Hairong Li for letting me use 3D rendered
images designed for his study as well as his helpful advice on my research.
Most of all, I would like to thank my family members. Without their love, trust,
and encouragement, I could have not completed this work. I am sincerely grateful to my
wife for being next to me always. Finally, I would like to thank God for everything.
TABLE OF CONTENTS
LIST OF TABLES ................................................................................. v1ii
LIST OF FIGURES ............................................................................... ix
LIST OF FORMULA ............................................................................. x
I INTRODUCTION .............................................................................. 1
H LITERATURE REVIEW ...................................................................... 4
B2C E-Commerce ........................................................................... 5
Virtual Experience ........................................................................... 6
Perceived Costs ............................................................................. 10
Product Categorization ..................................................................... 16
III HYPOTHESES ................................................................................ 19
IV METHODOLOGY ............................................................... ' ............ 25
Research Design ......................................................................... 25
Participants .................................................................................. 27
Stimulus ..................................................................................... 28
Procedures ................................................................................... 31
V ANALYSIS .................................................................................... 34
VI DISCUSSION AND LIMITATIONS ................................................... 40
Hypothesis 1 ................................................................................ 4O
Hypothesis 2 .............................................................................. 43
Hypothesis 3 and 5 ......................................................................... 46
Hypothesis 4 .............................................................................. 48
Hypothesis 6 .............................................................................. 48
Other consideration and implication for future studies ............................... 49
Limitation .................................................................................... 51
vi
Vll CONCLUSION ................................................................................ 52
APPENDICES ................................................................................. 54
BIBLIOGRAPHY ............................................................................ 58
vii
LIST OF TABLES
Table l: Anticipation of consumers’ preferences ............................................ 18
Table 2: 2 X 2 X 2 Between Subject ANOVA design for the study ........................ 19
Table 3: Overview of descriptive statistics for each experimental condition .............. 34
Table 4: Results of the hypothesis tests ......................................................... 36
Table 5: Possible explanations for the findings of this study ................................. 41
Table6: Attitude toward delays .................................................................. 46
viii
LIST OF FIGURES
Figure 1: Broadband access in OECD countries per 100 inhabitants ........................ 14
Figure 2: Screen shot of the experience good (a box of cereal) ............................... 29
Figure 3: Screen shot of the search good (a camcorder) ...................................... 30
Figure 4: Screen shot of the loading page with an active loading bar ...................... 31
Figure 5: HTML coding for a 30 second delay ................................................ 31
Figure 6: Screen shot of the index page ......................................................... 32
/
ix
LIST OF FORMULA
Figure 1: Added value of rich content in B2C e-commerce ................................... 9
Figure 2: Perceived Costs ........................................................................ 11
I INTRODUCTION
The resurgence of the US. economy in the 1990’s coincides with the growing
use of the Internet, including the rapid growth of electronic commerce (e-commerce). The
web is now considered a mainstream and viable channel for conducting commerce (Amor
2000), and is thereby transforming traditional business and consumer life (ITU, 1999).
Despite the collapse of dot-com bubbles, e-commerce is still expanding to banks, stock
trading, auction, entertainment, and all other industries. According to the International
Benchmarking Study (IBS) 2000, 83% of business is on-line, 26% of business trades on-
line in the United States. As a result, many e-commerce merchants are considering
turning to 3D product presentations to stand out in such a competitive environment and to
provide consumers with a greater sensory experience (Nash, 2000).
In accordance with this trend, many studies have demonstrated the positive
effects of virtual experiences (created by 3D environments) on product knowledge and
consumer attitude toward the products (Biocca, Li, Daugherty, & Chae, 2001; Hoch &
Deighton, 1989; Klein, 2001; Hauble & Figueroa, 2001; Li, Daugherty, & Biocca, 2001).
However, most of these studies focus on the relationship between virtual experience and
consumer behavior without considering the current Internet infi‘astructure or connection
speeds. Since dial-up connections are still predominant in most countriesl, these findings
may not hold up in the field. The added value of 3D presentation may generally diminish
under low connection speeds. In fact, a significant relationship between download delay
and the likelihood of aborting a page load has important implications for e-Retailers to
balance their page content and download speeds (Rose, Lees, & Meuter, 2001).
Therefore, the goal of the study is to identify significant factors affecting the
added value of rich content in business-to-consumer (BZC) e-commerce. Among possible
factors affecting B2C e-commerce, virtual experience, connection speeds, and product
attributes were selected for this study based on relevant theoretical works. The
relationships among virtual experience, connection speeds, and product attributes will be
examined to provide a clear understanding of how to balance the Web page content and
loading time delays across different product categories in B2C e-commerce. Furthermore,
an economic model for the added value of rich content, modified from the value creation
model by Brandenburger and Stuart, Jr. (1996) will be conceptualized to help understand
the relationships better.
1 The number of home broadband subscribers was about 24 million, 20% of all Internet users in the US.
in May 2002 (Surmacz, 2002)
In this paper, I will review relevant theoretical works centered on three key
concepts: virtual experience, connection speeds, and product attributes. Then, I will
describe a research project and test the hypotheses using an experimental research
method.
II LITERATURE REVIEW
There are three key concepts in this paper that need to be explained in order to
understand the added value of rich content in BZC e-commerce. As stated, virtual
experience created by 3 dimensional (3D) presentation of products may enhance
consumers’ attitudes toward products in B2C e-commerce. However, inevitable
download delays caused by virtual experience under the current last mile Internet
infrastructure may lead to negative consequences at the same time. Furthermore, the
relationship between virtual experience and connection speeds may vary depending on a
product’s attributes. A certain type of dominant product attribute can likely be enhanced
by virtual experience.
Therefore, this chapter will summarize the basic theoretical concepts underlying
virtual experience, connection speeds, and product attributes by reviewing the literature
supporting these concepts. Then, an economic model for the added value of rich content
will be conceptualized based on the three basic theoretical concepts. Finally, several
hypotheses suggested from these concepts will be proposed in next chapter.
B2C E-Commerce
First of all, B2C e-commerce must be defined in advance to any other theoretical
concepts for this paper since the definition of e-commerce varies widely. The concept of
electronic commerce is not new, and so, there is no one generally accepted definition of
e-commerce. Among many definitions by scholars and institutions, the following are
definitions used in the study.
. The term electronic commerce refers generally to commercial transactions,
involving both organizations and individuals, that are based upon processing
and the transmission of digitized data, including text, sound and visual images
and that are carried out over open networks (like the Internet) or closed
networks (like AOL or Minitel) that have a gateway onto an open network -
OECD, 1997
. Electronic commerce means using an electronic network to simplify and speed
up all stages of the business process from design and making to buying, selling
and delivery - DTI, 1998
As shown above, the definition of e-commerce may differ with respect to the
media under consideration. Some focus on the Internet, some include all sorts of direct
electronic distribution channels and others include all forms of electronic market places
(Schmitz, 2000).
Among various definitions, business activities such as marketing, selection,
payment, and delivery, all take place over the Internet to achieve a business goal. This
will be used as the definition of e-commerce for this study. In addition, e-commerce will
be limited to B2C transactions, Iwhich means Business-to-Consumer retail transaction for
goods and services.
VirLual Experience
The consumers’ best source of product and service information is direct product
experience, commonly referred to as trial or usage consumers (Hoch & Deighton, 1989;
Hoch & Ha, 1989). Direct experience has been defined as “an experience that stems out
of an unmediated interaction between the consumer and the product, with a person’s full
sensory capacity, including visual, auditory, taste-smell, and orienting” (Gibson, 1966;
Edwards & Gangdharbatla, 2001).
This direct experience has several advantages. First, evidence in direct
experience is self-generated and thus the most trustworthy for consumers themselves.
Second, consumers may manage the way a product is experienced by controlling the
focus and pace of an inspection to maximize informational input. Third, such an
interaction may result in more effective responses in consumers than indirect experience
(Miller & Millar, 1996; Biocca, Daugherty, Li, & Chae, 2001). Therefore, one of the e-
commerce marketer’s goals is to strive for verisimilitude in indirect communications with
consumers in e-commerce business where consumers cannot have direct experience
(Klein, 2001).
The Internet has the ability to serve as a more powerful medium than traditional
media in the sense that consumers are able to interact with products in 3D environments,
thus simulating a new form of experience, virtual experience (Biocca, Daugherty, Li, &
Chae, 2001). Virtual experience has been defined as a “vivid, involving, active, and
affective psychological state occurring in an individual interacting with three dimensional
computer simulation” (Li, Daugherty, & Biocca, 2001). As indicated earlier, many studies
support that virtual experience in 3D environments is superior to other mediated forms of
experience for facilitating information processing (Edwards & Gangdharbatla, 2001).
Many studies also have demonstrated the positive effects of virtual experience
(created by 3D environments) on product knowledge and consumer attitude toward the
products (Biocca, Li, Daugherty, & Chae, 2001; Hoch & Deighton, 1989; Klein, 2001;
Hauble & Figueroa, 2001; Li, Daugherty, & Biocca, 2001). However, one characteristic
of dynamic virtual experience is large data size, which requires a larger bandwidth
compared to simple and static 2D display. It may cause traffic problems on the Internet
and cannot bring a true virtual experience to consumers. Thus, the added valire of 3D
presentation may generally diminish under low connection speeds.
Formula 1 is an economic model for the added value of rich content in B2C e-
commerce, which has been modified from the formula of value creation by
Brandenburger and Stuart, Jr. (1996). Normally, willingness to pay refers to the actual
amount of money, for which consumers are willing to pay. But willingness to pay in
Formula 1 is conceptualized in terms of time value. Therefore, it represents the level of
how much time consumers are willing to spend to get particular information about
products on the Web. The concept of opportunity cost is defined in an analogous fashion
to willingness to pay and will be explained more detail in Perceived Costs section later.
Formula 1: Added value of rich content in B2C e-comrnerce
. Added Value of 3D Presentation == (1 x Willingness to Pay — [3 x Opportunity
Costs
. Where a = index from product attributes. Search goods may have higher index
than experience goods
. Where B = index fiom Internet connection speeds. High connection speeds
may have smaller index numbers than low connection speeds
. Where Willingness to Pay = consumers’ willingness to spend time to get
information about products on the Web.
Consumers generally have higher willingness to
pay for rich content or information
. Where Opportunity Costs = consumers’ perceived time value of getting
information on the Web.
Consumers are likely to have higher desire to spend more time to get detailed
information about products, which leads to higher willingness to pay for virtual
experience in Formula 1. In addition, willingness to pay for virtual experience may vary
depending on product categories since some product categories require more direct
experience while others are amenable to both indirect and direct experience. Therefore,
virtual experience and product attributes are considered positive factors for the added
value of rich content in Formula 1.
Perceived Costs
The concept of perceived costs is similar to opportunity cost in terms of its
influence on decision-making. Economists use the term opportunity cost to emphasize
that making choices in the face of scarcity implies a cost.
. The opportunity cost of any action is the best alternative forgone. The best thing
that you choose not to do — the forgone alternative — is the cost of the thing that
you choose to do - Parkin, 1993
Thus, consumers make decisions based on their perception of alternative values,
which could be money, time, or anything they prefer. Among these values, this study will
primarily focus on time value in consumers’ decision making. As a consequence,
consumers’ opportunity costs are likely to be affected by their Internet connection speeds
in B2C e-comrnerce. Therefore, the combined effects of consumers’ opportunity costs
10
with Internet connection speeds on the consumers’ minds, compared to their foregone
alternative in terms of time value will be defined as perceived costs in the study. After all,
perceived costs will be consumers’ opportunity costs for getting information in terms of
time value. Formula 2 shows a simple formula for consumers’ perceived costs, which is a
negative factor for the added value of rich content in Formula 1.
Formula 2: Perceived Costs
. Perceived Costs = B x Opportunity Costs
. Where B = Index based on consumers’ Internet connection speeds.
. Where B under high connection speeds < l
. Where B under low connection speeds > 1
Perceived costs may vary depending on each person’s value in his/her own mind,
but I assume that longer loading times cause relatively higher perceived costs and shorter
loading times cause relatively lower perceived costs to the same person if Ceteris Paribus
While there are many factors or values affecting the online consumer experience,
download speeds and the reliability of the Web sites directly impact them all. According
to Rose and Straub, there are six key technological impediments to e-commerce:
ll
download time, measurement of web application success, security weakness, lack of
internet standards, limitations in the interface, and requests for hypermedia (Rose &
Straub, 2000). Among the six key technological impediments to e-commerce, Rose, Huoy,
and Straub (1999) rank download time as the second most important. In addition, a recent
panel of experts in a Delphi study also ranked download delay as the single most
worrisome issue (Khosrowpour & Herman, 2000).
Loading time delays are also often cited as the reason for Web site abandonment.
In a survey by the Industry Standard, download delay was found to be the leading cause
for why consumers leave a site (Lake, 2001). In fact, only 47.3% of consumers are
satisfied with the speed of e-commerce Web sites. Even fewer consumers - 34.7% - view
the Internet as fast (Pedersen, 2001). According to a study conducted by ZDNet, e-
business is estimated to be losing US $362 million a month because visitors will not wait
for heavy pages to load, which shows that download time effects profitability. Obviously,
consumers with high-speed Internet access will have a different experience than those
who access the Internet through a dial-up connection, or narrowband, especially when the
Web site requires broadband connections for larger files or heavy content such as 3D
graphics.
12
The load time is measured fi'om the moment when a user clicks on a selection to
the point when all the new elements of the page have been downloaded to the user’s
computer (WebCriteria, 1999). In other word, download time is the time it takes for a
Web client to fully receive, process, and display files submitted by a Web server once
those files are requested (Rose & Straub, 2000).
While server side technologies can be improved and are within the control of e-
Retailers, client side technologies and last mile infrastructure used for reception of pages
and transmission of pages/ requests across the Internet are beyond the control of e-
Retailers, (Rose & Straub, 2001). Thus, loading time delays for rich content seem to be
inevitable despite the enormous growth of the Internet, since dial-up connection still
dominates the Internet infrastructure in the world.
According to the OECD report, just about 4.4 per 100 inhabitants, on average in
OECD countries, were subscribers to high speed Internet access in September 2002.
Although the United States is the largest market for broadband services in the OECD and
was ranked 3rd in terms of overall broadband penetration at the end of 2000 with 3.7
million cable modem subscribers and 2.4 million DSL subscribers, only about 6.2 per 100
inhabitants were subscribers to DSL, cable modem, or other broadband lines 2 in
2 fixed wireless broadband, direct satellite broadband, or fiber to the resident lines
13
September 2002 (OECD, 2002). Figure 1 shows the broadband penetration rates of
OECD countries in September 2002.
Figure 1: Broadband access in OECD countries per 100 inhabitants
20 J I Other broadband technologies subscribers
I Cable Modem subscribers
"if DSL subscribers
15']
10~
Iceland
Japan
Austria
E
M
§
5
’5
E
8
9
Sweden
Netherlands
Switzerland
(Source: OECD)
There is not any standardized definition about broadband, but definitions vary
widely. According to [TU-T Recommendation 1.113, broadband means transmission
14
capacity that is faster than primary rate ISDN (i. e. 1.5 or 2 Mbps). The FCC defined
‘broadband’ as having the capability of supporting, in both the provider-to-consumer
(downstream) and the consumer-to-provider (upstream) directions, a speed (in technical
terms, ‘bandwidth’) in excess of 200 kilobits per second (Kbps) in the last mile. In this
paper, the FCC definition is used and modified. Thus, broadband will be defined as
telecormnunication that provides multiple channels of data over a single communications
medium with a speed-rate over 200 Kbps in this paper.
Broadband technologies such as cable modem, DSL, or fixed wireless
broadband are expected to evolve fast. In fact, the number of residential broadband
subscribers was about 24 million, 20% of all Internet users in the United States, in May
2002 according to a recent report by the Pew Internet and American Life Project, a
nonprofit research group based in Washington, DC. (Surmacz, 2002). Although the
penetration rate of broadband access is growing fast in spite of technical and economical
problems in its early stages, the last mile infiastructure is still not mature enough to
support the latest 3D environments due to the predominance of dial-up connections.
Therefore, under the current last mile infrastructure, e-Retailers are likely to
reduce their content reluctantly without a clear understanding of its value, just to avoid
consumer fi'ustration and the propagation of negative attitudes toward their products and
15
services. The added value of rich content model might help e-Retailers to understand
when to reduce content or when to add more content.
Product Categorization
Product categorization is important in B2C e-commerce because some product
categories lend themselves better to brick and mortar (ofiline) shopping, which requires
more experience, while others are amenable to both online and offline shopping (McCabe
& Nowlis, 2001).
Nelson defined two types of product attributes, search and experience attributes.
Any product can have either one of or both of them (Nelson, 1976; 1981). Search
attributes are those features of a product that can be used to evaluate the quality of the
product without actual use of the product such as color, price, shape, size, and so on.
Good examples of search products can be personal computers or electronics, which are
pretty much standardized. On the other hand, experience attributes are those features of a
product that cannot be assessed until actual use or through direct experience, such as taste,
or texture. Good examples of experience products are music, books or foods (Li,
Daugherty, & Biocca, 2002).
16
Because of these different attributes of products as well as the limitation of
information that can be delivered electronically on the Internet, consumers’ willingness to
spend time to get information about products on the Web could be affected by product
categories. For search goods, consumers might be willing to sacrifice more time to get the
detailed information of search attributes on the Web, which means higher willingness to
pay. On the other hand, consumers might not be willing to spend much time in getting the
detailed information of experience attributes on the Web since such information may not
be helpfirl or usefirl for their decision-making. As a result, consumers’ willingness to pay
to get information about products would be lower in experience goods than search goods
on the Web.
After all, one of the purposes of this study is to find out what type of dominant
product attribute is more likely enhanced by virtual experience. Table 1 shows an
estimation of consumers’ preferences of product presentations under four different
conditions based on the concepts summarized earlier in this section.
As stated, consumers are likely to prefer virtual experience to simple
information if there were no loading time delays. On contrary, consumers may prefer
simple content if loading time delays for virtual experience are intolerable under low
connection speeds. However, there is also a chance that consumers still prefer virtual
l7
experience for search goods in spite of loading time delays. For search goods, the (1
index in Formula 1 could be big enough to make total willingness to pay larger than
perceived costs caused by loading time delays.
Table l: Anticipation of consumers’ preferences
Low Connection Speeds High Connection Speeds
(56.6 kbps or lower) (Broadband connections)
Experience Goods Simple content Virtual experience
Search Goods Virtual experience Virtual experience
As stated,
18
III HYPOTHESES
To summarize preceding discussion, consumers are likely to have more positive
attitudes toward products and Web sites under 3D environments than under simple 2D
environments as supported by many previous studies (Biocca, Li, Daugherty, & Chae,
2001; Hoch & Deighton, 1989; Klein, 2001; Hauble & Figueroa, 2001; Li, Daugherty, &
Biocca, 2001). However, these findings may not hold up in the field due to the current
Internet infrastructure. In addition, consumers’ attitudes toward products in 3D
environments may vary depending on the attributes of the products. Therefore, the
following hypotheses are formulated. Table 2 shows eight difl‘erent conditions based on
Connection Speeds (high vs. low), Product Categories (Search vs. Experience), and
Virtual Experience (3D vs. 2D).
Table 2: 2 X 2 X 2 Between Subject AN OVA design for the study
3D presentation of product 2D presentation of product
Experience Good Search Good Experience Good Search Good
Lower speed 1 2 5 6
Higher speed 3 4 7 8
l9
Hl: Consumers’ attitudes toward products are higher in the 3D presentation than in
the 2D presentation of products under high connection speeds (comparing
condition 3, 4 and 7, 8 in Table 2).
Hypothesis 1 was constructed to examine the consumers’ overall preferences of
virtual experience under the condition where there is no delay in downloading time. This
is more like a replication of previous virtual experience studies (Biocca, Li, Daugherty, &
Chae, 2001; Hoch & Deighton, 1989; Klein, 2001; Hauble & Figueroa, 2001; Li,
Daugherty, & Biocca, 2001) with an assumption of no traffic problems in the last mile
Internet infiastructure in the future.
H2: Added value of 3D presentation of products generally diminishes under lower
connection speeds (Consumers’ attitudes toward the 3D presentation of products
are lower under low connection speeds than under high connection speeds:
comparing condition 1, 2 and 3, 4).
Hypothesis 2 was constructed to examine if loading time delay affects the added
value of rich content. While no studies to date have demonstrated the negative effect of
20
loading time delays on a consumer’s attitudes toward products especially in 3D
environments or with rich content, it seems likely that the longer loading time may affect
a consumer’s perceived costs significantly. The added value of virtual experience likely
diminishes or may even result in being negative due to high perceived costs under lower
connection speeds, which may lead the consumers to abort the page load.
H3: Consumers’ preferences of the 3D presentation of products are higher in search
goods than in experience goods (Search goods are more benefited by virtual
experience than experience goods: comparing the difference between condition
<1, 3 and 5, 7> and <2, 4and 6, 8>).
Although the positive effects of virtual experience on product knowledge and
consumer attitudes toward products have been demonstrated by prior studies, the variable
effects of virtual experience across different product categories were not clearly indicated
in these studies. Search goods are more likely benefited by virtual experience due to their
search attributes, and experience goods are less likely benefited by virtual experience due
to their experience attributes. Even if a significant eflect of virtual experience on
consumers’ attitudes toward products is found in general, the amount of the effects would
21
r3... - . p
be different depending on product attributes. Therefore, the difference in consumers
attitudes toward products between 3D and 2D environments is likely larger in search
goods than in experience goods.
H4: Consumers prefer the 2D presentation of products to the 3D presentation of
products for experience goods under low connection speeds (comparing
condition 1 and 5 in table 2).
H5: Consumers prefer the 3D presentation of products to the 2D presentation of
products for search goods under low connection speeds (comparing condition 2
and 6).
In accordance with hypothesis 3, hypothesis 4 and 5 were constructed to
examine the relationship between loading time delay and virtual experience across
different product categories. Hypothesis 4 will be tested to see if perceived costs have
more significant effects on consumers’ attitudes toward products than willingness to pay
for experience goods. The (1 index for experience goods in Formula 1 might be so small
that the difference in willingness to pay caused by virtual experience is likely to be
22
smaller than the difference in perceived costs caused by Internet connection speeds. Thus,
the added value of rich content is more likely to be affected by Internet connection speeds
rather than by the richness of content itself. On the other hand, hypothesis 5 will be tested
to see if willingness to pay has more significant effects on consumers’ attitudes toward
products than perceived costs for search goods. The (1 index for search goods might be
so big that the difference in willingness to pay caused by virtual experience is likely to be
larger than the difference in perceived costs caused by Internet connection speeds. Thus,
the added value of rich content is more likely to be affected by the richness of content
rather than by Internet connection speeds.
H6: Consumers’ attitudes toward products are higher in the 3D presentation of
search goods under low connection speeds than the 2D presentation of search
goods under high connection speeds (comparing 2 and 8 in table 2).
Hypothesis 6 was constructed to examine the effect of virtual experience on a
consumer’s attitude toward search goods under different connection speeds. If hypothesis
5 is supported, it means that consumers prefer rich content in spite of loading time delays.
However, do rich content with loading time delays appeal better to consumers than
23
reduced content without loading time delays in B2C e-commerce? This is probably the
question that e-Retailers will want to know the most. It may upgrade the current BZC e-
commerce if e-Retailers provide dynamic and rich content regardless of a consumer’s
connection speed for a certain type of products.
24
IV METHODOLOGY
Regarch Design
A lab experiment was conducted to test the hypotheses since laboratory
experiments were commonly used in nearly all prior download delay and computer
response time studies, implying a causal model (Rose & Straub, 2001). Thus, a between-
subjects factorial analysis of variance design was used to test the hypotheses (See Table 2
and 3).
There are three independent variables rmnipulated for the experiment. First, the
operational definition for virtual experience was adopted from the work in Li, Daugherty,
and Biocca (2002). Three dimensional interactive was operationally defined as users’
ability to rotate, zoom-in and out the image of products for inspection. In turn, 2D static
was defined as the presentation of a static image of products for inspection.
Second, connection speeds with different download delay treatments were
defined as 5 seconds for a near-zero delay representing a high connection speed, 15
seconds for a delay for the 2D presentation of products under a low connection speed,
and 30 seconds for the 3D presentation of products under a low connection speed. The
delay interval of 15 seconds was adopted from previous delay studies (Rose, & Straub
25
2001; Rose, Evaristo, & Straub, 2002; Rose, Lees, & Meuter, 2001). A 5 second delay
was used for a near-zero delay condition for a realistic setting. A zero second delay could
be programmable, but it takes at least a few seconds to load Web pages even under most
broadband connections in real world. Although a 30 second delay was used as a moderate
delay in the previous delay study (Rose, Evaristo, & Straub, 2002), a 15 second delay
was used as a moderate delay in this study since the Internet is getting faster with
increasing broadband penetration. Besides, 15 seconds was about the average amount of
time to fully download the 2D version of the experiment Web site under a 56 Kbps dial-
up connection. Download delay for the experiment was programmatically controlled for
an accurate delay interval since delays could differ widely on the Web.
Last, product categorization was operationally defrned as search and experience
goods based on product attributes (Nelson, 1976; 1981). A camcorder was used to
represent a search good while a box of cereal was used as an experience good. Each
prodirct has the unique product attribute to represent either an experience or search good,
and thus seemed to be appropriate for the study.
Furthermore, consumers’ attitudes toward products served as the dependent
variable. Product attitudes were measured by asking participants how they felt about the
products using a proven six-item seven point semantic different scale common in
26
o . -_ _ ‘ u ' u
'1
'a “- ”.35,
advertising effectiveness measurement (Bad/Good, Not Appealing/Appealing,
Pleasant/Unpleasant, Unattractive/Attractive, Interesting/Boring) with 7 being the highest
and 1 being the lowest scale (Brunet, 1998).
Participants
A total of 232 undergraduate students enrolled at a major Midwestern university
participated in the experiment and were randomly assigned to eight different conditions
with 29 participants per condition (See Table 3). College students are seen as appropriate
in similar studies (Li, Daugherty, & Biocaa, 2002; Rose & Straub, 2001) because prior
computer knowledge and Web experience were required. In addition, college students
were shown to have the same attitudes and beliefs as compared to typical consumers
(Durvasula, Mehta, Andrews, & Lysonski, 1997; Rose & Straub, 2001).
The participants consisted of 157 males (67.7%) and 75 females (32.3%) with an
average age of 21 (SD = 2.69). Among 232 participants, 191 participants (82.3%) have
some sort of high speed Internet access at home. A large number of participants had
broadband access, which is way more than industry average (6.2 per 100 inhabitants:
OECD, 2002) since university dormitories provide Internet access via Ethernet LAN. The
average Internet usage of participants was 16.62 hours per week (SD = 14.47). Thus,
27
l . i as.» at n,
.k-A“
. l
participants were familiar with the Web and possessed required computer skills.
Stimulas
A fictional B2C e-commerce Web site selling tangible products was created in
both 3D and 2D environments. The Web site had four difierent conditions including the
3D presentation of an experience good (a box of cereal), the 2D presentation of the same
experience good, the 3D presentation of a search good (a digital camcorder), and the 2D
presentation of the same search good (See Figure 2 and 3 for screen shots). These 3D
rendered images were created for and used in the previous study of virtual experience (Li,
Daugherty, & Biocca, 2002).
To avoid any unnecessary biases, the information about the products and Web
site other than 3D and 2D presentations such as product descriptions, message appeal, or
the template of the Web site was held constant across all conditions. Therefore, the only
difference noticeable for the same product Web page was virtual experience with specific
interactive features such as the ability to rotate and zoom-in and out for the 3D Web page.
28
W” “““““l'
Figure 2: Screen shot of the experience good (a box of cereal)
The Web site was generated locally on an individual PC to programmatically
control loading time delays. The address bar was hidden to prevent participants from
noticing that the Web site was not generated from remote servers via the Internet. In fact,
exit interviews with 30 randomly selected participants confirmed that they believed to
have examined actual Web pages on the Web and no one found any unnatural settings or
weird experiences during the experiment.
29
Figure 3: Screen shot of the search good (a camcorder)
AG—Em
um Size mm Fern-108w
‘DEEE 'WWWMW
'IZXMIMM ”also-Ml.
‘ 6"
'zsmmurcownu "‘0me
- m Flow-mm
"who.”
‘MTEW‘M
~mmazmr-fia’ummm0-‘dp—m
'Zmflmrmhm
'mt’vm mun-ann-
HSK‘WCDRO.“ m
‘ch‘IflN—dIO—wmw :WW'Ouhfi
tint! on: [Am] MW|WGW
Huron”
'Imurogu Mm... M”. pawl-
‘DWSI WW umlrmvm
'Mownmmncwlflmwmmm
‘WM Maw-.1 mm
won-n, Al 9mmmrnan hwy-mum“
‘Iury Dram-cl: w .Mhn hm
‘% “’9‘”
Next
“1.31141"! .,. . . ‘. . ..... ""“‘f“ Ejfiy‘tonipméi" ‘ ~
HTML coding was used to make the accurate loading time delays and then to
redirect the viewers to go to the actual Web page containing product information. Figure
4 is the page, which participants viewed during the delay. Figure 5 is the actual coding
used to create a 30 second delay. An active loading bar was created in a Flash animation
to give viewers more realistic settings. The loading bar was filled in as the delay
continued.
30
Figure 4: Screen shot of the loading page with an active loading bar
apyn‘»|rvljlnfil5. and L‘mml Emerrence - Microsoft Internet E2-plorer
BI. Edit ylew Favorites Iuols nary ’ " $3
v.u.->.ro r: 35 “51.419 slayer ”1
”Each, :g'ru‘. 7777777530er flolosh _ items firSaarchir FiavoritesrfiMedia W History, W Mail W Print an DISCUSS ‘
LOADING
5.73114.
Figure 5: HTML coding for a 30 second delay
Procedures
Participants were given a large option of time blocks to choose from and were
asked to sign up for the date and time when they wanted to participate in the experiment.
The experiment continued for two weeks to collect enough data and give participants
more time flexibility. Once participants came to the laboratory, they were escorted by a
research administrator into a large laboratory where there were 20 personal computers
31
with seventeen inch monitors. Each participant was randomly seated at a computer
station corresponding to the 8 assigned conditions.
Before participants started to examine the Web site, they were given brief
instruction including how to navigate the Web site and the purpose of the study. An index
page was used to leverage their understanding of how to navigate the Web site, especially
for the virtual experience conditions (See Figure 6 for a screen shot of the index page).
Figure 6: Screen shot of the index page
. w: .. ', ..; ~. ,3 ;..-.--i.rii;.. - If‘i’v’vlfmit your
Edit ylow Favorites Iools‘ t1 , ‘ - . ‘7
i. a . o: o ~'a‘*~f“a' cal @5351; a g; E2]
' L”... -..-.E?f?!?.’§--_--- $99.-.. .5 [953... _ll°m°.--z‘_,,393'°h --F°V9'l‘°5-_.. 3391?..- Lilli!” l . 3” .31“! ; .-__" -_. 40‘9““
.. _,_.a_
a
.3 L..__..._.__._______...
i
l
l
“--._‘. _. .
',——-.--9--v~- n- 44"; .. - - r-a 7-..- --——-~c-- . a-“1-_--. .,-._'
'.'._ , . I . . .: . .-1a.,i,, .-
QSPARTANMAIL i- ’
5' . .
-_A I. . i_.m
Web Page Section Instructions: The following section asks that you view a series of two web pages and answer some
related questions.
The web pages contain content referring to brands of camcorder and cereal. Included among these are brands with which
you are likely familiar.
In the following section. before answering any questions about a specific page. please make sure the web page you are
evaluating has completely downloaded. Once the Web page has fully downloaded. please explore the web page as much
time as you want before answering questions. You can rotate and zoom in/out the 3D rendered image of the product by
using your mouse (hold down left button and move your mouse to rotate the product and use roll-uo/down for zoom
in/out).
Next step: Please click on 'Next‘ button in the browser. This will open the web page. which you are to evaluate for this
research. When that page has finished loading. please explore the web page thoroughly and then. answer the
questionnaires by clicking 'Next‘ button.
Next
’ L . w|. .
,. ”iv . _, . . . , ‘ _ ... _ .. . 'i «i - ‘D M c p . 1 ‘
. > - . . ‘ _ ~ ,. , . - - . . - , \ .
-‘£.".. ... --. ...J..,. ,-._ '.- I." . . ;. -.. r .z» -w . of. {pr-an .n ‘2. ) 1.‘ .i...-r . .l.:‘. ».h. " ‘~.._.-..ll _-l Ll - ll . .y..-°m ute'........-~... .:
’I ’ ‘Ja'v‘v'l’r - n““'" y"-"."" -.' W" "m-" In. -~ ‘ ""q‘0fl ' f'"‘ - ‘K ”I I ~ " -‘r "-‘v ‘f' ‘- ‘ \
“El
1 I
J
32
A next button was used to go to the next step or page for easy navigation.
Participants were instructed that they needed to wait to evaluate the Web site until the
page was fully loaded. Participants with low connection speed conditions were told that
their Web connection was dial-up because of network problems in the laboratory.
Participants did not know different conditions and were not allowed to talk to
each other. Participants were told to take as much time as they wanted to explore the Web
site, and then to click the next button after they finished examination, Finally, participants
were told that upon finishing their examination, they would be asked to complete a
survey with 15 questions to record their evaluation (See Appendices for questionnaires).
33
V ANALYSIS
Attitudes toward products were analyzed in a Connection Speed (low vs. high) x
Virtual Experience (3D vs. 2D) x Product Categorization (Search vs. Experience)
between subjects factorial analysis of variance to test the hypotheses. The results are
summarized in Table 3.
Table 3: Overview of descriptive statistics for each experimental condition
Virtual Experience
3D presentation
2D presentation
Experience Search Experience Search
Good Good Good Good
Condition 1 Condition 2 Condition 5 Condition 6
011:29 011:29 013:29 011:29
Low
. M=4.989 . M=5.592 o M=4.155 . M=5.l90
. . SD=.955 . SD=.719 . SD=1.111 . SD=1.031
Connection
Speeds Condition 3 Condition 4 Condition 7 Condition 8
on=29 on=29 on=29 .n=29
High
. M=4.943 . M=5.414 . M=4.213 . M=4.644
. SD=1.174 . SD=.804 . SD=1.196 - SD=.935
34
In general, the results showed a significant main eflect of virtual experience, E
2
(l, 224) = 26.94, p < .01, ‘l = .12, across different product categories, E (l, 224) = 23.23,
2
p < .01, 11 = .10. Participants were more likely to have higher attitudes toward products
in 3D environments (M = 5.2342, _S_Q = .9581) than in 2D environments (M = 4.5503, SD
= 1.1375). However, there was not a significant main effect of connection speeds, F (l,
224) = 1.83, _ns.; possible interpretations of which will be discussed later.
In addition, the me number showed 2.767, which did not exceed the
generally accepted guideline of 3. Three is roughly the point beyond which heterogeneity
of variance begins to inflate the actual a level beyond acceptable limits (Keppel, 1991).
Thus, the homogeneity of variance assumption has not been violated in this study.
Furthermore, the results of the hypothesis tests are summarized in Table 4.
Hypothesis I predicted participants’ attitudes toward products would be higher
in 3D environments than in 2D environments under high connection speeds with the
assumption of no delay in the last mile Internet infrastructure. In accordance with a
significant main effect of virtual experience, hypothesis 1 was supported in the study.
Specifically, participants were likely to have higher attitudes toward products in 3D
presentations (M = 5.178, _S_Q = 1.025) than in 2D presentations (M = 4.4253, SQ =
1.091), F (l, 224) = 16.20, p < .01, under the high connection speed.
35
Table 4: Results of the hypothesis tests
Short Descriptions of Hypotheses F P < Supported?
The sitive effects of virtual ex rience 3D
H1 p0 pe ( ) 16.20 .01 Yes
under high connection speeds
HZ The added value of 3D generally diminishes under
. .361 ns. No
low connection speeds
Search oods are more benefited b virtual
H3 g , y .55 ns. No
experience than experience goods
Consumers prefer 2D environments under low
H4 _ _ 10.0 .01 No
connectron speeds for experience goods
Consumers prefer 3D under low connection
H5 2.33 ns. No
speeds for search goods
Consumers prefer 3D environments under low
H6 connection speeds to 2D environments under high 12.95 .01 Yes
connection speeds for search goods
Hypothesis 2 predicted the negative effects of download delays on consumers’
attitudes toward the 3D presentation of products under low connection speeds. The
results of the hypothesis test found no significant difference in consumers’ attitudes
toward the 3D presentation of products between the low connection speed (M = 5.292,
.—
36
D = 0.8916) and the high connection speed (M = 5.1782, SD = 1.025), E (1, 224) = .36,
n_s_. As a result, it seems that there is no support of a significant relationship between
increases in delay on the Web and changes in attitudes toward the 3D presentation of
products.
Hypothesis 3 predicted that the difference in consumers’ attitudes between the
3D and 2D presentation of products would be larger in search goods than in experience
goods since search attributes are more likely to be enhanced by such visual effects.
However, the result showed no interaction effects between virtual experience and product
categories, 13 (1, 228) = .55, ns. Specifically, the difference of mean attitude between the
3D presentation (M = 4.9655, SD = 1.06) and the 2D presentation (M = 4.1839, _S_12 =
1.144) of the experience good was .7816. On the other hand, the difiemme of mean
attitude between the 3D presentation (M = 5.5029, SQ =.7612) and the 2D presentation
(M = 4.9167, SD = 1.014) was .5862. The difference in consumers’ attitudes between the
3D and 2D presentation of products was even larger in the experience good than in the
search good unlike the hypothesis although the difference was not statistically significant.
Hypothesis 4 predicted that consumers’ attitudes would be higher in 2D
presentations than in 3D presentations for experience goods under low connection speeds.
Although the result of the hypothesis test found a significant difi‘erence in participants’
attitudes between the 3D and 2D presentation of the experience good under the low
37
connection speed, E (1, 224) = 10, p < .01, the finding was quite opposite to the
hypothesis. Specifically, participants were likely to have higher attitudes in the 3D
presentation (M = 4.989, SD = .955), than in the 2D presentation (M = 4.1552, SD =
1.111) of the experience good under the low connection speed contradictory to hypothesis
3. Ironically, this opposite finding is quite natural and can be expected with the absence
of the main efl‘ects of connection speeds or loading time delays.
Hypothesis 5 predicted that consumers’ attitudes would be higher in the 3D
presentation than in the 2D presentation for the search good under the low connection
speed. The results showed a non significant difference in participants’ attitudes between
the 3D and 2D presentation of the search good under the low connection speed, E (1, 224)
é 2.33, gs. Although participants’ attitudes were higher in the 3D presentation (M =
5.592, §_I_)_ = .719) than in the 2D presentation (M = 5.190, SD = 1. 031) as anticipated,
the finding was not statistically significant.
Hypothesis 6 predicted that consmners’ preferences of 3D presentation would be
high for search goods regardless of connection speeds. The results of the hypothesis test
found a significant difference in participants’ attitudes between the 3D presentation of the
search good under the low connection speed and the 2D presentation of the search good
under the high connection speed, E ( 1, 224) = 12.95, p < .01. Specifically, participants
38
were likely to have higher attitudes in the 3D presentation of the search good under the
30 second delay condition (M = 5.592, SD = .719) than in the 2D presentation of the
search good under the no delay condition (M = 4.6437, _S_D = .935).
Possible explanations for the findings as well as limitations of this study will be
discussed in next chapter.
39
VI DISCUSSION AND LIMITATIONS
The value of a scientific finding of no effect can be just as good as an expected
significant outcome in a sense that non-published insignificant findings may cause a
distorted view of research questions, which is called a “file drawer” problem (Rosenthal,
1979). However, there may be possible explanations or interpretations for non significant
findings in the study, which are briefly summarized in Table 5.
Hyp_othesis 1
In accordance with many previous virtual experience studies, participants’
attitudes were higher in 3D than in 2D presentations under the high connection speed. A
possible interpretation of this finding is that consumers would prefer rich content in B2C
e-commerce if there were no download delays. If this finding is true, then e-Retailers
need to develop strategies for adding more relevant, interesting, and informative content
on their Web sites. However, no delay condition was assumed in the hypothesis 1 without
considering the possible impacts of loading time delays. Therefore, this assumption
should be kept in mind when interpreting the result of the hypothesis 1 test.
40
Table 5: Possible explanations for the findings of this study
Findings
Interpretation of the findings
H1
Higher attitudes in 3D than in
2D presentations under the
high connection speed
Consumers would prefer rich content in
B2C e-commerce if there were no download
delays.
e-Retailers need to develop strategies for
relevant and sales inducing content
No difierence in participants’
preference of 3D presentations
between the low and high
connection speed
Loading time delay has no impact on
consumers’ preference of 3D
No lingering effects from loading time
delays on e-Retailer success once the page
has loaded without being aborted
Opportunity costs were close to zero in the
experiment, thus the added value was
affected only by willingness to pay
Delay interval of 15 seconds was not
appropriate to measure the effects of loading
time delays
H3
No significant interaction
effect between virtual
experience (3D vs. 2D) and
product category (Search vs.
Experience goods)
Consumers prefer the 3D presentation of
products regardless of product categories
Consumers’ willingness to pay to get visual
effects by 3D environments is different
between search and experience goods
No interaction effect between virtual
experience and product category without the
effect of connection speeds
41
Findings
Interpretation of the findings
H4
Participants preferred the 3D
to 2D presentation of the
experience good under the low
connection speed
Consumers prefer the 3D presentation of
products regardless of loading time delays
and product categories
interpretations as
Same suggested in
hypothesis 2 can be applied
H5
No difference in participants’
attitudes between the 3D and
2D presentation of the search
good under the low connection
speed
Sample size was not big enough since there
was a difierence found in attitudes but it
was not just statistically significant
Consumers also got necessary information
about the camcorder from search attributes
other than visual images
Perceived cost was increased due to loading
time delay but willingness to pay was also
increased for 3D presentation, thereby
causing no difierence in attitudes
H6
Participants preferred the 3D
presentation of the search
good under the low connection
speed to the 2D presentation
of the search good under the
high connection speed
Consumers prefer the 3D to 2D presentation
of search goods in spite of loading time
delays in 3D environments and no delay in
2D environments
The difference in attitudes was mainly due
to the positive effects of virtual experience
rather than combined interaction effects
42
Hypothesis 2
Similarly, Rose and Straub could not find the relationship between download
delays and attitudes toward the retailers in their study (Rose & Straub, 2001). There are 4
plausible explanations for the result of the hypothesis 2 test including the one suggested
in the work by Rose and Straub. The first possible interpretation is that loading time
delays do not have any impact on consumers’ attitudes toward products. If so, then e-
Retailers need to change their strategies and add relevant, sale-inducing content without
counteracting the negative impact of delays on the Web by reducing content. However,
anecdotal evidence and previous studies (Rose, Lees, & Meuter, 2001; Rose, Khoo, &
Straub, 1999) showed the negative impacts of loading time delays in B2C e-commerce as
indicated earlier in the paper. Therefore, it would not be a good idea for e-Retailers to
create the Web site without balancing content and its loading time.
The second interpretation is that there may be no lingering effects fiom loading
time delays on e-Retailer success once the page has loaded without being aborted as
suggested in the previous study (Rose and Straub, 2001). It is possible that consumers
may not form attitudes toward products based on their experience on loading time delays
once the page has been successfully loaded. If so, then e-Retailers need to convince
consumers to wait until the page has been loaded successfully by giving them incentives
43
such as e-coupons or credit points, etc. In this case, it is essential to find out the exact
time point when visitors on the Web start to abort the page load, not to mention finding
many incentives, which appeal to visitors in future studies.
The third interpretation is to use the added value of rich content model
suggested in the study. The added value of rich content was defined as (1 times
willingness to pay minus B times opportunity cost where a is related to product
categories and B is related to Internet connection speeds. Given the economic model, it
is possible that participants’ opportunity costs could be close to nearly zero in a
laboratory setting. By the time participants had signed up for the experiment, they already
decided to devote a certain amount of their time for the experiment. Thus, they might not
consider other altematives to do during the experiment, which could cause zero
opportunity cost for them. Internet connection speeds could not affect perceived costs if
opportunity cost is nearly zero no matter how much difference the B index shows. If so,
then perceived costs become also nearly zero. Therefore, the added value of rich content
is pretty much determined by willingness to pay only.
If this interpretation is correct, then the remedy is to replicate the experiment
with some options for participations to have an ability to abort the page load or similar.
Although participants were allowed to quit at any time during the experiment in the study,
44
nobody actually quit. Thus, it is hard to say that aborting the page load option was
provided in the study. Therefore, at least two options of choice for participants with brief
instruction of what they are to see and how long it may take to load the chosen page are
suggested for the future study. If a participant chose to see the Web page that had a 3D
presentation version with a 30 second delay over a 2D presentation version with a 15
second delay, it would show the participant’s preference clearly. It is also suggested for
the firture study that options for aborting the page load and switching to the different
version of product presentations during delays need to be provided. It would improve the
reliability of the effects of product attributes if a series of different products were
presented to a participant, and then to see how the participant chooses different options
for each product.
The fourth interpretation for hypothesis 2 is that delay interval might not be
appropriate to measure the effects of delays. A 30 second delay may not be slow enough
to cause negative impacts on participants’ attitudes toward products. However, the results
of attitudes toward delay measure showed a significant difference between the no delay
conditions and the 30 second delay conditions, F (1, 224) = 40.28, p < .01 (See Table 6).
Specifically, participants were likely to have more positive attitudes toward delay in the
no delay conditions (M = 3.3707, SD = .6919) than in the 30 second delay conditions (M
45
= 2.569, SD = .7972). The attitude toward delay measure was adopted fiom previous
studies (Hui & Tse, 1996; Rose, Lees, & Meuter, 2001) with a four-point scale as
intolerable delay being number 1, excessive but still tolerable delay, acceptable delay, and
no significant delay being number 4 (See Appendices). Thus, the delay seems to be
appropriate, and this interpretation cannot be applied to the result of hypothesis 2 test.
Table 6: Attitude toward delays
No delay 15 second delay 30 second delay
Attitude toward . M = 3.3707 . M = 2.6724 . M = 2.569
delays . SD = .6919 . SD = .9250 . SD = .7972
Hmthesis 3 and 5
There are two plausible interpretations other than no significant interaction
effects for the result of the hypothesis 3 test. The first is that willingness to pay for 3D
presentations may vary depending on product categories. I assumed that willingness to
pay to get rich information would be the same regardless of product attributes. However,
participants might obtain good information for their decision on the camcorder from other
search attributes such as price, technical specifications, or size, since visual effects by 3D
presentation is one of the search attributes. If so, then participants’ willingness to pay for
46
the 3D presentation of the search good could be lower than willingness to pay for the 3D
presentation of the experience good. Thus, the added value of 3D presentations was
smaller than expected in search goods probably due to the reduced willingness to pay for
visual effects, in spite of relatively higher (1 index for search goods. Relatively lower
willingness to pay and higher 0. index might negate each other, and thus likely to result
in not much difl'erence in participants’ preference of 3D presentations between search and
experience goods assuming no difference in perceived costs.
If this interpretation is true, it can be also applied to the result of the hypothesis
5 test. Participants received necessary information for their decision about the camcorder
from other search attributes in the 2D presentation of the product, which might cause
relatively high attitudes even in 2D presentations. Therefore, the difference in attitudes
was likely to be insignificant between 3D and 2D presentation of the search good. Other
interpretation could be small sample size. The remedy for small sample size is to
replicate the experiment with larger sample size.
No significant difference between the 3D and 2D presentation of the search good
under the low connection speed could also be considered as natural and significant if
there were effects of loading time delays found in the study. Perceived cost could be
increased due to loading time delays while willingness to pay could be increased for 3D
47
presentation. As a result of the increase in both willingness to pay and perceived cost, the
final added value of 3D presentation could remain as before, thereby showing no
difference in attitudes between the 3D and 2D presentation of the search good.
Hypothesis 4
\Vrthout the main effect of delays, the result of the hypothesis 4 test could be
considered as natural and consistent with other findings. Only willingness to pay affects
the added value of 3D presentation when the impact of connection speeds is ignored.
Therefore, consumers are likely to have higher attitudes toward the 3D than the 2D
presentation even for experience goods.
Hypothesis 6
It would have been a significant finding if there were a main effect of connection
speeds since hypothesis 6 was supported. However, the significant difference found in the
result of the hypothesis 6 test is more likely due to the effects of 3D presentations rather
than due to interaction effects among connection speeds, virtual experiences, and product
categories since loading time delays showed no significant impacts on participants’
attitudes in this study. If there were a significant main effect of delays, then this finding
48
will help e-Retailers do more targeted marketing with dynamic versions of Web services
and give a clearer understanding of how to balance their content and loading time delays.
Other consideration and implication for future studies
One more thing needed to be discussed is the main effects of product category.
Although participants were likely to have higher attitudes toward products when a search
good was presented (M = 5.29, SD = .9398) than when an experience good was presented
(M = 4.5747, SD = 1.166), E (1, 224) = 23.23, p < .01, 712 = .10, it is hard to suggest
that the main effect of product categorization was found. Each product has the unique
product attribute to represent either an experience or search good. However, these two
products are not much equivalent in terms of price, or the level of consumers’
involvement. The price gap between a camcorder and a box of cereal was large enough to
cause the difference in participants’ attitudes, not to mention that college students usually
show higher interest in electronics than cereal. Thus, the difference in participants’
attitudes found in the experiment is likely due to the difference between a camcorder and
a box of cereal rather than product attributes they are to represent.
Besides, participants’ pre-existing experience with the products could affect the
results since both products are well known brands with relatively high reputations. This
49
potential prejudice from real experience could affect the results although the questions
were designed to ask how the Web site made participants feel about the products.
Therefore, it is suggested to use a series of non-existent products, which are
equivalent in terms of price, size, or the level of involvement in future studies. Having a
participant examined a series of different products could measure the efiect of product
categories more accurately. A Latin-Square design, randomly changing the sequence of
presentation order, should be used to prevent practice efi‘ects or carry-over effects in this
case.
For future study, I would like to suggest having more levels in connection speeds
with a smaller delay interval, having a series of different products with randomly changed
sequence of presenting orders, having abort-like options to choose from with brief
instruction of what participants are to see and how long it may take, and providing the
actual ability to abort the page load and switch to the different version of product
presentation during delays. I also would like to suggest adding more dependent variables
to indicate other possible areas rather than attitudes toward products where the factors
may affect BZC e-commerce for the future study.
50
Limitation
A limitation for this study is the experiment conducted in a laboratory setting
using college students as samples. Since this type of experiment restricts the external
validity, this limitation should be kept in mind when interpreting the results.
51
VII CONCLUSION
The purpose of this study was to examine the relationships among virtual
experience, connection speeds, and product attributes for a clear understanding of how to
balance the Web page content and loading time delays across different product categories
in B2C e-commerce. Without clear understanding of the added value of rich content on
the Web, both e-Retailers and e-consumers cannot appreciate the maximum benefits of e-
commerce using the Internet as a powerfirl medium.
The positive effects of virtual experience created by 3D environments on
product knowledge and consumer attitude toward the products and the negative impacts
of excessive download delays on B2C e-commerce seem to be a package deal under the
current Internet infrastructure. Unfortunately, since the consequences of delay, as well as
the impact of delay on the added value of rich content are still unclear, e-Retailers may
not be able to manage their Web sites effectively by adding more relevant and sales-
inducing content for targeted marketing. Hopefully, the findings of this study and
suggestions for future studies will address guidelines for a clear understanding of the
added value of rich content in B2C e-commerce.
52
If the impacts of delays on the added value of rich content in B2C e-commerce
are found in future studies, then e-Retailers can provide more personalized and targeted
Web services based on a consumer’s choice of products and their connection speeds
dynamically. This could maximize willingness to pay and minimize perceived costs in
Formula 1 at the same time, thereby maximize the added value of rich content.
In fact, many Web sites already provide visitors with options to choose from
text-only Web pages to more graphic oriented Web pages. But since technologies evolve
so fast, if server technologies were smart enough to detect a visitor’s connection speed
automatically, then e-Retailers would not even need to provide such options but could
dynamically provide consumers with different versions of product presentation and
relevant, sale-inducing content.
F ast-evolving technologies will help e-Retailers achieve their goal of providing
dynamic versions of Web pages, once they understand how to balance content and its
loading time clearly. In this way, e-Retailers could differentiate their e-commerce fiom
other e-Retailers, thereby standing out in competitive BZC e-commerce environments
without facing a price war.
53
APPENDICES
54
Survey Questionnaires
Indicate your level of agreement with the following statements by circling the most
appropriate number (question 1 through 3).
Strongly Strongly
Disagree Aggee
1. I feel very knowledgeable about the product I just
examined. ................................................. 1 2 3 4 5 6 7
2. If I had to purchase the product today, I would
need to gather very little information in order to
make a wise decision. ................................... 1 2 3 4 5 6 7
3. I feel very confident about my ability to judge the
quality of this product. .................................. l 2 3 4 5 6 7
55
For each of the items below, circle the number that best describes your overall feelings
about the product you have evaluated (question 4 through 9).
4. Bad 1 2 3 4 5 6 7 Good
5. Unappealing 1 2 3 4 5 6 7 Appealing
6. Unpleasant 1 2 3 4 5 6 7 Pleasant
7. Unattractive 1 2 3 4 5 6 7 Attractive
8. Boring 1 2 3 4 5 6 7 Interesting
9. Dislike 1 2 3 4 5 6 7 Like
56
10. Please indicate your attitude toward the download time delay for the page you just
viewed (choose one):
[ ] Intolerable delay
[ ] Excessive but still tolerable delay
[ ] Acceptable delay
[ ] Not significant delay
11. What is your PID
12. What is your sex? Male [ ] or Female [ ]
13. What is your age? Years (optional)
14. How many hours in a typical week do you spend on the Internet? Hour
15. If you have Internet connection at home, please indicate your connection speed:
Dial-up connection with 64k or lower[ ] High Speed [ ]
(LAN, Cable Modem or DSL, etc.)
57
BIBLIOGRAPHY
Amor, Daniel (2000), “The E-Business (R)evolution,” NJ: Prentice Hall PTR.
Biocca, Frank, Terry M. Daugherty, Hairong Li, and Zoo-Hyun Chae (2001), “Effect of
Visual Sensory Immersion on Presence, Product Knowledge, Attitude toward the
Product, and Purchase Intention,” Paper read at 2001 Experiential E-Commerce
Conference, September 27 - 28, at East Lansing, MI.
Brandenburger, Adam M, and Harborne W. Stuart, Jr. (1996), “Value-Based Business
Strategy,” The Massachusetts Institute of Technology, Journal of Economics &
Management Strategy, Volume 5, Number 1, Spring 1996, 5 -— 24.
Bruner, Gordon C. (1998), “Standardization & Justification: Do and Scales Measure
Up?,” Journal of Current Issues and Research in Advertising 20(1): 1 -1 8.
Department of Commerce (2000), “Digital Economy 2000 report.” [online] Available:
http://www.esa.doc.gov/de2000.pdf.
DTI (1998), “Net benefit: The electronic commerce agenda for the UK.” [online]
Available: http://www.dti.gov.uk/public/fi'ame8.html,
Durvasula, S., S.C. Mehta, J.C. Andrews, and S. Lysonski (1997), “Advertising Beliefs
and Attitudes: Are Students and General Consumers Indeed Different?,” Journal of
Asian Business (13:1), pp. 71-84.
Edwards, Steve and Harshavardhan Gangdharbatla (2001), “Separating the Wheat fi'om
the Chafi: Understanding the Novelty of Interacting with 3D Products and the
Implication of Wearout for e-Commerce,” Paper read at 2001 Experiential E-
Commerce Conference, September 27 — 28, at East Lansing, MI.
European Initiative in Electronic Commerce (1997), “Communication to the European
Parliament, the Council, the Economic and Social Committee and the Committee of
the Regions.” [online] Available: http://www.cordis.lu/esprit/src/ecomcomhtm.
58
Gibson, J.J. (1996), “The Senses Considered as Perceptual Systems,” Boston: Houghton
Mifflin,
Haubl, Gerald and Pablo Figueroa (2001), “Virtual-Reality Product Demos and Buyer
Behavior in Electronic Shopping Environments,” Paper read at 2001 Experiential E-
Commerce Conference, September 27 — 28, at East Lansing, MI.
Hoch, Stephen J. and John Deighton (1989), “Marketing In Hypermedia Computer-Based
Environments: Conceptual Foundations,” Journal of Marketing; 50 - 68.
Hoch, Stephen J. and Young-Won Ha (1986), “Consumer learning: Advertising and the
ambiguity of product experience,” Journal of Consumer Research, 13; 221-233.
International Telecommunication Union (1999), “CHALLENGES TO THE NETWORK:
Internet for Development.” [online] Available: http://www.itu.int/ti.
Keppel, Geofli’ey (1991), “Design and Analysis: a researcher’s handbook —— 3rd ed.,”
Prentice-Hall, Inc., p220.
Klein, Lisa R (2001), “Understanding Consumer Response to Virtual Product
Experiences: The Role of Telepresence,” Paper read at 2001 Experiential E-
Commerce Conference, September 27 — 28, at East Lansing, MI.
Hobley, Christopher (2001), “Just Numbers: Numbers on Internet use, electronic
commerce, IT and related figures for the European Community,” Published by the
European Commission’s Electronic Commerce Team (Information Society
Directorate General).
Hui, Michael K. and David K Tse (1996), “What to Tell Consumers in Waits of Different
Lengths: An Integrative Model of Service Evaluation,” Journal of Marketing, vol.
60, no. 2, pp. 81 - 90.
Lake, David (2001), “Quick and Easy,” The Industry Standard, 5 March. [online]
Available: http://www.thestandard.com/article/display/O,1 15 1 ,22342,00.html.
59
Li, Hairong, Terry M. Daugherty, and Frank Biocca (2001), “Consumer Learning and 3-D
Ecommerce: Sequential Exposure to a Virtual Experience Relative to Indirect and
Direct Product Experience,” Paper read at 2001 Experiential E-Commerce
Conference, September 27 — 28, at East Lansing, MI.
Li, Hairong, Terry M. Daugherty, and Frank Biocca (2002), “Effectiveness of 3-D
Visualization: The Role of Presence,” The Journal of Advertising, Special Issue on
new Media and Advertising.
McCabe, Deborah Brown and Stephen M. Nowlis (2001), “Information Integration
Across Online and Offline Shopping Environments and Its Effect on Consumers’
Purchase Decisions,” Paper read at 2001 Experiential E-Commerce Conference,
September 27 - 28, at East Lansing, MI.
Millar, Murray G. and Karen U. Millar (1996), “The Effects of Direct and Indirect
Experience on Affective and Cognitive Responses and the Attitude - Behavior
Relation,” Journal of Experimental Social Psychology, 32: 561 — 579.
Nash, Sharon (2000), “Online Catalogs Will Come To Life And You’ll View
Merchandise From Every Side,” PC magazine. [online] Available:
http://www.zdnet.com/pcmag/stories/trends/O,7607,2418240,00.html.
Nelson, Phillip (1976), "Economic Value of Advertising,” Advertising and Society, NY:
New York University Press.
Nelson, Phillip (1981), “Consumer Information and Advertising. in Malcolm Galatin and
Robert D. Leither (Eds.),” Economics of Information, Boston: MA: N. Nijhoff
Publishers.
OECD (1997), “Policy Brief, No. 1 on Electronic Commerce.” [online] Available:
http://www.oecd.org/pub1ications/Po l__brie f/ 970 l _Pol.htm.
OECD (2001), “The Development of Broadband Access in OECD Countries, Paris.”
[online] Available: http://www.oecd.org/pdf7M00020000/M00020255.pdf.
60
OECD (2002), “Broadband access in OECD countries per 100 inhabitants, September
2002.” [online] Available: http://www.oecd.org/EN/document/0,,EN-document-O-
nodirectorate-no-4-39262-0,00.html
Parkin, Michael (1993), “Microeconomics —2nd ed,” Addison- Wesley Publishing, 10 — 11.
Pedersen, Mark (2001), “PERFORMANCE METRICS IN CONTEXT,” Go’mezPro Brief
[online] Available: http://a520.g.akamaitech.net/7/520/915/a2687de89900ca/
images.gomez.com/images/Networks_marketing/MetricsinContext.pdf.
Rose, Gregory M., Roberto Evaristo, Detmar Straub (2002), “Culture and Consumer
Responses to Web Download Time; A Four-Continent Study of Mono- and
Polychronism,” Special Issue of IEEE Transaction on Engineering Management.
Rose, Gregory M., John Lees, and Matthew L. Meuter (2001), “A Refined View of
Download Time Impacts on E-Consumer Attitudes and Patronage Intentions
Toward E-Retailers,” JMM — vol. 3. [online] Available:
http://www.mediajournal.org/netacademy/publications.nsf/all_pk/1946.
Rose, Gregory M, Huoy Khoo, and Detmar Straub (1999), “Current Technological
Impedirnents to Business to Consumer Electronic Commerce,” Communications of
the AIS, vol. 1.
Rose, Gregory M., Detmar W. Straub, John D. Lees (2000), “The Eflect of Download
Time on Consumer Attitude Toward the Retailer in E-Commerce,” Proceedings of
the Sixth Americas Conference on Information Systems, August, pp. 1352-1354.
Rose, Gregory M., Detmar W. Straub (2001), “The Efiect of Download Time on
Consumer Attitude Toward the e-Service Retailer,” e-Service Journal, vol. I, no. I.
Rosenthal, R. (1979), “The ‘File Drawer Problem’ and Tolerance for Null Results,”
Psychological Bulletin, pp. 638 — 641.
Samsung Economic Research Institute (2000), “World Report 3” [online] Available:
http://www.seri.org.
61
Schmitz, Stefan W. (2000), “The effects of electronic commerce on the structure of
intermediation,” JCMC. [online] Available:
http://www.ascusc.org/jcmc/vo15/issue3/schmitz.html#re£
Surmacz, Jon (2002), “Broadband Appeal: Home subscribers vault to 24 million.”
[online] Available: http://www2.cio.com/metrics/2002/metric392.html
Tchong, Michael (2000), “More concentrated than the leading brand,” ICONOCAST
[online] Available: http://www.iconocast.com/issue/macroview.
WebCriteria (1999), “Max and the Objective Measurement of Web Sites,” [online]
Available: http://www.inetz.com/work/pdf/objective_measurement.pdf.
62
lllilllilliilllll
_L‘- ‘