fi- .- , h. . . ... . {a
2.. 5.... .. ..:.
. 5 «..:... awn”?!
..: .. ”5.5.? ... 7.1....
. .2»... .._............n!. .. 3...
. .1! $11... _\ . .
, . , . W:
u.......u..,:.,n£ . 598». G . ..
new.” umvu .f .. .
. Monaurkw . I
. . . . : . ..u. 4.... .. ..
. . I ..: o L,
. . y» 09 1A. .
:2: . . Raw. .1 .
. “Haw WJ"!I¢.. .. . . ‘14 I,
u

t 3... "..:. z.

I ) .ll of
2.: 2.4.4... .. 2. n...
.

. Jwr I.

i .. pl! «

. . c. .

. win

, . . . N...“ ,2

. . . , . , . to. . .
. . , x...- . 7. . .

. f... i. . , .

a. ..:...I.

. :.
$.65 ....... y

.mwfi. . u

.355;
Ital}!
n... In! .
.: \1
.15... ...
I... .3!

. II. is 9. ..: w.
. z 7!... 1:: a. H.755}?
r... ... .. . . r. IV.

5..
ti. ’3. 6:! .5
53g." 2.0%.... I}? .
p’ltflr . . fl
. n

h.

 

.........u
3......

7.. » JV.“

2...... gram 3&5... wtuqrsjauhh
(a 3L...“ L912... (.160 ..1

.2 .11 \
4.. . ire-1521.2. .3.‘
3.5:...6} 3 13...)...
’35. It»: 3')

.’
v‘HIOND .
0.5: v.

, 1.5.?! ‘.

.
WWW.” Oahu... -

 

 

 

 

THESIS

2 LIBRARY
1% Michigan State
University

This is to certify that the
dissertation entitled

A DUAL PROCESSING MODEL OF VIRTUAL EXPERIENCE

presented by

Ki-Young Lee

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Media and Information Studies

 

 

/

 

Maior Pr<)/fe70r'sé Signature

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLEE wigaflfi dfldate if requested.

 

 

 

 

 

DATE DUE DATE DUE DATE DUE
fi ‘ .
9.3512157 2037 20.8 0.1 92,
0 Wilfimmg 1
All? i. @2909

 

 

 

 

 

 

 

 

 

 

 

2105 p'JCIRC/DateDuejndd-pJ

 

A DUAL PROCESSING MODEL OF VIRTUAL EXPERIENCE
By

Ki-Young Lee

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Advertising

2006

ABSTRACT
A DUAL PROCESSING MODEL OF VIRTUAL EXPERIENCE
By

Ki-Young Lee

This study proposed and tested a dual-processing model of virtual experience in
which discursive processing and experiential processing modes constitute two distinct
routes of virtual experience as they relate to attitude accessibility and attitude confidence.
It also examined the boundary condition in which each processing mode is more likely to
operate, by specifying product type (utilitarian vs. hedonic) as a possible moderator for
relations between the two processing modes and attitude accessibility, and attitude
confidence. A two-factor mixed experiment (N = 227) was conducted, with presentation
mode as the between-subject factor (2-D or 3-D) and product type (utilitarian or hedonic)
as the within-subj ect factor. Data were analyzed using AN OVA, AN COVA, and latent
variable structural equation modeling (SEM). The results of SEM analyses suggest that,
as expected, the two processing modes mediated the effects of virtual experience, as
simulated in 3-D product visualization, to attitude accessibility and attitude confidence.
Also the relative influence of each processing mode on attitude accessibility and attitude
confidence was found to be affected by product type. However, decomposition of
significant effects suggests that regardless of product type, experiential processing played
a dominant role in differences observed in the two dimensions of brand attitude strength.
The results are discussed in light of research on interactive advertising and consumer

information processing.

Copyright by
KI-YOUNG LEE
2006

ACKNOWLEDGMENTS

I am greatly indebted to those who made this dissertation possible.

I thank Dr. Hairong Li, my advisor, for his direction, unflinching support, keen
insight and hard work, all of which inspired me — and in no small way enabled me — to
complete this project. I also thank my fabulous committee members, Drs. Frederick Fico,
Frank Biocca, and Steven M. Edwards, for their investment of time, energy and guidance
in this project, and for giving me the opportunity to work with them during my time at
Michigan State University. I will always remember fondly these fine experiences.

My special thanks go to Young-Seock Cha for his technical assistance. He willingly
sacrificed his valuable time to develop reaction time programs, without which this
dissertation would still be far from completion. I also thank my friends in both Korea and
the United States, and my MSU colleagues, for their friendship and support during my
study.

This dissertation would not have been made possible without the love, support, and
understanding of my wife, Sung-Hee J 00. I cannot find good words to express how
thankful I am to her, who has always been there for me as a wonderful critic, colleague,
and source of everyday happiness.

Finally, I dedicate this dissertation to my parents, J in-Ho and Suk-Sang Lee, for all
their encouragement and countless prayers over the years. I cannot thank them enough for

their love and sacrifices for me.

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................ vii
LIST OF FIGURES ....................................................................................................... ix
CHAPTER 1. Introduction ............................................................................................... 1
CHAPTER 2. Literature Review ..................................................................................... 6
2.1. Direct, Indirect, and Virtual Experience ........................................................... 6

2.2. Virtual Experience and Attitude Strength ......................................................... 8

2.3. Attitude Accessibility and Attitude Confidence ............................................. 10

2.4. Discursive and Experiential Processing .......................................................... 12

2.5. Utilitarian and Hedonic Product Types ........................................................... 17
CHAPTER 3. Proposed Research Model and Hypotheses ............................................ 19
3.1. Integrated Model of Virtual Experience ......................................................... 19

3.2. Hypotheses ...................................................................................................... 23
3.2.1. Cognitive Elaboration as Response to Virtual Experience .................. 23

3.2.2. Mental Imagery and Emotion as Responses to Virtual Experience ..... 25

3.2.3. Diagnosticity of Product Information as Mediator .............................. 27

3.2.4. Cognitive Elaboration, Attitude Strength, and Product Type .............. 31

3.2.5. Imagery-induced Emotion, Attitude Strength, and Product Type ....... 33
CHAPTER 4. Methodology ........................................................................................... 37
4.1. Research Design .............................................................................................. 37

4.2. Participants ...................................................................................................... 37

4.3. Stimulus .......................................................................................................... 38

4.4. Procedures ....................................................................................................... 41

4.5. Scales .............................................................................................................. 42
4.5.1. Scales of Endogenous Variables .......................................................... 43

4.5.2. Scales of Exogenous Control Variables ............................................... 49

5. Data Analysis ..................................................................................................... 50

CHAPTER 5. Results ................................................................................................. 53

5.1. Manipulation Checks ................................................................................... 53
5.2. Measurement Model Assessment ................................................................ 55
5.2.1. Item Purification ............................................................................... 55
5.2.2. Scale Reliability ................................................................................ 62
5.2.3. Construct Validity ............................................................................. 63
5.2.4. CFA for Attitude Accessibility and Attitude Confidence ................. 64
5.3. Effects of Presentation Mode on Attitude Accessibility and
Attitude Confidence .................................................................................... 67
5.4. Structural Model Assessment ...................................................................... 68
5.5. Hypothesis Testing ...................................................................................... 73
5.6. Decomposition of Effects ............................................................................ 78
CHAPTER 6. Discussion and Conclusion ................................................................. 84
6.1. Effects of Virtual Experience on Attitude Accessibility and
Attitude Confidence .................................................................................... 84
6.2. Role of Discursive and Experiential Processing in Mediating
Virtual Experience Effects .......................................................................... 85
6.3. Moderating Role of Utilitarian and Hedonic Product Types ....................... 87
6.4. Other Significant Findings .......................................................................... 88
6.5. Contributions ............................................................................................... 89
6.6. Limitations and Future Research ................................................................. 91
REFERENCES ........................................................................................................... 94
APPENDIX 1. QUESTIONNAIRE FOR MAIN EXPERIMENT ........................... 110
APPENDIX 2. INSTRUCTION FOR EXPERIMENT ............................................ 118
APPENDIX 3. INSTRUCTION FOR REACTION TIME MEASURE .................. 121
APPENDIX 4. DESCRIPTIVE STATISTICS OF DATA ...................................... 126

vi

LIST OF TABLES

Table 1. Results of Pretest 1 .......................................................................................... 39
Table 2. Results of Pretest 2 .......................................................................................... 40
Table 3. Operational Definitions of Thought Categories ............................................... 44
Table 4. Descriptive Statistics for Product-relevant Thought

(IBM laptop computer). .................................................................................. 45
Table 5. Descriptive Statistics for Product-relevant Thought

(Kodak digital camera) .................................................................................... 45
Table 6. Results of Presentation Mode Manipulation Check ......................................... 54
Table 7. Results of Product Type Manipulation Check ................................................. 54
Table 8. Summary of Fit Indices of Final Measurement Models .................................. 57
Table 9. Proposed Measurement Items for Constructs (IBM laptop computer) ............ 58
Table 10. Proposed Measurement Items for Constructs (Kodak digital camera) .......... 60
Table 11. Descriptive Statistics and Cronbach’s Alphas for Constructs ....................... 62
Table 12. Observed Correlations (IBM laptop computer) ............................................. 65
Table 13. Observed Correlations (Kodak digital camera) ............................................. 65
Table 14. Predicted Correlations and Residuals (IBM laptop computer) ...................... 66
Table 15. Predicted Correlations and Residuals (Kodak digital camera) ...................... 66
Table 16. Adjusted Means for Attitude Accessibility and Attitude Confidence ........... 68
Table 17. Summary of Fit Indices of Proposed Research Model .................................. 69
Table 18. Results of Hypothesis Testing ....................................................................... 78

vii

Table 19. Decomposition of Standardized Effects of Proposed Research Model
(IBM laptop computer) ................................................................................... 82

Table 20. Decomposition of Standardized Effects of Proposed Research Model
(Kodak digital camera) ................................................................................... 83

Table 21. Descriptive Statistics for Measurement Items (IBM laptop computer) ....... 127

Table 22. Descriptive Statistics for Measurement Items (Kodak digital camera) ....... 128

viii

LIST OF FIGURES

Figure 1. Proposed Research Model - Dual Processing Model of
Virtual Experience ......................................................................................... 22

Figure 2. Results of Standardized AMOS Analysis for Proposed Research Model
(IBM laptop computer) .................................................................................. 71

Figure 3. Results of Standardized AMOS Analysis for Proposed Research Model
(Kodak Digital Camera) ................................................................................. 72

ix

Chapter 1. Introduction

Virtual experience is an emerging concept in consumer psychology and e-
commerce. Prior research indicates that virtual product interactions, as simulated in three-
dimensional (3-D) product visualization, lead to increased product knowledge, more
favorable brand attitude, and stronger purchase intention than does indirect product
experience via traditional two-dimensional (2-D) online presentations (Edwards and
Gangadharbatla 2001; Li, Daugherty, and Biocca 2001; Li, Daugherty, and Biocca 2002,
2003; Schlosser 2003). Virtual experience resembles direct experience with respect to
interactivity and vividness, and these shared properties are believed to account for the
effectiveness of virtual experience (Li, Daugherty, and Biocca 2001).

Research on virtual experience until now has focused on traditional measures of
advertising effectiveness such as brand attitude and purchase intention in response to
virtual product interactions. Considering similarities between virtual and direct
experience, however, these standard valence attitude measures may not be able to detect
some of other subtle, yet important changes accompanying virtual experience. Although a
recent study suggests that virtual experience can influence attitude-behavior consistency
by fostering the formation of accessible and confident brand attitude (Lee and Li 2005),
evidently more research is needed to add robustness of this finding. Moreover, no
research has been conducted yet to account fully for the psychological processes of
virtual experience as it relates to these two dimensions of attitude strength. Thus, we do
not know yet how virtual experience actually shapes strong brand attitude.

In explicating the psychological processes of virtual experience, research on the

concept of experiential processing to persuasive communication (e. g., advertising) is
relevant. The “experiential processing” view emphasizes the role of mental imagery,
sensory pleasures, daydrearns, fantasies, and emotional responses for persuasion
(Hirschmann and Holbrook 1982; Holbrook and Hirschmann 1982; MacInnis and Price
1987; Meyers-Levy and Malaviya 1999), and is contrasted with the “discursive
processing” view focusing on verbal encoding/retrieval, cognitive responding, and
rational/analytical processing for persuasion (Ajzen and Fishbein 1980; Bettman 1979;
Greenwald 1968; MacInnis and Price 1987; Olson, Toy, and Dover 1982). When this
experiential processing strategy operates, judgments are not based on droughts or
cognitive elaborations generated in response to message content per se but rather on
sensations or feelings prompted by the very act of processing (Meyers-Levy and
Malaviya 1999; Strack 1992). Research indicates that interactive and well-designed
Internet user interfaces, 3-D product simulations in particular, are able to create an
enjoyable shopping episode via simulating a sensory and emotional experience of direct
product encounter (Li, Daugherty, and Biocca 2001; Schlosser 2003). This finding
suggests the experiential view may be relevant to the study of virtual experience, and
needs to be incorporated for a better understanding of what can be accomplished by this
phenomenon.

To date, however, discussion on the merit of virtual experience is limited, often,
to its non-experiential aspects such as elaborated processing, learning, efficiency of
information acquisition and decision-making (Ariely 2000; Haubl and Trifts 2000; Hoque
and Lohse 1999; Tremayne and Dunwoody 2001), failing to acknowledge the

experiential benefits it may offer. In contrast, other studies have been devoted

predominantly to exploration of the probable experiential ramifications of virtual
experience such as flow, escapism, immersion, and imagery (Hoffman, Novak, and
Duhachek 2003; Mathwick and Rigdon 2004; Novak, Hoffman, and Yung 2000),
ignoring non-experiential aspects of virtual experience.

Although these studies offer valuable insight into factors and processes related to
virtual experience, what seems needed is an integrated framework organizing and
combining the various theories and findings, and a delineation of the boundary conditions
in which each is likely to operate. The central goal of this dissertation is to develop such a
framework. In brief, this dissertation proposes a dual processing model of virtual
experience in which discursive processing and experiential processing constitute two
distinct routes of virtual experience as it relates to the strength of brand attitude. Given
the distinct roles theses two processing modes may play in online consumer behaviors
(Hoffman, Novak, and Duhachek 2003; Schlosser 2003), research on virtual experience is
reasoned to benefit from simultaneously taking them into account. By incorporating these
perspectives together into the existing virtual experience literature, the explanatory power
of such an integrated model in explicating online consumer behavior is expected to
improve.

In addition, this dissertation attempts to explore the conditions in which each
processing mode is likely to fimction, by proposing and testing product type as a
moderator. Specifically, the proposed model posits that the relative size of the effect of
each processing mode on the strength of brand attitude (attitude accessibility and attitude
confidence) is influenced by whether the product being examined is perceived as

primarily utilitarian or as primarily hedonic.

This dissertation is aimed to answer the following three questions.

i) Does virtual experience lead to the formation of strong brand attitude with

respect to attitude accessibility and attitude confidence?

ii) Do both discursive processing and experiential processing of virtual

experience contribute to the strength of brand attitude?

iii) If both processing modes contribute to attitude strength, then what factors
would determine the relative influence of each mode in shaping attitude
strength? More specifically, will the nature of a product of interest (e. g.,
utilitarian or hedonic) play a significant role in deciding which of the two

modes ensues, and mediate the effect?

This dissertation is divided into six chapters. Chapter 1 introduces the phenomena
to be examined. It provides an overview of the dissertation and its intended contribution.
Chapter 2 reviews the research on direct, indirect, and virtual experience in regard to their
impact on attitude accessibility and attitude confidence. In Chapter 2, the research on
discursive and experiential processing modes is also reviewed in their relations to the
distinction between utilitarian and hedonic products. Based on the review of literature,
Chapter 3 proposes an integrated framework of virtual experience and discusses the
research hypotheses derived from it. Chapter 4 describes the methodology to be used to

test the research hypotheses, along with a description of the scales that are used in the

empirical data collection. Chapter 5 reports the results of statistical analyses for
hypotheses testing. The proposed research model is examined in terms of its
psychometric properties and hypothesized causal relationships using confirmatory factor
analysis and structural equation modeling. Chapter 6 presents discussions on the research
findings presented in Chapter 5. Theoretical and practical implications, as well as
limitations of the present study are also discussed. An exploration of future study

directions concludes this dissertation.

Chapter 2. Literature Review

2.1. Direct, Indirect, and Virtual Experience

The classification of direct, indirect, and virtual experience is primarily based on
the degree of possible interaction between consumer and product, and the intensity of
sensory engagement involved in such interactions (Li, Daugherty, and Biocca 2001).
Direct experience involves the interaction of a product with the consumer’s full sensory
capacity, including visual, auditory, taste-smell, tactile, haptic, and orienting (e.g., trial,
sampling, or use of products; Gibson 1966), whereas indirect experience stems from a
secondhand, mediated exposure to the product, for example, via word of mouth
communication, TV commercials, print ads, or two-dimensional Internet ads.

Research on direct/indirect experience on consumer behavior suggests that direct
product experience is generally superior to advertising for persuasion (Marks and Kamins
1988; Smith and Swinyard 1982, 1983, 1988). In particular, the information integration
model (Smith and Swinyard 1982) predicts that when consumers directly experience the
product, they tend to develop strong (higher order) beliefs and attitudes, which culminate
in higher attitude-behavior consistency. Advertising usually elicits weak (lower order)
beliefs, attitudes, and conations because the tendency to accept the presented information
is low. The information integration model has been refined by Wright and Lynch (1995)
and Kempt and Smith (1998), whose work suggests direct experience should be superior
in conveying experience attributes, whereas advertising should be superior in conveying
search attributes. Of particular importance to this dissertation is the finding that strong

attitudes are generated by direct experience. Specifically, direct experience has been

shown to positively relate to attitude accessibility (Doll and Ajzen 1992; Fazio et a1.
1982) and attitude confidence (Fazio and Zanna 1978a, 1978b; Smith and Swinyard 1982,
1983; Wu and Shaffer 1987).

Virtual experience is formally defined as “psychological and emotional states that
consumers undergo while interacting with products in a 3-D environment” (Li, Daugherty,
and Biocca 2001 , p. 14). Virtual experience possesses certain properties of direct
experience, including interactivity and vividness (Li, Daugherty, and Biocca 2001), yet it
remains a mediated experience (Heeter 2000) that can be provided to larger audience
(Edwards and Gangadharbatla 2001). Virtual experience, therefore, may be able to enjoy
the advantages of both direct and indirect experience (Edwards and Gangadharbatla
2001; Li, Daugherty, and Biocca 2001).

Research indicates that virtual experience has positive effect on consumer
learning, brand attitude, and purchase intention (Coyle and Thorson 2001; Edwards and
Gangadharbatla 2001; Griffith and Chen 2004; Li, Daugherty, and Biocca 2002, 2003;
Schlosser 2003). The effectiveness of virtual experience stems from its ability to simulate
consumers’ first-hand evaluations of product claims. Virtual experience resembles direct
experience in that consumers interact with 3-D products and real products similarly as far
as visual input is concerned, which is quite different from the passive absorption of
information from traditional advertising (Li, Daugherty, and Biocca 2001). The effect of
virtual experience, however, appears qualified by the extent to which experiential
attributes can be manifested and thus conveyed online. Griffith and Chen (2004), for
example, report that virtual experience is more effective than online advertising based on

text information alone for products with fully digitizable attributes. However, its

effectiveness is restricted to affect and product evaluations for products of which
attributes cannot be fully manifested online (e. g., apparel). Behavioral measures are not
affected by virtual experiences which can not adequately convey products’ experiential
attributes. The relative effectiveness of virtual experience also appears affected by the
online-Shopper’s goal (searching vs. browsing) and individual difference in imagery

ability (Schlosser 2003) and expertise (Griffith and Chen 2004).

2.2. Virtual Experience and Attitude Strength

This dissertation assumes that an interactive and vivid 3-D mode leads to stronger
attitudes than a static and pallid 2-D mode, largely due to the similarities between direct
and virtual experience. Thus, more relevant studies are those that document the effect of
3-D visualization on non-evaluative dimensions of attitudes such as attitude intensity
(Klein 2003) and decision quality (Li, Daugherty, and Biocca 2003). These studies
suggest that standard valence/evaluative measures of attitudes (e. g., brand attitude,
purchase intention) may be inadequate for capturing the effect of 3-D product
visualization and therefore non-evaluative attitude dimensions should be measured for a
full grasp of what can be accomplished by 3-D product visualization.

The need for incorporating non-evaluative dimensions of attitudes can be further
justified by the fact that people do not always behave in accordance with their evaluative
dimensions of attitudes. The adoption of evaluative/valence attitude measures as
indicators of advertising effectiveness has been a common practice in advertising and
consumer behavior literature. The popularity of this approach derives from the implicit

assumption that these valence attitude measures are good predictors of actual behavior.

However, empirical evidence regarding the attitude-behavior relationship shows that,
although such a link is generally positive, in many situations the strength of the
association is rather weak (Fazio et al. 1982; Fazio and Zanna 1981; Smith and Swinyard
1983; Wicker 1969).

Responding to the lack of consistency in the attitude-behavior relationship, some
researchers proposed and empirically identified non-evaluative dimensions of attitudes as
moderators of the attitude-behavior relationship (Fazio and Zanna 1981; Zanna and Fazio
1982). These include attitude extremity, attitude ambivalence, affective-cognitive
consistency, attitude accessibility, and attitude confidence/certainty (Krosnick et a1. 1993;
Krosnick and Petty 1995). Indeed, these non-evaluative dimensions of attitudes comprise
the concept of attitude strength. Attitude strength has been defined as “a latent
psychological construct that is presumably represented in memory by various attributes
of the attitude” (Krosnick and Petty 1995, p. 3). The concept of attitude strength rests on
the premise that attitudes with equivalent extremity (either positive or negative) can differ
as to their underlying strength, and these differences in attitude strength moderate the
extent to which attitudes influence subsequent information processing and behavior
(F azio 1995). In fact, research on attitudes has demonstrated that some attitudes are more
firmly internalized and consequential, i.e., strong, whereas others are more flexible and
less influential, i.e., weak, on one’s thought and action (Krosnick et a1. 1993; Raden
1985). More specifically, attitudes are deemed strong when they are persistent over time,
resist change, have strong influence on information processing and judgments, and
strongly predict behavior (Haugtvedt and Priester 1997; Krosnick and Petty 1995).

Of several attitude strength variables, attitude accessibility and attitude confidence

are of primary interest largely due to their positive relationship with direct experience
(Fazio and Zanna 1981; Smith and Swinyard 1983). Research on these constructs also
implies that they are influenced by various advertising techniques, such as repeated
advertising and comparative advertising (Berger and Mitchell 1989; Yr, Phelps, and
Roskos-Ewoldsen 1998). Thus, aided by virtual experience technologies, it may be

feasible for advertisers to foster consumers’ attitude accessibility and attitude confidence.

2.3. Attitude Accessibility and Attitude Confidence

Attitude accessibility refers to the ease or quickness with which an attitude is
retrieved from memory. F azio (1986) defines attitudes as a learned association in memory
between an attitude object and one’s evaluation of the object. Attitude accessibility
indicates the strength of this object-evaluation association. In Fazio’s accessibility theory
(1986), behavior is posited to be a function of one’s perceptions of the attitude object in
the immediate situation in which the object is encountered. A highly accessible attitude
from memory is likely to be activated automatically upon the individual ’8 encounter with
the object (Fazio et a1. 1986). Once activated, the attitude serves to filter or bias the
individual’s immediate perceptions of the object and the situation (F azio and Williams
1986; Houston and F azio 1989) and thereby ‘guides’ subsequent behavior regarding the
object. In contrast, when the attitude is not readily accessible from memory, immediate
perceptions are less likely to be influenced by any previously formed evaluation toward
the object. Instead, these immediate perceptions tend to be based on other momentarily
salient features of the object and the situation that are not necessarily consistent with the

preexisting attitude. As a result, higher attitude-behavior consistency is more pronounced

10

when attitudes are highly accessible than when attitudes are relatively inaccessible.

Considerable empirical support for this proposition has been found in the
literature of consumer behavior (Berger and Mitchell 1989; Fazio, Powell, and Williams
1989; Priester et a1. 2004) and social psychology (Bassili 1993; Fazio et al. 1982; Fazio
and Williams 1986; Millar and Millar 1996). Typically, the activation potential of
attitudes in memory, namely attitude accessibility, is operationalized as the latency of
response to an attitudinal inquiry (Fazio 1986, 1990). Presumably, the speed at which
people respond to attitudinal inquiries reflects the likelihood of automatic activation of
attitudes upon their encountering the object.

Attitude accessibility and attitude confidence appear to be related, but they are
psychologically distinct constructs with different antecedents (Berger 1992; Berger and
Mitchell 1989). In the social psychology literature, attitude confidence has been defined
as the degree to which an individual is certain that his or her attitude toward an object is
correct, and it is usually measured by self-report of certainty or confidence (Krosnick and
Schuman 1988). Although people are motivated to hold correct attitudes (Petty and
Cacioppo 1996), they vary in the extent of their confidence in the correctness of attitudes.
In the consumer behavior literature, attitude confidence has been defined as the buyer’s
subjective feeling of “certainty,” “accuracy,” or “sense of feeling sure” in making a
judgment of the quality of a particular brand (Berger 1992; Howard 1989; Howard and
Sheth 1969). This definition appears to reflect two different theoretical dimensions of
attitude confidence. It may refer to the buyer’s overall confidence in the brands.
Alternatively, it may refer to the buyer’s confidence in his or her ability to judge or

evaluate attributes of the brands (Bennett and Harrell 1975). In this study, attitude

11

confidence is defined as the consumer Is subjective belief that his or her evaluation of a
product presented online is accurate.

As with attitude accessibility, attitude confidence has been proposed as a
moderator in the attitude-behavior relationship. Prior research shows attitudes held with
greater confidence are more predictive of subsequent behavior than attitudes with weak
confidence (Berger and Mitchell 1989; Fazio and Zanna 1978a, 1978b; Smith and
Swinyard 1983). Attitude confidence is also believed to provide some indication of the
degree of resistance an attitude enjoys in the face of counter-attitudinal attack: an attitude
held with a higher degree of confidence is less vulnerable to a counterattack aimed to

change it (Fazio, Powell, and Herr 1983; Fazio and Zanna 1981).

2.4. Discursive and Experiential Processing

Research on information processing to persuasive communication (e.g.,
advertising) has traditionally focused on discursive processingl (MacInnis and Price
1987). Researchers of discursive processing regard individuals as logical information
processors in arriving at a decision (Bettrnan 1979), and are usually interested in “how
symbols (e.g., words and numbers) are combined in a working memory and represented
to solve the problem” (MacInnis and Price 1987, p. 473). Accordingly, researchers in this
tradition have emphasized the role of verbal encoding/retrieval, cognitive responding, and

rational and analytical scrutiny of information as important mediators for persuasion

 

' In this dissertation, the term, “discursive processing” does not refer to a particular information processing
model. Rather, the term is used as a generic label that subsumes many of cognitively based information
processing models in persuasion, including multiattribute attitude model (F ishbein and Azjen 1975),
cognitive response model (Greenwald 1968; Olson, Toy, and Dover 1982; Wright 1973, 1980), the dual
processing models such as the ELM (Petty and Cacioppo 1981, 1986) and the HSM (Chaiken (1980).

12

(Bettman I979; Greenwald 1968; MacInnis and Price 1987; Olson, Toy, and Dover 1982;
Wright 1973, 1980).

Discursive processing assumes that when an individual is exposed to a persuasive
message, it is not the message itself that is important, but rather thoughts, ideas, cognitive
responses to the message that mediate persuasion (Brock and Shavitt 1983). In the
context of advertising, when consumers attend to an ad for brand evaluation purpose,
they ofien attempt to ‘bridge’ or ‘connect’ the information in the ad to related, preexisting
knowledge that they have about the product advertised and its product category
(Krugrnan 1965). For example, they may retrieve from memory previously stored criteria
normally employed for evaluating alternatives within the category and apply the criteria

for assessing the ad’s information. They also may respond to the ad with elaborated
cognitive responding such as counterargument and support argument, resulting in
agreement or disagreement with the ad’s propositions (Petty and Cacioppo 1986). These
cognitive activities can be described on an elaboration continuum. At the low end of the
continuum are simple responses such as the retrieval of a verbal label of an object.
Toward the high end of the continuum are more complex responses such as
counterarguments, support arguments, source derogations, and attributions (Bone and
Ellen 1992; MacInnis and Price 1987; Wright 1980).

In sum, when discursive processing strategy operates, persuasion is thought to be
a function of consumers’ reflections on, and cognitive responses about, the presented
message (Greenwald 1968; Wright 1980). As a result, attitude formation and change are
contingent on the net favorableness of the cognitive responses that occurred during

exposure to the message (Cacioppo and Petty 1979; Kisielius and Stemthal 1984, 1986):

13

when the elaboration is favorable, a positive attitude change is expected, whereas when
the elaboration is unfavorable, an attitude is adversely affected.

It should be noted that discursive processing postulates that messages can be
conveyed in the form of both text (e.g., ad copy) and visual (e.g., pictures and images).
Although much research on discursive processing in advertising and consumer behavior
has focused on consumers’ cognitive responses in response to ad copy (e. g., Olson, Toy,
and Dover 1982), semantic, reasoned cognitive activities also can be evoked fiom
pictures and images of the ad through their ability to imply verbal information (Mitchell
and Olson 1981; Rossiter and Percy 1980). In addition, in the context of this study, the
term cognitive responses is used in a rather limited manner in which it refers to only
thoughts and inferences related to the product being presented online. Although
cognitive responses have been defined broadly as any thoughts and inferences that arise
during exposure to a persuasive communication (Meyers-Levy and Malaviya 1999; Olson,
Toy, and Dover 1982), prior research indicates that the extent of issue- or topic-relevant
elaboration is directly responsible for the development of strong attitudes and
attitudinally consistent behavior (Haugtvedt and Priester 1997; Petty, Haugtvedt, and
Smith 1995). By the same logic, the extent of product-relevant elaborated thoughts that
emerge during exposure to a stimulus web site will be associated more closely with the
formation of strong brand attitudes than will other product-irrelevant thoughts that may
emerge concurrently.

Experiential processing emphasizes the role of experiential components of
responses to a persuasive communication, including mental imagery, sensory experiences,

daydreams, fantasies, and emotional responses (Hirschmann and Holbrook 1982;

14

Holbrook and Hirschmann 1982; MacInnis and Price 1987; Meyers-Levy and Malaviya
1999). Researchers following this tradition, therefore, have offered the possibility that
persuasion may be mediated by sensations or feelings that emerge during the very act of
processing rather than by cognitive responses to available information (Meyers-Levy and
Malaviya 1999; Stract 1992).

In the experiential view, the concept of mental imagery assumes a central role.
Mental imagery has been defined as a form of information processing that invokes
perceptions and sensory experiences in working memory (Baddeley 1986), and is “a
conceptually distinct process of representing information, a way that is ‘very like
picturing and very unlike describing’ (MacInnis and Price 1987, p. 473).” When
consumers are exposed to an ad, they not only process its messages in a semantic,
reasoned analytical manner, but may respond to it by creating mental images about their
potential behaviors in an imagined scenario in which they are the main characters
(Escalas 2004; MacInnis and Price 1987). For example, when a consmner sees a red Ford
Mustang in a TV commercial, he or she not only perceives and encodes multi-sensory
information from the commercial (e.g., visuals, sounds), but also may react to them by
generating multi-sensory images within himself or herself. He or she may imagine
himself or herself behind the wheel of the red Mustang, driving along a beautiful beach
road on a sunny day, feeling the breeze rustling his or her hair, all of which are also
“experienced” G-Iirschmann and Holbrook 1982).

When mental imagery operates, it leads to persuasion via reduced negative
cognitive responding, realism of experience, and associated emotional responses (Green

and Brock 2000). This way of processing information via mental imagery represents a

15

processing that is qualitatively different from discursive processing in which logical
consideration is critical for persuasion and thus is unrelated to cognitive elaboration
(Escalas 2004; Green and Brock 2000; Schlosser 2003). As with discursive processing,
mental imagery can be described also on an elaboration continuum, which ranges from
simple retrieval of a fleeting, simple image of an object to very rich and engrossing ‘real’
experiences induced by fantasies, daydreams, and visual problem solving (Bone and
Ellen 1992; Holbrook and Hirschmann 1982; MacInnis and Price 1987).

Mental imagery has several qualities that fundamentally distinguish it from
discursive processing. First, mental imagery implies mental transportation into a virtual
world created in a consumer ’s mind (Green and Brock 2000). Especially, when mental
imagery operates in the form of ‘mental simulation’ (Taylor and Schneider 1989) or
‘role-taking’ (MacInnis and J aworski 1989) with hypothetical scenarios and stories, it is
possible for consumers to be completely immersed into their self-generated stories that
captivate their attention and imagination. In narrative transportation literature, this
phenomenon is referred to as “immersion into a text” or “lost in a story” (Genig 1994;
Green and Brock 2000). Second, mental imagery can be multi-sensory in nature (Bone
and Ellen 1992; MacInnis and Price 1987). Although prior research on mental imagery
has mainly focused on visual imagery (i.e., generation of mental pictures of objects or
situations), mental imagery can incorporate a combination of different sensory
dimensions such as taste, smell, sight, and tactile sensations. In contrast, discursive
processing stresses semantic and cognitive processing, which are only remotely related to

sensory experiences.

16

Finally, mental imagery is intimately associated with affective responses and
emotions (Bone and Ellen 1992; MacInnis and Price 1987; Sheikh and Jordan 1983),
especially when the imagery involves the self (Taylor and Schneider 1989). In contrast,
discursive processing is relatively affect-free. Mental imagery shares the same
physiological mechanism with perception (sensation) and is thus able to produce a broad
range of physiological responses such as muscular reactions, heart rate, eye movements,
which reflect actual perceptual processes (Lang 1979; MacInnis and Price 1987). These
physiological responses are typically accompanied by emotions such as pleasure,
negative feelings, and arousal. Once evoked, these emotions are further to be related to
attitudes toward the ad and brand (Sheikh and Jordan 1983). Specifically, greater imagery
is shown to be associated with stronger emotions and affective responses to the ad and
the brand (Bone and Ellen 1992).

Although mental imagery is a multidimensional concept, encompassing vividness,
quantity, ease, and ability to manipulate images (see Bone and Ellen 1982 for a review),
this study focuses on vividness of mental imagery. Vividness of imagery appears most
relevant in virtual experience studies (Schlosser 2003). Specifically, in this study, the

vividness of a mental image is defined as the degree of clarity of an imagined episode.

2.5. Utilitarian and Hedonic Product Types

One of the main goals of the present study is to specify some boundary conditions
in which each of discursive and experiential processing modes is likely to operate. In this
regard, the idea of classifying products as being either utilitarian or hedonic is useful as it

is inherently related to the conceptualizations of the two processing modes.

17

Prior research indicates that products can be distinguished based on whether the
primary purpose associated with the consumption of the product is utilitarian/functional
or hedonic/experiential (Batra and Ahtola 1990; Hirschmann and Holbrook 1982; Mano
and Oliver 1993). Generally speaking, utilitarian products refer to goods whose purpose
of consumption is mainly instrumental and goal-oriented and accomplishes a functional
and practical task (e. g., toothpaste, cereal, detergent, laptop computer), whereas hedonic
products refer to goods whose purpose of consumption lies mainly in an sensory and
experiential experience of fun, pleasure, and excitement (e.g., movie, ballet, vacation
resort, painting) (Dhar and Wertenbroch 2000; Hirschmann and Holbrook 1982).
Accordingly, for a utilitarian product brand selection is primarily determined by the
consumer’s overall assessment about the functional utility of tangible product attributes.
In contrast, for a hedonic product a key determinant of brand selection is the consmner’s
anticipation about how well (the consumption of) the product will serve to satisfy his or
her emotional wants (Batra and Ahtola 1990; Hirschmann and Holbrook 1982). Although
the consumption of a product may involve both dimensions (Batra and Ahtola 1990),
products can be easily arrayed on a firnctional-experiential continuum (Batra and Ahtola
1990; Batra and Ray 1985; Hirschmann and Holbrook 1982; MacInnis and Jaworski
1989; MacInnis and Price 1987). As a result, it is possible that consumers can
characterize some products as primarily utilitarian and others primarily hedonic (Dhar

and Wertenbroch 2000).

18

Chapter 3. Proposed Research Model and Hypotheses

3.1. Integrated Model of Virtual Experience

The main thesis of this dissertation is that discursive processing and experiential
processing constitute two distinct processing modes of virtual experience relating to the
formation of strong brand attitudes — accessible and confident brand attitudes, and that
both of these processing modes can be united in a general theoretical fi'amework of
virtual experience — a dual processing model of virtual experience. Considering the
distinct roles these two processing modes may play in online consumer behaviors (N ovak,
Hoffman, and Duhachek 2003; Schlosser 2003), the explanatory power of such an
integrated model in explaining online consumer experiences is expected to improve.
Specifically, in the dual processing model, cognitive elaboration is posited to represent
the discursive processing aspect, whereas mental imagery and imagery-evoked emotions
represent the experiential components of virtual experience.

The proposition of discursive and experiential processes as two distinct routes of
virtual experience does not necessarily imply that they should occur in a mutually
exclusive manner. Rather, I suggest, consumers can experience both simultaneously. The
integration of discursive and experiential processes in this manner is compatible with the
premise that cognition and affect are parallel but integrated parts of human mental
framework (Epstein 1994; MacInnis and Price 1987; Zajonc 1968). Likewise, cognitive
elaboration, and mental imagery and imagery-evoked emotions are different but integral
parts of virtual experience, and the relation between these components is analogous to

that of cognition and affect in general.

19

Although theoretically interesting, integrating discursive and experiential
processing modes in a single framework is a methodologically difficult task since the two
processing modes might be intertwined in their effects on processing outcomes (MacInnis
and Price 1987). For example, it can be argued that a consumer may envision an image
and give it a verbal label or engage in elaborated cognitive responses based on an
imagined episode (MacInnis and Price 1987). If this happens, it makes the
disentanglement of the unique effects of each mode on processing outcomes challenging.
Not entirely dispelling this possibility, however, this research maintains that discursive
and experiential processing modes represent two “qualitatively” different processing
modes, and that their effects on attitude formation may be separable (Escalas 2004; Green
and Brock 2000; Schlosser 2003).

Figure 1 describes the proposed research model. Specifically, the model posits
cognitive elaboration and imagery-evoked emotions have a direct and separate effect on
the two dimensions of attitude strength, attitude accessibility and confidence. At the same
time, the model also allows for cognitive elaboration and mental imagery to affect the
two attitude strength dimensions indirectly via their joint influence on the perceived
diagnostic value of product information. The latter hypotheses can be thought of
reflecting the possibility of the two processing modes being intertwined on their effects
on processing outcomes.

In addition, this study attempts to shed light on some boundary conditions in
which each processing mode is likely to operate, by specifying product type as a possible
moderator for the relations between the two processing modes and attitude strength. In

doing so, an idea of classifying products as being either utilitarian or hedonic is

20

introduced as it is inherently related to the conceptualizations of discursive and
experiential processing. In essence, the dual processing model posits that the relative
contribution of discursive and experiential processing modes to attitude strength is
affected by the nature of a product in question. Specifically, the relative contribution of
the discursive processing on attitude accessibility and attitude confidence will be more
prominent for utilitarian products, whose goal of consumption is mainly functional and
goal-driven than for hedonic products, whose goal of consumption is mainly for
enjoyment and arousal. On the other hand, the relative contribution of the experiential
processing on attitude accessibility and attitude confidence will become more prominent
for hedonic products than for utilitarian products. The final two sets of hypotheses deal
with the moderating roles of utilitarian vs. hedonic product type for the relationships
between cognitive elaboration and attitude strength (H7a & H7b) and between imagery-
evoked emotion and attitude strength (H8a & H8b).

It is worth noting that in the proposed research model attitude accessibility and
attitude confidence are represented as two distinct constructs, rather than as alternative
dimensions of an overarching latent construct that can be labeled attitude strength (i.e., as
two first-order dimensions of attitude strength). Conceptually, they can be thought of as
two manifestations of attitude strength. However, existing empirical evidence does not
support a single factor structure encompassing these dimensions (Krosmck et a1. 1993).
In addition, given the clear distinctions between accessibility and confidence in terms of
definitions and operationalizations, combining these constructs into a single second-order
factor appears counterproductive. Consequently, attitude accessibility and attitude

confidence are treated as two separate constructs in this study.

21

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a»: _ 3.833— mm.—
_ 3:335”.— Down—:—
M .5: r 382
N:
352.280 m:
emu—533‘ aw:
Earneuua: cannon.»
5:258:— 563 .5
hag—mung: 3:
83:2 4:
:—
a ..:.
green—EH
as: _ 3.35

 

«2.3.—2.5— 3.5.5» .«e .352 wumnmooEm .55 I 3.52 seasoned .5258.— A95»:—

22

3.2. Hypotheses
Each path in the proposed research model (Figure 1) constitutes research
hypotheses of this dissertation. The following section describes theoretical rationale and

supporting empirical evidence for each hypothesis.

3.2.1. Cognitive Elaboration as Response to Virtual Experience

Interactivity and vividness, two key characteristics of virtual experience, are
assumed to contribute jointly to the degree of cognitive elaboration consumers engage in
for 3-D products. First, the direct manipulation of virtual products using 3-D features is
expected to enhance consumers’ cognitive elaboration of the product presented because it
requires them to think about how to interact with the product properly. Furthermore,
interactive 3-D features allow for control over the flow of information to gather
information relevant to users’ unique informational needs and interest (Ariely 2000). An
online shopper for a laptop computer, for example, is able to examine the laptop at
different angles, zoom in and out for close inspection, and even try for different colors for
the laptop’s lid to see what it looks like prior to purchase. The important point for this
interaction is that these visual and behavioral simulations are conducted at a pace and in a
sequence of the consumer’s own choice. As a result, each individual consumer is able to
receive his or her own unique combination of messages and experiences (Sundar and
Kim 2005). This added opportunity for customizing information flow and acquisition
appears to facilitate the consumer’s elaboration of the product-relevant information. In

support of this view, an individual’s ability to control the pace at which the information is

23

delivered has been found to increase the likelihood of elaborative cognitive processing
(Chaiken and Eagly 1983; Krugrnan 1965).

Being different from a comparatively pallid 2-D product presentation, the
vividness of a 3-D product presentation is also likely to induce consumers to engage in
product-relevant cognitive elaboration with richer information cues (Kisielius and
Stemthal 1984, 1986). Kisielius and Stemthal (1984, 1986), for example, have shown
that vivid information affects the degree to which consumers will engage in cognitive
elaboration of the product advertised, which in turn mediates judgmental attitudes.
According to their availability-valence hypothesis, vividly presented information evokes
cognitive elaboration that may either enhance or undermine the persuasiveness of a
message, depending on the relative favorableness of the thoughts elaborated.

There is direct empirical support for this proposition. Li, Daugherty, and Biocca
(2001) have found that when interacting with 3-D products, participants tend to produce
more product-relevant thoughts than when interacting with 2—D products. In fact, this
active cognitive process found in the 3-D condition is one of the most consistent findings
witnessed by participants exposed to 3-D products (Li, Daugherty, and Biocca 2003).

Given the above discussion and findings, this dissertation predicts that virtual
experience created by 3-D online environments leads to greater cognitive elaboration of
the product presented as compared to static 2-D presentations. The following hypothesis

reflects this proposition.

H1: Participants exposed to products in the 3-D condition will engage in more
product-relevant cognitive responding than those exposed to products in
the 2-D condition.

24

3.2.2. Mental Imagery and Emotion as Responses to Virtual Experience

The interactive nature of 3-D product presentations is expected to lead to more
vivid mental imagery of product use, compared to static 2-D presentations online
(Schlosser 2003). One of the most fundamental features that distinguishes 3-D
presentations from traditional 2-D presentations is their ability to directly manipulate
virtual products (Li, Daugherty, and Biocca 2001). With direct manipulation, there is an
immediate and constant change of graphics reflecting user behaviors, which resembles
physical actions (Schlosser 2003). What is implied in this interaction is that the computer
system is capable of giving feedbacks in response to the actions users perform on the
computer, resulting in a sense of “as if” responding — “responding as if the events were
occurring in the physical world (Schlosser 2003, p. 185).” According to Schlosser (2003),
it is this “as if” responding that blurs the boundary between virtual and physical reality
(Biocca 1992) and causes users to be immersed into an imaginative world where they are
being transported as participants (Green and Brock 2000). This whole process can be
captured properly by the term, mental imagery. Especially, when mental imagery occurs
in the form of ‘mental simulation’ (Taylor and Schneider 1989) with hypothetical
scenarios and stories, it is possible for users to be immersed completely into their self-
generated stories that captivate their attention and imagination.

In line with this reasoning, direct exploration of virtual products using 3-D
features is expected to stimulate mental imagery of product use. With interactive 3-D
features, consumers are able to move, rotate, zoom in and out, and customize virtual
products, and each of these direct manipulation capabilities of 3-D interface appears

contributive to the creation of more vivid mental images of product use.

25

The vivid product descriptions of 3-D interface also facilitate mental imagery by
enhancing the realism of the experience. Vividness as a media characteristic is defined as
“the representational richness of a mediated environment as defined by its formal
features; that is, the way in which an environment presents information to the senses”
(Steuer 1992), and consists of two dimensions — sensory breadth, representing the
number of sensory dimensions simultaneously a medium can appeal to, and sensory
depth, representing the depth or quality of information a medium can provide to the
senses. Although current 3-D technologies are limited in their capacity to simulate tactile
and olfactory sensations (Biocca 1992; Li, Daugherty, and Biocca 2001), 3-D
presentations of products are still capable of appealing to sight via vivid visual and
behavioral simulations, and this is more powerful than pallid 2-D presentations. This
appealing to sight may then foster vicarious experience of product use by bringing
consumers closer to a more realistic environment that imitates real experience. In line
with this assrnnption, Edwards and La Ferle (2003) suggest that appealing to sight using
the various images and textures associated with the product presented online will enhance
the believability of the virtual experience, thus helping consumers imagine themselves
using the product.

Given the above discussion and findings, this study predicts that virtual
experience created by 3-D online environments leads to more vivid2 mental images of
product use than do static 2-D presentations. The following hypothesis reflects this

proposition.

 

2 Vividness of mental imagery refers to the clarity of mental images and is the participant’s response to the
stimuli. In this regard, it differs vividness as an inherent characteristic of the media interface.

26

H2: Participants exposed to products in the 3-D condition will have more vivid
mental imagery of product use than those exposed to products in the 2—D
condition.

Mental imagery of having and using a product should involve a sensory and
emotional experience, which in turn leads to the generation of strong judgments about
the product. Prior research indicates that mental imagery evoked by stimuli can be
affect-laden (Lang 1979; MacInnis and Price 1987; Sheikh and Jordan 1983),
especially when the imagery involves the self (Taylor and Schneider 1989). Mental
imagery shares the same physiological mechanism with actual sensation, thus is able to
produce a broad range of physiological responses such as muscular reactions, heart rate,
eye movements, which reflect actual perceptual processes (Lang 1979; MacInnis and
Price 1987). Specifically, it is expected that vivid and elaborated imagery of product
use to be associated with stronger emotional responses than dull and less-elaborated
imagery. Evidently, fantasizing or daydreaming about experiencing the product alone
or with significant others in an imagined scenario should be a more vivid sensory and
emotional experience, than merely envisioning an image of the product in mind.

Given the above discussions and findings, the following hypothesis is

developed.

H3: Vivid mental imagery will evoke stronger emotional responses.

3.2.3. Diagnosticity of Product Information As Mediator
Consumers may use their online experiences as an information source when

they form judgments about the product being presented. And when an online

27

experience serves an informational function, consumers are expected to evaluate the
quality of the experience as an information source about the product. In doing so, they
may consider several aspects of their online experience, including perceived
diagnosticity of product information (e.g., how helpful and relevant was the
information acquired fi'om it for product evaluations), self-perceived expertise (e. g.,
am I knowledgeable enough to process the information provided from it), and the
duration of experience (e. g., was it long enough to allow for a thorough evaluation)
(Kempf and Smith 1998). Among these aspects, particularly interesting to the present
study is consumers’ perception of the diagnostic value of the information obtained.
Specifically, they may assess the usefulness or utility of product-relevant information
obtained in forming accurate judgments about the product. A key question to be raised
in this process will be whether the experience provided useful information for product
evaluations. Answer to this question is critical because the experience will be factored
into judgments of the product to the degree to which consumers perceive the
information obtained to be useful and diagnostic for an accurate judgment. of the
product.

Consistent with Kempf and Smith (1998), this study defines diagnosticity of
product information as the degree to which the consumer believes that product-relevant
information acquired while interacting with a product web site is useful in correctly
inferring his or her attitude toward the product presented. That is, a given piece of
product-relevant information is regarded as diagnostic to the extent to which the
consumer perceives the information is helpful in forming accurate judgments about the

quality and performance of a product.

28

In the context of this study, both cognitive elaboration and mental imagery are
assumed to contribute jointly to the increase of the perceived diagnostic value of the
product-related information acquired. Cognitive elaboration as well as mental imagery
leads a consumer to think about and evaluate possible consequences of having and
using a product in question, but in drastically different ways. When engaging in
cognitive elaboration about the product, the consumer is likely to carefully assess pros
and cons of having the product before arriving at an overall judgment about the
product. This process is task-oriented and analytical in nature (Hirschmann and
Holbrook 1982). Alternatively, the consumer may vicariously experience the product
by visually imagining the product in use and use that imagined scenario as a basis for
judgment formation about the product. This process tends to be experiential and
holistic in nature (MacInnis and Price 1987). Despite the underlying differences in
processing information, however, product-relevant information elicited as a result of
effortful cognitive activity and vivid mental imagery will be perceived as a diagnostic
basis for attitude formation about the product because both activities represent an
added opportunity to elaborate on potential consequences for having and using the

product. Based on this reasoning, the following set of hypotheses is developed.

H4: Cognitive elaboration will be positively associated with diagnosticity of
product information acquired.

H5: Mental imagery will be positively associated with diagnosticity of product
information acquired.

29

 

The greater the perceived diagnosticity of the informational basis of an attitude,
the more accessible and confidently held the attitude is. According to Fazio (1995),
people are sensitive to the diagnostic value of the evidentiary base upon which they are
relying. Through their previous learning, individuals may come to trust certain kinds of
information as a more reliable, relevant, and thus, more diagnostic indication of their
attitudes, than are other kinds of information. For example, people may learn that
information obtained via direct behavioral experience is more diagnostic of their attitudes,
than information acquired via indirect experience such as word of mouth communication.
More diagnostic grounds of an attitude are of concern here because the evaluation formed
by such kinds of information may be noted more readily and held more confidently
(Fazio 1995). In fact, this relationship of attitudinally diagnostic information to attitude
accessibility and attitude confidence seems able to readily explain differential effects of
direct vs. indirect experience with respect to the formation of strong attitudes (Berger and
Mitchell 1989; F azio 1995; Fazio and Zanna 1978a, 1978b, 1981; Smith and Swinyard
1983). Despite the lack of empirical evidence to directly support the diagnosticity-attitude
accessibility relation, several studies exist to support the diagnosticity-attitude confidence
relation (Kempf and Smith 1998; Koriat, Lichtenstein, and Fischoff 1980). The following

set of hypotheses aims to reflect these propositions.

H6a: Higher levels of diagnosticity of product information will lead to more
accessible brand attitude.

H6b: Higher levels of diagnosticity of product information will lead to more
confident brand attitude.

30

3.2.4. Cognitive Elaboration, Attitude Strength, and Product Type

The extent of cognitive elaboration a consumer engages in for the product
presented is associated with the development of strong brand attitudes. Petty and
Cacioppo’s influential dual-processing model of persuasion (I 981, 1986), the Elaboration
Likelihood Model (ELM), is relevant to this proposition. The ELM distinguishes two
distinct routes to persuasion - central and peripheral routes. The central route refers to
attitudes that are based on an effortful cognitive activity, whereby individuals focus their
attention on issue-relevant arguments and product-related attributes, and draw on their
prior experience to assess and elaborate on presented information. In contrast, the
peripheral route refers to attitudes that are based on relatively simple “cues” in the
persuasive setting that determine attitudes via simple decision rules (Petty, Cacioppo, and
Schumann 1983).

A key postulate in the ELM is that the strength of an attitude is based on the
extent of cognitive elaboration the individual has done about the attitude object. That is,
when an attitude changes as a result of careful deliberation about the merit of the product
(i.e., central route), the resulting attitude will be stronger than if an attitude changes via
simple decision rules such as memory-based heuristics (i.e., peripheral route) (Petty and
Cacioppo 1986; Petty, Haugtvedt, and Smith 1995). In explicating the underlying

processes of this phenomenon, Petty and Cacioppo (1986) argued that:

the process of elaborating issue-relevant arguments involves assessing the
schema for the attitude object in order to evaluate each new

argument. . . .Under the peripheral route, however, the schema may be accessed
only once to incorporate the affect or inference elicited by a salient

cue. . . .Under the central route then, the issue-relevant attitude schema may be
accessed, rehearsed, and manipulated more times, strengthening the

31

interconnections among the components and rendering the schema more
internally consistent, accessible, enduring, and resistant than under peripheral
routes. (p. 22)

In line with the above reasoning, it is hypothesized that brand attitudes formed as
a result of more product-related cognitive activity are more accessible than brand
attitudes formed in response to less product-related cognitive activity. An unpublished
work by Petty, Haugtvedt, and Rennier (1995) provides direct empirical evidence of this
proposition, reporting that attitudes stemming from elaborative cognitive processing are
more accessible from memory than those resulting from less-elaborated cognitive
processing (i.e., peripheral processing). Here, the very act of deliberation and elaboration
may foster attitude rehearsal as the examination of different pieces of information leads to
the same evaluation repeatedly, and this increased attitude rehearsal appears to account
directly for the effect of cognitive elaboration on attitude accessibility (Berger and
Mitchell 1989; Fazio I995). Attitude rehearsal is also likely to increase the perception of
consumers’ confidence in their brand attitudes as it may cause them to make an attitudinal
decision repeatedly about the product. Einhom and Hogarth (1978) provide evidence that
confidence in judgments increases as a function of how often the decision is made.
Furthermore, a greater degree of thoughtful consideration and deliberation on product-
relevant thoughts themselves should increase the level of confidence consumers have
toward their attitude about the product (Petty, Cacioppo, and Schumann 1983).

In discussing the efiects of cognitive elaboration on attitude accessibility and
confidence, the role of product type is important. Especially, an idea of classifying
products as utilitarian or hedonic becomes valuable in this study since it is inherently

related to the conceptualizations of discursive and experiential processes, thus serving to

32

illuminate some boundary conditions in which each mode is likely to operate.

Utilitarian products are goods whose purpose of consumption is primarily
functional and goal-driven (Dhar and Wertenbroch 200; Hirschmann and Holbrook 1982).
Thus, an analytical analysis of tangible product qualities and utility performance serves
as a main determinant of the formation and change of the consumer’s attitude toward a
utilitarian product (Hirschmann and Holbrook 1982). In contrast, for hedonic products,
whose purpose of consumption is mainly to have fun and enjoyment, this cognitively
charged analytical reasoning is less valued. Rather, for these products, a key determinant
of brand attitude is the consumer’s anticipation about how well the consumption of the
product will serve to satisfy his or her emotional wants (Batra and Ahtola 1990;
Hirschmann and Holbrook 1982).

Based on the preceding literature and discussion, it is hypothesized that greater
product-related cognitive activity leads to more accessible and confident brand attitude.
The relative contribution of cognitive elaboration on attitude strength dimensions,
however, is likely to be moderated by whether the product under examination is
perceived as primarily utilitarian or hedonic. The following set of hypotheses reflects this

proposition.

H7: Product type (utilitarian vs. hedonic) will moderate the relationship between
cognitive elaboration and (a) accessibility and (b) confidence. For both
relations, a stronger effect will be pronounced for participants exposed to a
utilitarian product, than those exposed to a hedonic product.

3.2.5. Imagery-induced Emotion, Attitude Strength, and Product Type

Once evoked, emotions are associated with the product. It has been shown that

33

ad-induced emotions can influence attitude toward the ad and attitude toward the brand
(Aaker, Stayman, and Hagerty 1986; Batra and Ray 1986; Burke and Edell 1989; Edell
and Burke 1987; Holbrook and Batra 1987). Positive emotional responses (e.g.,
amused, happy, hopeful, elated) are associated with favorable attitude toward the ad
and the brand, whereas negative emotional responses (e. g., skeptical, offended,
depressed) are associated with unfavorable attitude toward the ad and the brand (Burke
and Edell 1989). More relevant to this study are the results of studies of effects of
emotional reactions to direct product experience, since virtual experience allows
simulated direct experience with products. Prior research has identified “pleasure” and
“arousal” as the two major dimensions of emotions most relevant to a direct experience
situation (Havlena and Holbrook 1986; Mano and Oliver 1993). These two dimensions
have been found to influence attitude toward the brand in the context of a trial (Kempf
1999; Kempf and Smith 1998).

Many previous studies in advertising and consumer behavior point to the
significant role of emotions in attitude formation and change, whereas empirical
studies directly examining the connection of imagery—evoked emotions to attitude
formation and change are rare. Despite this dearth, a hint can be found in the research
on the effect of emotions on attitude-behavior consistency in social psychology. Fazio,
Zanna, and Cooper (1978), for example, conducted an experiment where two groups of
participants were indirectly exposed to a set of puzzles by watching a videotape of
people working out puzzle examples, while a narrator describes the puzzles. One group
of participants was simply asked to listen and watch carefully while the other group

was asked to emphasize and imagine how they would feel if they were in the position

34

 

of the people working on the puzzles. Interestingly, attitudes among participants
focusing on feelings were found more predictive of later free-play behavior than were
attitudes among participants for whom focusing on feelings was not emphasized.
Although they did not directly assess attitude strength, it is reasoned that imagery-
evoked emotional reactions to the puzzles were perceived as a trustworthy and
diagnostic basis for evaluation about the puzzles (F azio 1995; Millar and Millar 1996).
Thus, the resulting attitude may have become readily accessible and held with more
confidence, leading to more attitudinally consistent behavior during the fi'ee-play
period. Millar and Millar (1996) found evidence that emotional responses to an attitude
object indeed are positively associated with attitude accessibility.

As hypothesized for the cognitive elaboration - attitude accessibility and
attitude confidence links, whether the product being examined is perceived as
utilitarian or hedonic will play a significant role in determining the relative
contribution of imagery-evoked emotions to the attitude strength dimensions.
Especially, for hedonic products, where a sensory and experiential aspect of
consumption matters, emotions elicited from imagined episodes of product use will
constitute a crucial part of brand evaluation and, ultimately, its selection. It is reasoned
that vicarious consumption of a product realized via mental imagery of experiential
processing will provide hedonic value by allowing consumers to enjoy the product’
benefits without purchasing it (Babin, Darden, and Griffin 1994; MacInnis and Price
1987). Consumers seek fun, pleasure, and excitement through the consumption of
hedonic products, thus vicarious consumption of how well they would serve to satisfy

emotional wants becomes an important informational source about those products

35

(Hirschmann and Holbrook 1982; MacInnis and Price 1987). In contrast, for utilitarian
products, these affectively charged feelings are less valued in the formation of
consumers’ brand attitudes because they are primarily evaluated on their practical and
instrumental values, rather than their experiential aspects of consumption.

In line with the preceding literature and reasoning, imagery-evoked emotions
are expected to lead to accessible and confident attitudes toward the imagined product
because emotional reactions elicited from imagined episodes are apt to involve a
product under examination, and consumers seem to trust their emotional reactions to
the product as diagnostic. The relative contribution of imagery-evoked emotions on
attitude strength dimensions, however, is likely to be moderated by whether the
product under examination is perceived as primarily utilitarian or hedonic. The

following set of hypotheses reflects this proposition.

H8: Product type (utilitarian vs. hedonic) will moderate the relationship between
emotional responses and (a) accessibility and (b) confidence. For both
relations, a stronger effect will be pronounced for participants exposed to a
hedonic product, than for those exposed to a utilitarian product.

36

 

Chapter 4. Methodology

4.1. Research Design

An experiment was conducted in a lab setting. Since the effects of presentation
mode and product type were of particular interest of this research, these two variables
became the treatment factors in a 2 x 2 mixed factorial design. The between-subj ect
factor was presentation mode (2-D or 3-D) and the within-subject factor was product type
(utilitarian or hedonic). Participants were randomly assigned to either the 2-D or the 3-D
treatment conditions and were exposed to both product categories — one representing a
utilitarian product (IBM laptop computer) and the other representing a hedonic product
(Kodak digital camera). The order in which the different types of products were displayed
was randomized for each participant to avoid possible order effect. Participants were in
groups of 5 to 10 in a spacious computer lab, where they were isolated each other so that
they could complete the experiment independently. Survey questionnaire and participant

instructions used in the experiment are shown in Appendix 1 and 2.

4.2. Participants

A total of 238 undergraduate students enrolled at a Midwestern university
participated in the experiment in return for class credit. Participants’ ages ranged from 18
to 25, and approximately 66% were female. All participants signed an informed consent
form prior to their participation in the experiment. A student sample is considered
appropriate for this experiment because they have high Internet literacy and represent

early adopters of e-commerce in 3-D product visualization (Li, Daugherty, and Biocca

37

2001). After eliminating five incomplete responses, a total of 233 usable surveys were
obtained. Further, seven surveys from the IMB laptop computer condition and six surveys
from the Kodak digital camera condition were deleted due to response error during the
response latency test (see Attitude Accessibility in section 4.5 Scales for further
explanation). As a result, 226 usable surveys were included in the final sample for
construct validation and hypothesis testing for the IMB laptop computer and 227
responses for the Kodak digital camera. For manipulation checks, the pooled sample with

221 surveys was used.

4.3. Stimulus

Four stimulus web sites were created to represent four experimental conditions. A
laptop computer and a digital camera were selected as the test product categories
representing the utilitarian and hedonic product classification for the experiment based on
the results of a pretest and previous research. A pretest, consisting of 35 participants taken
from the same population pool as those in the main experiment, indicated that laptops and
digital computers are relevant to the intended participant’s population (i.e., undergraduate
students). More importantly, these products were perceived to be significantly different in
terms of their utilitarian and hedonic natures. In the pretest, participants were asked to
categorize laptops and digital cameras as primarily utilitarian, defined as practical,
functional, or something that helps achieve a goal (e.g., pain killer), as primarily hedonic,
defined as pleasant and fun, or something that is enjoyable and appeals to senses (e. g.,
perfume), and as both utilitarian and hedonic (Dhar and Wertenbroch 2000).

As shown in Table 1, the majority of participants classified laptops as primarily

38

utilitarian (21 out of 35 participants) and digital cameras as primarily hedonic (20 out of
35 participants). Previous research also suggests that these two product categories
represent product types whose salient features can be adequately evaluated with 3-D
interfaces, not necessarily requiring tactile inspection (Li, Daugherty, and Biocca 2001;

2002; 2003).

Table 1. Results of Pretest 1

 

 

Product Category
Perceived Function Laptop computer Digital camera
Utilitarian 21 5
Hedonic 6 20
Both 8 10
Total 3 5 3 5

 

Note: x2@) = 11.37, p < .01 for laptop computer; x20) = 9.31, p < .01 for digital camera.

A second pretest (N=3 7) was conducted to select a brand name and salient
product attributes for each of the two test product categories. IMB and Kodak were
identified in the pretest as a brand with neutral brand attitudes in each category against
six major manufactures respectively (SONY, Dell, Sharp, Toshiba, HP, Gateway for
laptops, SONY, Olympus, Samsung, Minolta, Fuji, Pentax for digital cameras) and thus
chosen for the test brands (see Li, Daugherty, and Biocca 2002 for a similar case).
Participants were asked to rate their preferences for seven popular laptop and digital
camera brands using a seven-point semantic differential scale item (1=1ike/7=dislike). For
laptops, results indicate that SONY, HP, Dell comprise a more favored brands group,

while Gateway, Toshiba, and Sharp 3 less favored brands group, with IBM as a neutral

39

brand between these two sets of brands. For digital cameras, SONY, Olympus, and
Samsung comprise a more favored brands group, while Fuji, Minolta, and Pentax a less
favored brands group, with Kodak as a neutral brand between these two sets of brands

(Table 2).

Table 2. Results of Pretest 2

 

 

 

Laptop Computer brands Mean Digital camera brands Mean
SONY 5.88 SONY 6.19

HP 5.31 Olympus 5.25

Dell 5.28 Samsung 5.14

IBM 5.09 Kodak 5.00

Gateway 4.78 Fuji 4.75

Toshiba 4.25 Minolta 4.06

Sharp 3.94 Pentax 3.53

Note: N=37

Salient product attributes were also identified in the pretest using free elicitations.
During the pretest, participants were asked to list product attributes that they thought
were important to consider for the selection of a laptop and a digital camera. A total of 18
attributes were identified for a laptop and a digital camera each. Among those, the seven
most salient attributes (quality, size/weight, price, design/style, battery life, tech-
support/warrantee, and battery life for both laptops and digital cameras and wireless-
Intemet capability for only laptops and easy to use for only digital cameras) were used,
along with a company logo, a headline, and a sub-headline to construct the stimulus web

sites.

40

"

Four web sites, consisting'of 2-D and 3-D pages for a laptop and a digital camera
each, were created to represent the four treatment conditions. All features were identical
in the 2-D and 3-D versions except for the presentation mode: the 3-D version had the
product in a 3-D visualization where participants can move, zoom in or out, and rotate the
product; the 2-D version had the product in a series of still photos. In the 2-D conditions,
multiple photos of the test product (an IBM laptop or a Kodak digital camera) were
displayed in an attempt to equate the amount of information available between the 2-D
and 3-D conditions, thus minimizing possible confounding due to information effects.
The 2-D web site for the IBM laptop computer showed the front, side, bottom of the
laptop, all with its lid open, while the 2-D web site for the Kodak digital camera showed
the front, back, and the top of the camera. In fact, for both products, participants
perceived the amount of information from the 2-D and 3-D conditions not different
(Mzmapmp = 5.84 vs. M3pzapmp = 5.66, t(224) = 1.32, p > .05; szcmra = 5.92 vs. M3Dcamm

= 5.85, t(225) = .59, p > .05), suggesting this attempt was successful.

4.4. Procedures

Upon arriving at a computer lab for the main experiment, participants were
greeted and asked to fill out a consent form and a short survey. The survey included items
measuring participants’ abilities to imagine, product category experience, and brand
familiarity, which are the control variables of the study. Participants’ background
information and computer usage were also collected in the short survey. After completing
the short survey, they were seated before a computer terminal corresponding to the

assigned experimental condition. The first screen of the computer explained that the study

41

concerns “the evaluation of two products to be presented on web sites.” The screen also
contained instructions that upon completion of the web sites viewing, they would be
asked to report their opinions and thoughts. The instructions were necessary to have
participants pay close attention to test materials and actively engage in brand information
processing (Kempf and Smith 1998; Li, Daugherty, and Biocca 2002).

Afier reading the instructions, participants were directed to the first product
promotion web site where they were allowed to take as much time as necessary to
examine the web site (Griffith and Chen 2004; Li, Daugherty, and Biocca 2003). After
viewing the web site, participants were asked to complete a portion of the survey
inquiring about the dependent variables of the study for the product examined. Measures
for interactivity, vividness, and product type (i.e., whether the product is perceived as
utilitarian or hedonic) were also included as manipulation checks for the presentation
mode and product type treatments. After completing the relevant portion of the survey,
participants were directed to the second product promotion web site. After viewing the
second web site, participants were asked to complete the remaining portion of the survey
inquiring about the dependent variables for the second product examined.

Completing the survey, participants were then given instructions about the
response latency test to measure attitude accessibility. Prior to the collection of latency
data, participants were given a series of practice trials to familiarize them with test

procedures. The whole experiment took about 35 to 40 minutes to complete.

4.5. Scales

Most of the scales were adopted from previous research with some modified to fit

42

the context of this study. As recommended by Anderson and Gerbing (1984) and Bentler
and Chou (1987), all constructs of this study, except for cognitive elaboration and attitude
accessibility, were measured by at least three observable indicators (items). These items
were written in the form of statements, most using a seven-point Likert or semantic

differential rating system.

4.5.1. Scales of Endogenous Variables

Cognitive Elaboration. A standard thought-listing procedure (Cacioppo and Petty
1981) was used to measure the degree of participants’ elaborated thinking toward the test
products. Participants were given a lined, blank sheet of paper on which they were
instructed to write down all their thoughts that went through their mind while they
viewed the web site for the test product. They were told to ignore spelling, punctuation,
and grammar. A time limit of ten minutes was imposed on the task to increase the validity
of the measure. The purpose of imposing the time limit was to allow sufficient time for
complete recording of the honestly spontaneous thoughts, while reducing the probability
that purely reactive thoughts were listed (Cacioppo and Petty 1981; Wright 1973). Once
collected, the semantic contents of cognitive responses (i.e., listed thoughts) were coded
into self-thoughts, product-thoughts, and irrelevant thoughts3 (Schlosser 2003; Shavitt
and Brock 1986) by two judges blind to participants’ experimental condition and the

research hypotheses (Cacioppo and Petty 1981). Operational definitions of these thought

 

3 Initial coding categories of cognitive responses also included a category for ad-related execution thoughts
(e.g., participants’ evaluations of ad style). For the purpose of this dissertation, however, these thoughts
were classified as irrelevant. Although ad execution-related thoughts have been shown to affect attitude
toward the ad and subsequently brand attitude in certain situations (Batra and Ray 1986; Edell and Burke
1987), the link mainly represents what can be characterized as the peripheral or the heuristic attitude
change process, which has little bearing on the dimensions of attitude strength, the focal concepts of this
dissertation (Haugtvedt and Priester 1997; Petty, Haugtvedt, and Smith 1995).

43

categories are shown in Table 3.

Table 3. Operational Definitions of Thought Categories

 

Thought Categories

 

Definition

 

Self-Thought

Product-Thought

Irrelevant Thought

Statements involving references to own purchasing or trying of
the product, self-characterizations and associations, and
statements about others’ reactions.

e.g.) “I usually use this brand.” .
“This camera would work well with my computer.” *
“My girl friend would like the design of this camera

(laptop).”

Statements about the performance of the product, restatements
or elaboration of the selling ideas, including participants’ simple
playback and the references to the product class.

e. g.) “This camera has a red eye reduction function.”
“This laptop has two USB ports.”
“It is overpriced compared to other brands with similar
specs.”

All other statements not categorized in one of the above two
categories.

 

Specifically, coders were instructed to proceed hierarchically: they considered

whether a listed thought belonged to the self-thought category, or the product-thought

category, or, finally, irrelevant thought category (Shavitt and Brock 1986). When a listed

thought included both self- and product-thought, it was classified as a self-thought.

Finally, the number of self- thoughts and the number of product-thoughts were summed

for each participant to form an index for the degree of elaborated thinking about the

product for the participant. This summed index was used in the estimation of the

hypothesized effects of cognitive response on its presumed consequent variables in

44

Figure 1.

The judges coded a total of 2,1 1 8 thoughts (1061 thoughts for the IBM laptop
computer; 1057 thoughts for the Kodak digital camera) and agreed on 78% of the
classifications. The researcher resolved disagreements. Table 4 and 5 shows the
descriptive statistics for product-relevant thought by experimental conditions. It should
be noted that several participants recorded no product-relevant thought: ten participants
in the 3-D condition for both products (8.8% of participants per product condition) and
four participants in the 2-D condition for the Kodak digital camera (3.5% of participants)
recorded “0” product-relevant thought. Here, the response “0” does not mean the
participant did not record any thought. It simply means the participant did not record any
self- or product-thoughts, as operationalized in Table 3. In fact, when considering all

thought categories, no participant recorded “0” thought.

Table 4. Descriptive Statistics for Product-relevant Thought

 

 

(IBM laptop computer)
Mean SD. Min. Max.
2-D 5.24 (6.79) 2.11 (2.35) 1 (2) 12 (14)
3-D 4.16 (6.21) 2.60 (2.28) 0 (2) 14 (15)

 

Note: N = 112 for the 2-D condition, N = 114 for the 3-D condition; Numbers in parentheses refer to
when all the three thought categories are considered.

Table 5. Descriptive Statistics for Product-relevant Thought

 

 

(Kodak digital camera)
Mean SD. Min. Max.
2-D 4.90 (6.46) 2.59 (2.32) 0 (2) 14 (14)
3-D 4.41 (6.40) 2.96 (2.53) 0(1) 14 (14)

 

Note: N = 113 for the 2-D condition, N = 114 for the 3-D condition; Numbers in parentheses refer to
when all the three thought categories are considered.

45

""_‘_-l

Mental Imagery. Measures of mental imagery were adopted from prior research
(e.g., Babin and Burns 1997; Bone and Ellen 1982) and measured the degree of vividness
of a mental image. Although mental imagery is a multidimensional concept,
encompassing vividness, quantity, ease, and ability to manipulate images (see Bone and
Ellen 1982 for a review), vividness dimension of imagery seems most appropriate in the
context of the present study examining the effect of 3-D induced imagery on emotions.
To measure imagery vividness, a five-item seven-point Likert scale was used that asked
how clear, vivid, intense, lifelike, and sharp the images are (Bone and Ellen 1982).

Emotional Response. A five-item seven-point semantic differential scale was
used to measure pleasure and arousal in response to mental imagery. Pleasure and arousal
have been identified as the two major components of emotions relevant to consumption
experience such as product use and trials (Havlena and Holbrook 1986; Kempf and Smith
1998; Mano and Oliver 1993). The items included: “Happy — Unhappy,” “Pleased —
Annoyed,” “Excited — Calm,” “Stimulated — Relaxed,” and “Aroused — Unaroused”
(Kempf and Smith 1998; Mano and Oliver 1993).

Information Diagnosticity. Diagnosticity of product information was measured
using a three-item seven-point semantic differential scale. The first two items asked, “In
judging the quality and performance of the product, overall, how helpful would you rate
product-related information you obtained during exposure to the web site?” and “In
understanding different features of the product presented, overall, how helpful would you
rate the product-related information you obtained during exposure to the web site?”
bounded by “Not helpful at all - Extremely helpful.” The last item asked, “To what extent

did product-related information you obtained during exposure to the web site enable you

46

to accurately evaluate the product presented?” The end points were labeled “It did not
enable me to evaluate the product at all” and “It fully enabled me to evaluate the
product.” The midpoint was labeled “It somewhat enabled me to evaluate the product.”
The first and third items were adopted from Kempf and Smith (1998) and modified to fit
the context of this study. The second item was newly developed for the study.

Attitude Accessibility. Attitude accessibility was operationalized as the latency of
response to an attitudinal inquiry (Fazio 1986, 1990). A computer program was developed
to measure response latencies (i.e., reaction time). Participants were told that a series of
brand names fi'om various product categories will appear on a computer screen and that
they need to press one of two keys on the keyboard, “2” for “like” or “I” for “dislike” to
indicate their feelings about each brand (Jewell and Unnava 2003). Consistent with
established procedures for measuring response latencies, participants were encouraged to
respond as quickly and as accurately as possible (Fazio 1990; see Appendix 3).

In addition, to avoid confusion resulting from a single brand name being
represented in multiple product categories, the presentation of a brand name was paired
with its product category (e.g., EDDIE BAUER - clothing, DELL - laptop). A two-second
interval separated each trial, and the order of presentation of the brand names was
randomized for each participant to rule out order effects. The program recorded each
response (like or dislike) and the response latency in milliseconds (i.e., the duration of
time between the appearance of the brand name and the response). To familiarize
participants with the test procedure, participants performed ten trials involving brand
names different from those used in the actual latency test, with an extra opportunity to

repeat the trial session, if they wanted. Afier the trial session, participants were asked to

47

respond to 20 brand names, including two target test brands and 18 filler brands.
Once collected, raw latency scores were inspected for two kinds of possible errors.
First, the participant’s dichotomous response (like or dislike) during the latency test was
compared with his earlier response on a separate brand attitude scale. Specifically, prior
to the response latency test, participants were asked to rate their feelings about the two
test products on a four-item seven-point semantic differential scale, bounded by “Bad -
Good,” “Dislike quite a lot - Like quite a lot,” “Unpleasant - Pleasant,” and “Poor quality
- Good quality” (Gardner 1985; Machleit and \Vrlson 1988). These items were highly
correlated each other (Cronbach’s alpha = .91 for the IBM laptop computer; Cronbach’s
alpha = .93 for the Kodak digital camera), thus they were summed and averaged to
construct a single composite index for brand attitude. The participant’s score on this
composite index was then compared with his or her response during the computerized
latency test. An error was defined as the participant’s responding “like” during the latency
test but rating the product negatively below the neutral point on the attitude scale (i.e.,
“4”), or responding “dislike” during the latency test but rating the product positively
above the neutral point on the scale (Fazio, Powell, and Williams 1989). At the end,
seven surveys from the IMB laptop computer condition and six surveys fi'om the Kodak
digital camera condition were deleted due to such response error. It is a common practice
to limit the analysis to those latencies associated with a given response (Fazio 1990).
Second, the participant’s latency scores for the two target brands (one
representing a utilitarian product and the other representing a hedonic product) were
adjusted to control for error variance due to individual differences in general responding.

To standardize the latencies across participants, each participant’s latency score to the test

48

 

brand was divided by the mean of latency scores for the rest of 18 brands (test
latency/mean filler latency) (F azio 1990; Jewell and Unnava 2003).

Attitude Confidence. Attitude confidence was measured by a three-item seven-
point semantic differential scale, adopted from Berger and Mitchell (1989) and White,
Tashchian, and Ohanian (1991). Participants were asked to rate their certainty that their
judgments about the two test products were accurate on a three-item seven-point semantic
differential scale, bounded by “Not confident - Confident,” “Not sure - Sure,” and

“Uncertain - Certain.”

4.5.2. Scales of Exogenous Control Variables

Ability to Imagine. Given that individual differences in imagining ability may
affect the imagery evoked (see MacInnis and Price 1987 for a review), ability to imagine
needs to be controlled for mental imagery. Ability to imagine was measured using
Sheehan’ s (1967) shortened version of Betts’ (1909) Questionnaire Upon Mental
Imagery (QMI).

Product Category Experience. Product category experience was measured and
used as a control for cognitive elaboration and attitude confidence. Personal experiences
with a given product category may affect the degree of cognitive elaboration via
increased domain knowledge (Celci and Olson 1988). Ability to process information such
as domain knowledge has been found to increase cognitive elaboration (Petty and
Cacioppo 1986). Product category experience may also positively correlate with levels of
attitude confidence because they may serve as a valid cue (e.g., perceived expertise)

when the consumer assesses whether his or her confidence toward a product from the

49

category is correct.

Product category experience was measured using self-reports of amount of
product search, usage, and ownership. This use of search, usage, and ownership as
indicators of product category experience is consistent with Alba and Hutchinson’s
conceptualization of product experience (Alba and Hutchinson 1987). Amount of search
and usage was measured on nine-point items. Ownership was a dichotomous measure of
laptop computer and digital camera ownership at the time of data collection.

Brand Familiarity. Brand familiarity was measured as a control for mental
imagery, attitude accessibility, and attitude confidence. Brand familiarity needs to be
controlled because it may affect both the ability to imagine and the vividness of the
imagery (Burns, Biswas, and Babin 1993; MacInnis and Price 1987). Increased levels of
consumers’ familiarity with a given product have also been found to result in increased
attitude accessibility and attitude confidence (e.g., Berger and Mitchell 1989; Fazio and
Zanna 1978a, b). Brand familiarity was measured using a three-item seven-point
semantic diflerential scale, bounded by “Unfamiliar - Familiar,” “Inexperienced -

Experienced,” and “Not knowledgeable - Knowledgeable.”

5. Data Analysis

Manipulations of the presentation mode and product type were checked by
AN OVA and chi-square tests. Prior to testing the proposed research model, AN COVA
tests were performed to examine the effects of the presentation mode and product type
manipulations on attitude accessibility and attitude confidence, with product category

experience and brand familiarity as two varying covariates. The result of the AN COVA

50

tests helped determine the appropriateness of testing the proposed research model, by
demonstrating ultimate effects of the presentation mode manipulation on attitude
accessibility and confidence. Next, the proposed research model was tested. Testing of
the proposed research model itself was two-fold: the assessment of the measurement
model and the assessment of the structural model. AMOS 4.01 , a structural equation
modeling (SEM) program, with maximum-likelihood (ML) parameter estimation method
was employed. All the constructs of the study were represented as latent variables.

The use of latent variable SEM method in this study is appropriate for several
reasons. Firstly, since the impetus behind this research was to examine the causal
processes by which 3-D presentation influences attitude accessibility and confidence, the
proposed model should be inspected for evidence of substantial and continuous paths
from presentation mode to the two constructs. SEM is appropriate for situations where
examination of causal relationships between multiple variables is the focus of the study
(Anderson and Gerbing 1988; Bollen 1989). Secondly, SEM allows for testing a causal
model with multiple final dependent variables. As attitude accessibility and attitude
confidence constitute two final variables in the proposed research model, SEM can test
this model by allowing their disturbance terms to covary (see Chaudhuri and Holbrook
20001 for a similar example).

Thirdly, most items of this research reflect not only the construct they are
intended to measure, but also random and systematic measurement error. In multiple
equation systems with multiple variables, measurement error can inflate or deflate
estimates of the relations between the variables, thus leading to both Type I and Type II

errors (MacKenzie 2001). In this regard, latent variable SEM appears the right choice

51

because it considers measurement error when assessing the structural relationships
between the variables (Kline 1998). Moreover, latent variable SEM has also been used
for experimental data (Bagozzi and Yr 1989), and has the potential to fundamentally
improve experimental research in the field of consumer behavior (MacKenzie 2001).
Testing of the measurement and structural models was performed separately for
each test product. As a result, two measurement models (one for the IBM laptop
computer and another for the Kodak digital camera) and two structural models (one for
the IBM laptop computer and another for the Kodak digital camera) were developed.
Following the procedures advocated by Anderson and Gerbing (1988) and Hunter and
Gerbing (1982), model fit of the measurement models was assessed prior to estimating
the proposed structural relations between the variables. This two-step approach is
recommend because it minimizes the potential for “interpretational confounding” of the
constructs that may occur with the use of a one-step approach in which the measurement
and structural models are estimated simultaneously (Anderson and Gerbing 1988, p.418).
Specifically, the initial measurement model was subject to a series of respecifications so
that the final measurement model fits the data. To facilitate the moderation hypotheses
testing of product type, emphasis was placed on establishing two acceptable and
comparable measurement models for each of the two test products (i.e., two equivalent
confirmatory factor analysis models with the same observable items). Once a
measurement model with acceptable fit for both products was established, the researcher
subsequently tested the proposed structural relationships among the constructs. Research

findings following these methods of data analysis are presented in the next chapter.

52

Chapter 5. Results

5.1. Manipulation Checks

The presentation mode manipulation was assessed from participants’ ratings of
interactivity and vividness of the stimulus web sites. Participants were asked to indicate
how interactive and vivid they perceive the web sites on two seven-point Likert scale
items. The items asked “The web site I just visited included various interactive features”
and “The web site I just visited included vivid product description.” A two-way AN OVA
examined the impact of the presentation mode and the product type treatments on the
perceived interactivity and the perceived vividness.

The analysis showed significant main effects of the presentation mode on the
perceived interactivity (F (1 , 219) = 288.21, p < .001) and the perceived vividness (F (1 ,
219) = 49.15, p < .001). The effects were in predicted directions in that, compared to 2-D
participants, 3-D participants perceived the 3-D web sites to be more interactive, (M 20 =
3.15 vs. M30 = 5.81) and more vivid (M20 = 5.01 vs. M30 = 5.89). There were no
significant main or interaction effects involving product type for either perceived
interactivity (F(1, 219) = 1.14, p = .29 for main effect; F(1, 219) =.27, p = .61 for
interaction eflect) or perceived vividness (F(1, 219) = 1.22, p = .27 for main effect; F (1 ,
219) = .36, p = .55 for interaction effect). The researcher interpreted these results as
evidence of successful independent manipulation of presentation mode. Table 6

summarizes the results of the ANOVA test.

53

Table 6. Results of Presentation Mode Manipulation Check

 

 

 

 

Manipulation

Presentation Product
Items Mode (PM) Type (PT) F

2-D 3-D IBM Kodak PM PT PMxPT
Perceived no
Interactivity 3.15 5.81 4.43 4.52 288.21 1.14 .27
Pfrfelved 5.01 5.89 5.41 5.49 49.15‘” 1.22 .36
Vrvrdness

 

Note: N = 221; "*p < .001.

To verify the integrity of the product type manipulation, participants were asked
to categorize the two test products as primarily utilitarian, defined “as useful, practical,
functional, something that helps achieve a goal,” as primarily hedonic, defined “as
pleasant and firn, something that is enjoyable and appeals to the sense,” and as both
utilitarian and hedonic (Dhar and Wertenbroch 2000). As shown in Table 7, the majority
of participants classified the IBM laptop computer as primarily utilitarian (132 of 221
participants, xi» = 116.05, p < .001) and the Kodak digital camera as primarily hedonic
(148 of 221 participants, x2@) = 134.56, p < .001). The researcher interpreted this as

evidence of successful manipulation of product type.

Table 7. Results of Product Type Manipulation Check

 

 

 

Product Type Manipulation
Perceived Function IBM laptop computer Kodak digital camera
Functional 132 8
Hedonic 3 148
Both 86 65
Total 221 221

 

Note: x20) = 116.05, p < .001 for the IBM laptop computer; 12(2) = 134.56, p < .001 for the Kodak
digital camera.

54

5.2. Measurement Model Assessment
The assessment of the measurement models took place in three phases, as follows:
(i) item purification, (ii) scale reliability assessment, and (iii) construct validity

assessment.

5.2.1. Item Purification

Items were purified through the procedure recommended by Anderson and
Gerbing (1984) and Gerbing and Anderson (1988). Except for cognitive elaboration and
attitude accessibility, all seven latent constructs with multiple items in this study - mental
imagery, information diagnosticity, emotional response, attitude confidence, ability to
imagine, product category experience, and brand familiarity - were subject to
confirmatory factor analysis (CFA). Thus, a fully correlated seven-factor CFA model with
27 items was initially fitted to the data for each test product.

Specifically, the item purification process was carried out in two phases. Firstly,
the researcher purified items by assessing their unidirnensionality and their individual
item-construct loading (i.e., factor loading). Achieving unidimensional measurement is a
crucial undertaking in theory testing and development because it allows the most
unambiguous assignment of meaning to the estimated constructs (Gerbing and Anderson
1988). Empirically, unidimensionality is achieved when each item loads on a single factor
and the measurement error terms are uncorrelated (Gerbing and Anderson and 1988).
Secondly, an item with its factor loading less than .70 was considered unreliable and thus

removed from further analysis (Kline 1998).4 Individual item-construct reliability is

 

" Squared factor loadings are termed reliabilities. Thus, items with their factor loadings below .70 have

55

considered adequate when an item has a factor loading greater than .70 on its intended
construct, which implies more shared variance between the construct and the item than
error variance.

After eliminating items with substantial cross-loadings or low loading on their
intended constructs, the final measurement model with 25 out of the 27 original items
reflecting seven constructs, was found to fit the data for both products. Table 8 presents
various fit indices of the final measurement model for the IBM laptop computer and the
Kodak digital camera. SEM literature suggests that when sample sizes are less than 250,
the standard root mean squared residual (SRMR) and one of the other recommended fit
indices such as Tucker Lewis Index (TLI) and Comparative F it index (CFI) should be
reported (Holbert and Stephenson 2002; Hu and Bentler 1999).5 In general, SRMR values
less than .10 (Kline 1998) or close to .09 (Hu and Bentler 1999), along with TLI and CFI
values higher than .90, are considered representative of a well-fitting model (Kline 1998),
with Hu and Bentler (1999) recommending a cutoff point close to .95. The root mean
square error of approximation (RMSEA) was also used with a cutoff point of .06 (Holbert
and Stephenson 2002).

The indices for the final measurement model indicate an adequate fit with the
data: TLI = .94; CFI = .95; SRMR = .05, RMSEA = .06 for the IBM laptop computer;
TLI = .95; CFI = .96; SRMR = .04; RMSEA = .05 for the Kodak digital camera (Table 8).

The model x2 statistics were statistically significant (x2(254)= 435.86, p < .001 for the IBM

 

reliabilities less than .5 (.70 x .70 = .49), which implies that at least 50% of the observed variance is due to
random measurement error.

5 More traditional fit indices such as Goodness of Fit Index (GFI), Adjusted Goodness of Fit Index (AGFI),
Normed Fit Index (N F1), and xz/df ratio, are not reported here because SEM literature warns against use of
such fit indices due to their sensitivity to sample size, model specification, and distributional properties of
the model’s variables (Holbert and Stephenson 2002; Hu and Bentler 1999).

56

laptop computer; X2054): 429.66, p < .001 for the Kodak digital camera). The significant

x2 statistics were expected due to their sensitivity to sample sizes (Kline 1998).

Table 8. Summary of Fit Indices of Final Measurement Models

 

 

Final measurement model 2

(seven-factor model, 25 items) df TLI CFI SRMR RMSEA
IBM laptop computer 435.36’” 254 .94 .95 .05 .06
Kodak digital camera 429.66‘" 254 .95 .96 .04 .05

 

Note: N = 226 for the IBM laptop computer, N = 227 for the Kodak digital camera.
"‘" p < .001.

The descriptive statistics of items and item-construct loadings with t-values are
shown in Table 9 (IMB laptop computer) and Table 10 (Kodak digital camera). As shown
in Tables 9 and 10, all the item-construct loadings were significant (t-values between 6.78
and 19.48 for the IBM laptop computer; t-values between 5.26 and 19.42 for the Kodak
digital camera) and above the benchmark of .70 except for three items for ability to
imagine in both products (LAI3 = .67, LAI4 = .66, LAIS = .47 for the IBM laptop
computer; CAI3 = .67, CAI4 = .65, CA15 = .47 for the Kodak digital camera) and one
item for product category experience for the Kodak digital camera (CPCE3 = .61).
Despite their relatively low item-construct loading, these items are retained because the
scales with these items only serve as control variables in this study, and these items were
not substantially cross-loaded on other constructs, therefore, not undermining

unidimensionality of the measurement model for both products.

57

Table 9. Proposed Measurement Items for Constructs (IBM laptop computer)

Constructs Measured Items

MENTAL
IMAGERY

LMI 1
LMIZ
LM13
LMI4
LMIS

EMOTIONAL
RESPONSE

LERl
LER2
LER3
LER4

INFORMATION
DIAGNOSTICITY

LDIAGI

LDIAGZ

LDIAG3

ATTITUDE
CONFIDENCE

LAC 1
LAC2
LAC3

How would you describe the images that
occurred in your mind while examining
the web site materials? The images were:
Clear

Vivid

Intense

Lifelike

Sharp

Tell us how you felt while you were
viewing the web site.

Happy - Unhappy

Pleased - Annoyed

Excited - Calm

Stimulated - Relaxed

The following items concern your
evaluation about product-related
information obtained during exposure to
the web site. Product-related information
includes any ideas, thoughts, and images
about the product you had while
examining the web site, in addition to the
product information directly obtained
fiom the web site.

In judging the quality and performance of
the product, overall, how helpful would
you rate the product-related information
you obtained?

(Not helpful at all - Extremely helpful)
In understanding different features of the
product, overall, how helpful would you
rate the product-related information you
obtained?

(Not helpful at all - Extremely helpful)
To what extent did the product-related
information you obtained enable you to
accurately evaluate the product?

(It did not enable me to evaluate the
product at all - It firlly enabled me to
evaluate the product)

How certain are you of the accuracy of the
responses you gave regarding your
feelings toward the product?

Not confident - Confident

Not sure - Sure

Uncertain - Certain

58

Mean

4.98
4.69
3.83
4.58
4.65

4.26
4.37
2.99
3.49

4.97

5.07

4.71

5.71
5.71
5.58

SD.

1.17
1.25
1.45
1.39
1.38

0.79
0.88
1.15
1.17

1 .09

1 .06

0.98
0.97
1 .02

Factor Loading

 

Std.

0.85
0.87
0.76
0.80
0.85

0.82
0.83
0.85
0.88

0.83

0.76

0.79

0.92
0.91
0.92

t-value

15.49
16.11
13.05
14.11
15.57

14.57
15.01
15.40
16.31

14.22

12.67

13.41

17.82
17.56
17.87

Table 9. (Cont’d)

 

 

 

Factor Loading
Constructs Measured Items Mean SD. Std. t-value
ABILITY TO Think of some relative or friend whom
IMAGINE you frequently see, considering carefully

the picture that rises before your mind's
eye. And classify the images suggested by
the following questions as indicated
by the degree of cleamess and vividness.
(No image present at all - Perfectly clear
and as vivid as actual experience)

LAII The exact contour of face, head, shoulders 5.21 1.18 0.72 11.11
and body.

LA12 Characteristic poses of head, attitudes of 4.97 1.25 0.74 11.60
body, etc.

LA13 The precise carriage, length of step, etc. 4.31 1.31 0.67 10.25
in walking.

LAI4 The different colors worn in some familiar 4.84 1.47 0.66 10.03
costume.
Think of seeing the following, considering
carefully the picture that comes before
your mind's eye.

LAIS The sun as it is sinking below the horizon. 5.54 1.19 0.47 6.78

PRODUCT Indicate your experience with a laptop
CATEGORY computer in general.
EXPERIENCE

LPCEI How often do you use a laptop computer? 5.65 3.28 0.80 9.09
(9 pt item: Never - Everyday)

LPCE3 Overall, how much time have you spent 3.71 2.57 0.81 9.12
searching for laptop computer?
(9 pt. item: Very little - Very much)

BRAND Indicate your familiarity with IBM laptop

FAMILIARITY computer.
LBFI Unfamiliar - Familiar 3.32 1.92 0.88 16.79
LBF2 Inexperienced - Experienced 3.03 1.79 0.94 18.43
LBF3 Not knowledgeable - Knowledgeable 3.11 1.82 0.96 19.48

 

Note: N= 226; Std.: Standardized; LER5 (Aroused - Unaroused), LPCE2 (Do you currently own a laptop
computer? Yes or No) were deleted during the purification process.

59

Table 10. Proposed Measurement Items for Constructs (Kodak digital camera)

 

 

 

Factor Loading_
Constructs Measured Items Mean 5.1), Std. t-value
MENTAL How would you describe the images that
IMAGERY occurred in your mind while examining the
web site materials? The images were:
CMII Clear 5.35 1.13 0.87 16.39
CM12 Vivid 5.12 1.27 0.90 17.30
CM13 Intense 4.17 1.39 0.83 15.10
CMI4 Lifelike 5.07 1.43 0.84 15.47
CM15 Sharp 5.13 1.34 0.89 17.00
EMOTIONAL Tell us how you felt while you were
RESPONSE viewing the web site.
CERl Happy - Unhappy 5.1 1 1.00 0.91 17.89
CER2 Pleased - Annoyed 5.19 1.06 0.93 18.52
CER3 Excited - Calm 4.00 1.27 0.92 18.20
CER4 Stimulated - Relaxed 4.33 1.34 0.94 18.76
INFORMATION The following items concern your
DIAGNOSTICITY evaluation about product-related

information obtained during exposure to
the web site. Product-related information
includes any ideas, thoughts, and images
about the product you had while examining
the web site, in addition to the product
information directly obtained from the web
site.

CDIAGI 1n judging the quality and performance of 5.24 0.99 0.80 13.88
the product, overall, how helpful would
you rate the product-related information
you obtained?
(Not helpful at all - Extremely helpful)

CDIAGZ In understanding different features of the 5.18 1.11 0.81 14.24
product, overall, how helpfirl would you
rate the product-related information you
obtained?
(Not helpful at all- Extremely helpful)

CDIAG3 To what extent did the product-related 5.10 0.93 0.82 14.46
information you obtained enable you to
accurately evaluate the product?
(It did not enable me to evaluate the
product at all - It fully enabled me to
evaluate the product)

ATTITUDE How certain are you of the accuracy of the
CONFIDENCE responses you gave regarding your feelings

toward the product?

CACI Not confident - Confident 5.88 0.90 0.91 17.82

CAC2 Not sure - Sure 5.82 0.97 0.95 19.22

CAC3 Uncertain-Certain 5.77 1.01 0.93 18.18

60

 

Table 10. (Cont’d)

Constructs Measured Items

ABILITY TO
IMAGINE

CAII
CA12
CAI3

CA14

CA15

PRODUCT
CATEGORY
EXPERIENCE CPCEI

CPCE3

BRAND
F AMILIARITY

CBFl
CBF2
CBF3

Think of some relative or friend whom
you frequently see, considering carefully
the picture that rises before your mind's
eye. And classify the images suggested by
the following questions as indicated

by the degree of cleamess and vividness.
(No image present at all - Perfectly clear
and as vivid as actual experience)

The exact contour of face, head, shoulders
and body.

Characteristic poses of head, attitudes of
body, etc.

The precise carriage, length of step, etc.
in walking.

The different colors worn in some familiar
costume.

Think of seeing the following, considering
carefully the picture that comes before
your mind's eye.

The sun as it is sinking below the horizon.
Indicate your experience with a digital
camera in general.

How often do you use a digital camera?
(9 pt item: Never - Everyday)

Overall, how much time have you spent
searching for digital camera?

(9 pt. item: Very little - Very much)
Indicate your familiarity with Kodak
digital camera.

Unfamiliar - Familiar

Inexperienced - Experienced

Not knowledgeable - Knowledgeable

Mean

5.21

4.96

4.32

4.85

5.54

4.54

3.80

3.61
3.01
3.10

SD.

1.18

1.25

1.30

1.47

1.19

1.95

2.04

2.01
1.76
1.79

Factor Loading

 

Std. Mean

0.73

0.73

0.67

0.65

0.47

0.85

0.61

0.81
0.96
0.95

11.39
11.47
10.28

9.91

6.66

5.75

5.26

14.58
19.42
19.00

Note: N= 227; Std.: Standardized; CER5 (Aroused - Unaroused), CPCE2 (Do you currently own a digital
camera? Yes or No) were deleted during the purification process.

61

5.2.2. Scale reliability

Since latent constructs with multiple items are reflective, the assessment of the

measurement model includes the estimation of internal consistency for reliability (Bollen

1989). Thus, all seven latent constructs with multiple items in this study were tested for

internal consistency, using Cronbach’s alpha coefficient. The Cronbach’s alpha

coefficients for all the constructs were higher than the recommended cutoff point of .70

(Nunnally 1978). The alpha coefficient for product category experience for the Kodak

digital camera fell just below the cutoff point. Table 11 shows the descriptive statistics of

the constructs and Cronbach’s alpha coefficients of the scales for both products.

Descriptive statistics of cognitive elaboration and attitude accessibility are also shown in

the table.

Table 11. Descriptive Statistics and Cronbach’s Alphas for Constructs

 

 

Number IBM laptop computer Kodak digital camera
Constructs fIt

0 ms Mean SD. a Mean SD. o
Cognitive Thought
Elaboration listing 4.70 2.43 - 4.66 2.79 -
Mental Imagery 5 4.55 1.14 .91 4.97 1.17 .93
I”.f°"“a".°‘.‘ 3 4.92 .94 .84 5.17 .89 .85
D1agnost1crty
Em°“°“a' 4 3.78 .89 .90 4.66 1.10 .96
Response
Attitude Reaction
Accessibility time method '88 '28 - '81 '23 '
Attitude
Confidence 3 5.67 .94 .94 5.82 .91 .95
Ab'l‘iym 5 4.55 1.14 .78 4.98 .94 .78
Imaglne
Product
Category 2 4.68 2.66 .77 4.17 1.74 .68
Experience
B’a'id. . 3 3.15 1.76 .95 3.24 1.74 .93
Familiarity

 

Note: N = 226 for the IBM laptop computer; N = 227 for the Kodak digital camera.

62

5.2.3. Construct Validity

Construct validity of the measurement model was examined by convergent
validity and discriminant validity. Construct validity is defined as “the extent to which an
operationalization measures the concept it is supposed to measure” (Bagozzi, Y1, and
Phillips 1991, p. 421). Convergent validity refers to the extent to which the items for a
construct act as if they are measuring the underlying construct because they share
variance (Schwab 1980), whereas discriminant validity refers to the degree to which
items of the two constructs are empirically distinct (Bagozzi, Y1, and Phillips 1991).
Empirically, convergent validity is shown when each estimated factor loading on its
intended construct is greater than twice its standard error (Anderson and Gerbing 1988).
Each factor loading in the measurement model was greater than twice its associated
standard error for both IBM laptop computer (.04 - .29) and Kodak digital camera (.05 -
.29), suggesting convergent validity.

Discriminant validity is shown when the confidence interval around the estimated
correlation between any given two constructs does not include 1.0 (Anderson and
Gerbing 1988). The measurement model passed this criterion for both products,
suggesting discriminant validity. Further, a more stringent discriminant validity test was
performed using a constrained analysis method (Sharma 2000), which involves setting
the correlation between one pair of constructs (e. g., mental imagery and emotional
response) to unity (1.0) and running the measurement model again. A x2 difference test is
used to compare the results from the constrained and original model. Discriminant
validity is evidenced if the x2 difference is significant, thereby, supporting the original

model (Anderson and Gerbing 1988). Following this procedure, a total of 21 different

63

constrained models were run, each time setting to unity the correlation between one pair
of the seven constructs. For all 21 pairs, the x2 difi‘erence was significant (p < .01),

supporting discriminant validity of the measurement model.

5.2.4. CFA for Attitude Accessibility and Attitude Confidence

An additional CFA was performed to see whether attitude accessibility and
attitude confidence were indeed two distinct constructs in this study. The CFA showed
that the two factor measurement model fits the data adequately for both products: x20, =
1.95,p = .38; TLI = 1.00; CFI = 1.00; SRMR = .01, RMSEA = .00 for the IBM laptop
computer; x29, = .54, p = .76; TLI = 1.00; CFI = 1.00; SRMR = .01; RMSEA = .00 for
the Kodak digital camera. To further test for the appropriateness of the two factor
measurement model, a constrained analysis was conducted, which involved setting the
correlation between attitude accessibility and attitude confidence to unity (1.0) and
rlmning the model again (Sharma 2000). For both products, the model fit indices
worsened significantly (SRMR = .37 for the IBM laptop computer; SRMR = .43 for the
Kodak digital camera) and the x2 difference tests confirmed (p < .001) that the original
two factor model is a significantly better fit than a single factor model combining
accessibility and confidence.

The two factor measurement model was firrther tested for the parallelism theorem.
The parallelism theorem states that the correlation between two items for two separate

constructs is the product of their factor loadings and the correlation between the

64

constructs (Hunter and Gerbing 1982; Gerbing and Anderson 1988).6 For this test, the
correlation matrix was computed first (Tables 12 and 13). This observed correlation
matrix was then compared with the predicted correlation matrix (Tables 14 and 15),

which was constructed using the internal consistent theorem7 and the parallelism theorem.
This procedure was repeated for each product. In Tables 14 and 15, predicted correlations
are in the lower triangle of the matrix and residuals (the difference between the observed

and predicted correlations) are presented in the upper triangle.

Table 12. Observed Correlations (IBM laptop computer)

 

LACl LAC2 LAC3 LRT
LACl 1.00
LAC2 .84 l .00
LAC3 .84 .84 1 .00
LRT -.44 -.39 -.42 1.00

 

Note: N = 226; LACI, LAC2, and LAC3 are attitude confidence items for the IBM laptop computer; LRT
is reaction time score for the IBM laptop computer.

Table 13. Observed Correlations (Kodak digital camera)

 

CACl CAC2 CAC3 CRT
CACI 1.00
CAC2 .87 1.00
CAC3 .85 .89 1.00
CRT -.31 -.30 -.29 1.00

 

Note: N = 227; CACI, CAC2, and CAC3 are attitude confidence items for the Kodak digital camera; CRT
is reaction time score for the Kodak digital camera.

 

‘5 r XiYk = r XiTrTU rUYk. where Xi is an indicator of the construct T and l’k is an indicator of the construct U.
7 rXin = r XiT r X173 where Xi and Xj are two alternate indicators of the construct T.

65

Table 14. Predicted Correlations and Residuals (IBM laptop computer)

 

LAC 1 LAC2 LAC3 LRT
LAC l .00 .00 -.02
LAC2 .84 .00 .03
LAC3 .84 .84 .00
LRT -.42 -.42 -.42

 

Note: N = 226; Factor correlation = -.46; Predicted correlations are in the lower triangle and residuals are
in the upper triangle of the matrix.

Table 15. Predicted Correlations and Residuals (Kodak digital camera)

 

CACI CAC2 CAC3 CRT
CACl .00 .00 -.02
CAC2 .87 .00 .00
CAC3 .85 .89 .01
CRT -.29 -.30 -.30

 

Note: N = 227; Factor correlation = -.32; Predicted correlations are in the lower triangle and residuals are
in the upper triangle of the matrix.

As shown in Tables 14 and 15, residuals are not substantial in both product
conditions. Compared with the predicted correlations, all observed correlations were
within the 95% confidence interval.8 Again, the results confirm that the two factor
measurement model is consistent with the data adequately for both products.

Taken together, the above CFA results indicate that attitude accessibility and
attitude confidence were distinct constructs in this study, which is consistent with prior
research (Berger 1992; Krosnick et a1. 1993). Therefore, attitude accessibility and attitude

confidence were represented as separate constructs for structural model assessment.

 

l-r2
8 Sr = , where Sr is the standard error for the predicted correlation.
J N — 1

 

66

5.3. Effects of Presentation Mode on Attitude Accessibility and Attitude

Confidence

Prior to testing the structural relationships among the constructs, two-way
AN COVAs were performed to examine the effects of the presentation mode and product
type manipulations on attitude accessibility and attitude confidence with product category
experience and brand familiarity as two varying covariates. Test results helped determine
the appropriateness to test the proposed causal paths among constructs by SEM. For
analysis, summated scores were computed and used for latent constructs with multiple
items.

Controlling for the two covariates, presentation mode had significant main effects
on both attitude accessibility (F(1, 215) =21.52, p < .001, I72 = .09) and attitude
confidence (F(1, 215) = 33.90, p < .001, If = .14). The effects were in predicted
directions in that, compared to 2-D participants, 3-D participants accessed their attitude
faster (M20 = .910 vs. M 30 = .790) and held their attitude more confidently (M20 = 5.46 vs.
M 30 = 6.01). Unexpectedly, the product type main effect on attitude accessibility was also
significant (F(1, 215) = 7.17, p < .01, 17’ = .03), with participants exposed to the Kodak
digital camera holding more accessible attitude than those exposed to the IMB laptop
computer (M1314 = .881 vs. MKODAK = .819), controlling for the two covariates. Although
not a formal hypothesis here, it is clear that more work is needed to understand what has
brought this result. The product type main effect on attitude confidence was not
significant (F (1, 215) = .13, p = .72), controlling for the two covariates. No significant
presentation mode by product type interaction effects were found on either attitude

accessibility (F(1, 215) = .297, p = .59) or attitude confidence (F(1, 215) = 2.35, p = .13).

67

Table 16 shows adjusted means from the analyses.

Table 16. Adjusted Means for Attitude Accessibility and Attitude Confidence

 

 

 

Attitude Accessibility Attitude Confidence
2-D 3-D 2-D 3-D
IMB laptop computer .936 .827 5.43 5.89
Kodak digital camera .886 .753 5.49 6.14

 

Note: N = 221; Means adjusted with product category experience and brand familiarity as varying
covariates.

5.4. Structural Model Assessment

Following the development of the measurement model adequately fitted to both
products and the confirmation of the basic assumption behind this research - 3-D online
presentation (and consequent virtual experience) entails more accessible and confident
attitude, compared to 2-D online presentation, the researcher then estimated the model
parameters along with the fit of the proposed structural model and tested the research
hypotheses. Testing of the structural model was conducted separately per product type,
using AMOS 4.01. The researcher found no normality problem in the data (Appendix 4).

Specifically, the structural model assessment was performed as follows. First, all
constructs in the model were represented as latent variables. Cognitive elaboration and
attitude accessibility were represented as single-item latent variables with no
measurement error because those constructs were measured via thought listing and
reaction time, respectively. For ease of interpretation, reaction time scores were reverse-
coded by multiplying by —1 so that a positive path coefficient of attitude accessibility

with its antecedent means that the antecedent variable leads to more accessible attitude

68

(please recall that a small response latency means more accessible attitude). In addition,
presentation mode, one of the two experimental factors, was also represented as a single-
item construct with no measurement error. Presentation mode was dummy coded: “0” for
2-D, “1” for 3-D. Second, all paths from three exogenous control variables to six
endogenous variables were allowed in the model. The researcher also allowed all
exogenous variables to covary with one another. Finally, the paths among key model
variables were added as hypothesized in the proposed research model. To maintain
empirical interpretability of the model, the researcher allowed no path from indicators to
the constructs that were not intended to reflect.

The fit indices of the proposed research model indicate an adequate fit with the
data: TLI = .93, CFI = .94, SRMR = .05, RMSEA = .06 for the IBM laptop computer;
TLI = .95, CFI = .96, SRMR = .04, RMSEA = .05 for the Kodak digital camera (Table

17). The model )8 statistics were statistically significant (X2014): 552.07, p < .001 for the

IBM laptop computer; x2(314)= 522.71 , p < .001 for the Kodak digital camera), which are

expected given the relatively large sample sizes (Kline 1998).

Table 17. Summary of Fit Indices of Proposed Research Model

 

Proposed Research Model )8 df TLI CFI SRMR RMSEA
IBM laptop computer . 552.07." 314 .93 .94 .05 .06
Kodak digital camera $22.71” 314 .95 .96 .04 .05

 

Note: N = 226 for the IBM laptop computer, N = 227 for the Kodak digital camera.
"“" p < .001.

Figures 2 and 3 show the results of AMOS analysis for the proposed research

model with factor loadings and standardized path coefficients with p-values for each

69

product. All factor loadings are statistically significant, ranging from .48 to .96 for the
IBM laptop computer and from .47 to .97 for the Kodak digital camera. For the IBM
laptop computer, the proposed research model explained 35 % of the variance in attitude
accessibility and 35% of the variance in attitude confidence. For the Kodak digital
camera, the research model explained 32% of the variance in attitude accessibility and
63% of the variance in attitude confidence. Notably, despite comparable R2 values for
attitude accessibility between the two products, there was a substantial difference in R2
values for attitude confidence. Possible reasons for the difference in the explanatory

power of the model explaining the variance in confidence are presented in discussion.

70

co." «mm—2m ”3." gm so." in ”no." 5p :8. v a .233 u 35%

:5. V Q2: .8. V R:
5829:: con E52? 8: 2a 8:8 3593a? was 333.33 3.580 mucoEoSmaoE Seton firs 833.35 883 ESTo-mfim ma con—.823.—

fla b___£mmooo< 33:2. 93 £33.32”.— o>Euon .032 83:82.— 58qu8 @392 mS: new 08E. :28va .53 Aeneas—8 gan— 3 com 95%.—
aueefiv HS Accesses: ace: eeeseooeb 2?. sea consciences sages as. cease can ease Eocene canoes an: Bow ncautz “scz

n—Es—

           

 

     

           

 

 

En.— 23 En.— 324 «:5
3E.— asn :. e». S. :2.—
mm.
8...
ooonooooocoo.nooooooooooooooooooooooooo 8:38“ Eva's—
“US OIOOOOOOOOOOOOOOOOOO. —“--°mfloam .’.”bo -3=°:
8.-
fit in”.
3:06:50 mm.
x o _ a:
m ca.3< seenm. —
u use:
m .Eozmeauaa 538.38...— ESE
n Hanan—Fab:—
u :53.
m bm=nm§90< GB. Who a. no... .— ~. V"..-
.u 8.53. 853 .2.
m 35:.— GSB
m _ 8:225
m b: :2. 2:396
S.
_
HS

€859.39 n39.1— 526 .352 5.808% uoneaeum ..8 are: mOE Evan—=35 me 338% .N ”..:—ME

71

mo.“ aim—)2 ”we.” £32m 6o." EU ”no. u 1:. :8. v minmfi u @3an

:5. v a t... .3. v a.

56:9:6 :8 56% 8: Eu 3:2 85:53:. :5 832:; 3:80 3:082:33: auto: 5:» 83:53 283 EBTanm we 388032

on £58.83. 83?. e5 .eeeecenm ozeemoo see: Seneca: Essa Ease sauce. co: 25 8:685 Eu .Eoeao .sas capo: 6: was:
Ewsozb 4:0 Ease—aiui 3o: 5.5.—32$ 22m 35:: ESE:w_m:o: 88:2: 8:: 3:8 :5 min: Emu—mime. 3:22: 8:: Bow KNNHZ Munoz

 

 
 

 

   
 

3:6 2:6 «Eu
3:6 2:: m. .m. 8. 3. :20
. E.
:.
.tCOODOCIIICICIIOIIOIOIIOIIIIIIIIOOOIOOOI am-honmum remna—
_U<U 35:25 . .552
85 a. tram
ma. resmv.
Na.
nU<U «sew—e.
Sam—EEO me.
a. 85:2 3. r...
e e
u _
. one:
36:8:35 553.58...— ESE
u 53:53.2:
u earl-v.
u b._5_mmooo< 5. . .
I o o o o I he . . . cm on . .I

u 85:2 85:: S ..we 8
m Gino 65::
_ ..... .85—23:...“—
m Eu 8. 2:53:

gt

_
5:.0

Acne—:3 .535 £2.36 .25: £933.3— uomemfia— ..:.— amba:< mOE conga—5:5 a: 338% .m PEEK

72

5.5. Hypothesis Testing

Control Variables

All the relationships described in the following hypothesis testing, except for H1
and H2, were tested with ability to imagine, product category experience, and brand
familiarity as controls. First, as expected, ability to imagine was positively related to
mental imagery for both the IBM laptop computer (,6 = .12) and the Kodak digital
camera (,6 = .10), but barely approached or just fell below marginal significance for both
products (p = .10 for IBM; p = .14 for Kodak). It appears likely that with a larger sample,
these relationships would become significant. Second, product category experience was
positively related to attitude confidence for the Kodak digital camera (,6 = .20, p < .01),
as expected. The corresponding relationship was also positive for the IBM laptop
computer, but was not significant (,6 = .05, p = .49).

Lastly, brand familiarity was positively related to attitude accessibility for the
IBM laptop computer only (,8 = .16, p < .05). The corresponding relationship in the
Kodak digital camera model also was positive, but was not significant (,6 = .02, p = .68).
The path coefficient from brand familiarity to mental imagery was, at least marginally,
significant and positive for both IBM (6 = .14,p < .10) and Kodak (,6 = .19,p < .01). In
addition, the path from brand familiarity to attitude confidence was positive but trivial for
both IBM (,6 = .06) and Kodak (,6 = .09) and approached marginal significance for only
Kodak (p < .10). Unexpectedly, brand familiarity was positively related to cognitive
elaboration for the IBM laptop computer (fl = .19, p < .01). Although not a formal

hypothesis here, more work is needed for better understanding of this unexpected

73

relationship.
The results of hypothesis testing are presented below. Table 18 summarizes the

results.

HI: Effect of Presentation Mode on Cognitive Elaboration

H1 predicted that 3-D participants will engage in more product-relevant cognitive
responding than 2-D participants. As shown in Table 18, presentation mode was
significantly related to cognitive elaboration for the IBM laptop, but the direction of the
effect was the opposite of what was predicted (,6 = -.24, p < .001),9 indicating that 3-D
participants engaged less in product-relevant cognitive responding than 2-D participants.
For the Kodak digital camera, the path fi'om presentation mode on cognitive elaboration
also was negative, but did not approach significance ()6 = -.09, p = .19). Thus, H1 was

not supported.

H2: Eflect of Presentation Mode on Mental Imager

HZ predicted that 3-D participants will have more vivid mental imagery than 2-D
participants. As expected, presentation mode was positively related to mental imagery for
both the IBM laptop computer (,6 = .35, p < .001) and the Kodak digital camera (,6 = .48,
p < .001), thus supporting H2. Therefore, those in the 3-D condition had more vivid

mental imagery about the product than those in the 2-D condition, regardless of product

type.

 

9 Presentation mode was dummy coded as “0” for the 2-D conditions and “1” for the 3-D conditions.

74

H3: Eject of Mental Imagery an Emotional Responses

H3 hypothesized that vivid mental imagery will evoke higher levels of emotional
responses. As expected, vivid mental imagery elicited stronger emotional responses for
both the IBM laptop computer (6 = .78, p < .001) and the Kodak digital camera (6 = .88,

p < .001). Thus, H3 was supported.

H4: Effect of Cognitive Elaboration on Diagnostitcity of Product Information

H4 predicted that cognitive elaboration will be positively associated with
diagnosticity of product information acquired. Although the signs of the path coefficients
were in predicted directions for both products, they failed to reach significance at the .05
level (6 = .11, p = .08 for the IBM laptop computer; 6 = .04, p = .44 for the Kodak

digital camera). Thus, H4 was n_ot supported.

H5: Eflect of Mental Imagery on Diagnosticity of Information
H5 predicted that mental imagery will be positively associated with diagnosticity
of product information. Results show that mental imagery had a positive efiect on

information diagnosticity for both the IBM laptop computer (6 = .53, p < .001) and the

Kodak digital camera (6 = .75, p < .001), as hypothesized. Thus H5 was supported.

H6a & H 6b: Efl'ecm of Information Diagnosticity on Attitude Accessibility and Attitude
Confidence
This set of hypotheses assumed that higher level of information diagnosticity will

lead to more accessible and more confident attitude. As predicted, information

75

diagnosticity was positively associated with attitude accessibility for both the IBM laptop
computer (6 = .46, p < .001) and the Kodak digital camera (6 = .47, p < .001). As
previously mentioned, reaction time scores were reverse-coded by multiplying by -1 for
the AMOS analysis. Thus, the positive path coefficients suggest that higher level of
information diagnosticity leads to more accessible attitude for both products.
Information diagnosticity was also positively associated with attitude confidence
for both the IBM laptop computer (6 = .53, p < .001) and the Kodak digital camera (6

= .24, p < .05). H6a and H6b were, therefore, supported.

H 7a & H 7b: Moderating Role of Product Type for the Cognitive Elaboration-
Accessibility Link and for the Cognitive Elaboration-Confidence Link

This set of hypotheses predicted that cognitive elaboration will have a stronger
influence on attitude accessibility (H7a) and attitude confidence (H7b) for the utilitarian
product (i.e., the IBM laptop computer) than for the hedonic product (i.e., the Kodak
digital camera). As hypothesized, noticeable differences were found in the relationships
between cognitive elaboration and attitude accessibility, and attitude confidence,
depending on product type.

As hypothesized, a noticeable difference was found in the relationship between
cognitive elaboration and attitude accessibility, depending on product type. Specifically,
the effect of cognitive elaboration on attitude accessibility was significant and positive
only for the utilitarian product (6 = .18, p < .01), whereas the corresponding effect was
not significant for the hedonic product (6 = .09, p = .13), thus providing evidence for

H7a. Product type, however, did not moderate the relationship between cognitive

76

elaboration and attitude confidence. Although the sign of the coefiicient was in the
predicted direction for the utilitarian product, it failed to reach significant at the .05 level
(6 = .10, p = .10). The corresponding relationship was not significant for the hedonic
product (6 = -.02, p = .73). These results suggest that, as far as the relationship between
cognitive elaboration and attitude confidence is concerned, product type is not a

significant moderator. Thus, H7b was [19} supported.

H8a & H8b: Moderating Role of Product Type for the Emotional Response —
Accessibility Link and for the Emotional Response-Confidence Link

This set of hypotheses is in contrast to what the previous set of hypotheses
predicted: they predicted that the effects of emotional response on attitude accessibility
and attitude confidence will be more pronounced for the hedonic product than for the
utilitarian product. There was a noticeable difference in the relationship between
emotional response and attitude confidence, depending on product type. For the hedonic
product, the effect of emotional response on confidence was positive and significant (6
= .49, p < .001), but was not significant for the utilitarian product (6 = -.02, p = .80), thus
providing support for H8b.

The moderating role of product type, however, was not detected for the relation
between emotional response and attitude accessibility. In fact, the paths fiom emotional

response to confidence were not significant for both the utilitarian (6 = -.06, p = .46) and

hedonic products (6 = .11, p = .41). Thus, H8a was M supported.

77

Table 18. Results of Hypothesis Testing

 

 

 

H th _ d Std. Path
Hypothesis Path Description $6353? Coefficient Result
IBM Kodak
Presentation Mode 9
Not
_ ##4' _
H1 Cognitive Elaboration (+) '24 '09 supported
Presentation Mode 9 g,” g"
H2 Mental Imagery (+) .35 .48 Supported
Mental Imagery 9 *u M,
H3 Emotional Response (+) .78 .88 Supported
Cognitive Elaboration Not
H4 9 Diagnosticin (+) '11 '04 supported
Mental Imagery -) and: an
H5 Diagnosticity (+) .53 .75 Supported
Diagnosticity 9 u g * H
H6a Accessibility (+) .46 .47 Supported
Diagnosticity 9 *H ,,
H6b Confidence (+) .53 .24 Supported
Cognitive Elaboration Utilitarian > H
“73 -> Accessibility Hedonic "8 '09 supmm"
H7b Cognitive Elaboration Utilitarian > 10 _ 02 Not
9 Confidence Hedonic ' ' supported
H8a Emotional Response Utilitarian < - 0 6 11 Not
9 Accessibility Hedonic ' ' supported
Emotional Response Utilitarian < _ "H,
H8b 9 Confidence Hedonic '02 '49 Supported

 

Note: *p < .05, "p < .01, mp <. .001

5.6. Decomposition of Effects

All the significant relationships observed were decomposed into direct, indirect,

and total effects for further analysis. One of the main advantages of causal analysis such

as SEM and regular path analysis is that it enables one to decompose the relationship

between any two variables into a sum of direct and indirect paths with some of these

indirect paths being substantially meaningful and others not (i.e., nonsignificance) (Asher

1983)

78

The basic impetus behind this research was to illuminate causal processes by
which the presentation mode manipulation influences attitude accessibility and
confidence. And the SEM analysis results show that the presentation mode manipulation
had no significant direct effects on accessibility and confidence for both products.
Therefore, decomposition analysis was focused on whether differences in mediator
variables initially induced by the presentation mode manipulation would be actually
transmitted to differential levels of accessibility and confidence (i.e., indirect paths from
presentation mode to accessibility and confidence in the proposed research model). To
facilitate comparisons, standardized path coefiicients were used throughout for analysis.
Results are shown in Table 19 and 20.

In Tables 19 and 20, total effects were simply the sum of all the direct and indirect
effects of one variable on another. Indirect effects were the product of relevant path
coefficients. A significance of indirect effects involving three variables (e.g., X 9
Y19Y2) was tested by a Z test with the following formulas (Baron and Kenny 1986;

Kline 1998, p. 150).

axb
SE0,

 

Z:

 

where SE“, = (ESE: + a255,; + 553515:

a = path coefficient between X and Y1
b = path coefficient between Y1 and Y2
SE = standard error

For complex indirect effects with more than four variables, the indirect effect was

taken as significant if all of its component path coefficients are significant (Kline 1998,

79

p. 1 50).

Table 19 summarizes the results of effects decomposition in the IBM laptop
computer model. Here, presentation mode had a significant negative indirect effect on
accessibility through cognitive elaboration. Substantively, this implies 2-D participants
engaged in more elaborated cognitive responding than 3-D participants, which, in turn,
was transformed to more accessible attitude for 2-D participants. This contradicts what
was hypothesized. The size of this indirect effect, however, was trivial (-.04), and its
contribution to 2-D participants’ accessibility was offset completely by a stronger positive
indirect influence of presentation mode on accessibility (.08) through imagery and
diagnosticity, resulting in a positive total effect of presentation mode on accessibility
(.04). The other indirect effects of presentation mode on accessibility and confidence, via
either cognitive elaboration or cognitive elaboration and diagnosticity, were negligible
and not significant.

As expected, mental imagery mediated the effect of presentation mode to
diagnosticity that was a strong predictor for both accessibility and confidence. All the
path coefficients involved in these causal links were substantial and positive, resulting in
significant, positive, indirect effects of presentation mode on accessibility (.08) and
confidence (.10). Mental imagery also mediated the effect of presentation mode on
emotional response that was a nonsignificant predictor for accessibility and confidence.
The nonsignificant relationships between emotional response and accessibility, and
confidence, produced a break in these continuous paths from presentation mode to
accessibility, and to confidence.

The decomposition results for the Kodak digital camera are summarized in Table

80

20. None of the efiects of presentation mode on accessibility and confidence were
mediated by either cognitive elaboration or cognitive elaboration and diagnosticity. On
the other hand, mental imagery mediated the effect of presentation mode on diagnosticity
that was a reliable predictor for both accessibility and confidence. All the path
coefficients involved in these causal links were substantial and positive, resulting in
significant, positive, indirect effects of presentation mode on accessibility (.17) and
confidence (.09). Although the size of the path coefficient from diagnosticity to
confidence (.24) may be questioned as not substantial, it did not break the continuous
path from presentation mode to confidence.

Mental imagery also mediated the effect of presentation mode on emotional
response that was a strong predictor of confidence. All the path coefficients involved in
this causal link were substantial and positive, resulting in a significant, positive, indirect
effect of presentation mode on confidence (.20). The indirect effect of presentation mode
on accessibility via mental imagery and emotional response was not significant, however,

due to the break-up between emotional response and accessibility in that link.

81

._o>o_ no. .I. d a: Eco—baa 8: ”set: m_o>o_ no. u d a: ~5oE:w_m o5 $56526 o5 =< ”8:2

 

 

 

 

 

 

 

 

set: 2. .m.: .m.: .m: 2. bug—En: 58m
4: at: rat: .m: .m: .m.: 85598 bowoueo 6:55
.m: .m: .m: rot: ..o..: n: 255: 2 b=3<
833:3, .9550
.m.: .m: - - - - Coobo 30.5 «ooEo 69:0
85 mo: 3558.":
mm. 2.. - - - - :85 Bob ocean ace:
:6 5555.5—
mm. em. mm. mm. - - acute 58.
act: .m.: - - - - 85:5: 35:25 a? “ooh—o5
mm. vm. - - - - b65556 55585 a? «oobo “ooh—o:—
- - mm. mm. - - Soho 89:0
53%
co. m _ . - co. - - Soho :38.
.m: .m: - - - - 56:5:wmmu 5355.5 :3.» 82:65
a: m _. - at: - - Soho 885
5:80:15 92:50
3. vo. mm. 9. mm. emf Soho :38.
. . . . 85:5: 355:5 on
m : m : - - - -
bowag 3:2: :_> “coho 6265
. . essence accesses a.
2 no t - - - bows—E .858 a? .oobo «ooh—o5
- - mm. 2. - - bowmé 35.: a? “coho 8265
.m: .m: - - - - b63553: 5558:. can
538055 3558 :3 Soto 8265
.m: 3.1 - act: - - 55555 9,258 a? 8265
- - - - mm. em... Hoobo 625
32: 5343555
85358 b25658: 8:55: bmocmofirwflu bomg: 55555 25...!» 53:5...—
o:3_:< 33:3. 3558”: 55585 .852 3:5on
o3£5> 553m

 

£85955 5am:— Em—v .252 aaaomom .555...— ue floor—H moi—2:6:5m a: :eflmmeafieoon .3 25¢.

82

..o>o. mo. u d .: 8::h.:w.8 5: ”8.: ..o>o. no. u a 8:2 8: 8555.8 8: 855550 on. ..< No.02

 

 

 

 

 

 

 

 

 

 

8.: 8.: 8.: 8.: 2. 8.: b.::...8:.. :55
cm. 8.: 8.: 8.: 8.: 8.: 855:5 .0588 S58:
8.: 8.: 8.: 8.: 8.: 8.: o:.w:8. 8 b...:<
5.5.25 .9350
ow. 8.: - - - - cooho .80.... Soho S85
85 89. 35.58.”.
em. 2.. - - - - Cooho .80.... Soho S85
.28: :.: 5585.8.
6. 8. 88. 8:. - - Soho .83
M... 8.: - - - - 85:82 .:5.88o :.> So...::.
a .. cm. - - - - 5.2.858: 53:85.8. :.> Soho S9..::.
- - mm. mm. - - Soho Sth
98:532.
oo. oo. - oo. - - Soho .80...
8.: 8.: - - - - b.8558: 53:85.8. :.> Soh::.
8.: 8.: - 8.: - - Soho Sth
5.85:5 o>... :0
am. N. .. mu. :m. ..:. oo. Soho .88.
. . . 85:89. .::o.58o Q
cm 8 : - - - - bow:8. .858 :.> Soho Soh::.
. . b.o.88o:w:.: 5.88.5.8. Q
ac b. - - - - bow:8. .858 :.> Soho So:.::.
- - NV. 3. - - bow:8. .858 :.> Soho So...::.
8.: 8.: - - - - b.o.8::w:.: 53:85.8. Q
5.5555 o>...:w8 :.> Soho So:.::.
8.: 8.: - 8.: - - :o..:._on.:.o o>...:woo :.> S9..::.
- - - - 3. 8.: Soho S85
o:08 5.45585
oo:o:h5o 8.5.8830: 858.88 b.o.8::w:.: bow:8. 5.85:5 “..:—88> 5.3.5....
o:8.:< o:8.=< .:5.58m 5.88.5.8. .852 3...:on
o...:...:> 2.5%...

 

8585 13%. .: 88:36 5.5:. 5.883. 582.9... ..: 85h... :o8:..:::8m ..: 53.82.853— .cn o...:...

83

Chapter 6. Discussion and Conclusion

This dissertation proposed and tested a dual processing model of virtual
experience, in which discursive processing and experiential processing modes are united
in an integrated framework as they influence online consumers’ brand attitude strength,
specifically, attitude accessibility and attitude confidence. At the operational level,
cognitive elaboration was posited to represent the discursive processing aspect, whereas
mental imagery and imagery-evoked emotions represent the experiential aspects of
virtual experience. Specifically, this research was aimed to provide answers for three
testable propositions: (i) 3-D product presentation online (and consequent virtual
experience) leads more accessible and confident brand attitude than 2-D product
presentation online; (ii) there exist two distinct causal processes underlying those effects;
and (iii) product type (utilitarian vs. hedonic) plays an important role determining the

relative contribution of each process to the two attitude strength dimensions.

6.1. Effects of Virtual Experience on Attitude Accessibility and Attitude
Confidence
Findings lend support for the first proposition. Attitude was significantly stronger
for 3-D participants than 2-D participants in terms of both attitude accessibility and
confidence, regardless of product type. These findings are important, given the proven
impact of these attitude strength dimensions on the attitude-behavior relationship:
accessible and confident attitudes guide actual behavior (Berger and Mitchell 1989; Fazio,

Powell, and Williams 1989; F azio and Zanna 1978a, 1978b; Smith and Swinyard 1983).

84

Taken together, this lends credence to the notion that brand attitudes formed via 3-D
product visualization are readily accessible and held with greater confidence, thereby,
fostering attitudinally consistent purchase behavior, than brand attitudes formed via
product exposure in a static 2-D format.

The causal processes by which 3-D product visualization leads more accessible

and confident attitude are explained in the following section.

6.2. Role of Discursive and Experiential Processing in Mediating Virtual

Experience Effects

Findings also are supportive of the dual processing modes of virtual experience.
Cognitive elaboration, representing discursive processing, mediated the effect of the
presentation mode manipulation to accessibility for the utilitarian product. Cognitive
elaboration also mediated the effect of the presentation mode manipulation to confidence
for the utilitarian product, but this presentation mode — cognitive elaboration —
confidence link failed to reach significance at the .05 level (p = .12). With a larger sample,
it appears likely that this continuous path also becomes significant. These discursive
processes, however, were not present for the hedonic product.

Mental imagery and imagery-evoked emotions, representing experiential
processing, also proved to mediate differences created by the presentation mode
manipulation to confidence for the hedonic product. The path from emotional response to
accessibility, however, was not significant, thus producing a break in this continuous path
from presentation mode to accessibility. It remains unclear what may have caused the null

finding. Clearly, more work is necessary for better understanding of this observation. The

85

causal link representing such experiential processing, however, was not detected for the
utilitarian product. Nevertheless, these results generally support the existence of two
distinct processing modes of virtual experience regarding the strength of brand attitude,
despite some variations depending on product type, which will be discussed in the section
6.3.

It should be noted that the significant relationship between presentation mode and
cognitive elaboration found for the utilitarian product was negative, which was
completely opposite of the predicted effect. This finding was also contrasted with the
nonsignificance of the corresponding relationship for the hedonic product. A possible
explanation exists for this unexpected result. Although speculative, it can be argued that
in the face with various interactive features of the interface, 3-D participants may have
been distracted, with their attention being turned away from processing information in a
more analytical manner. In contrast, 2-D participants had no such a distractor. A
theoretical ground for this speculation can be found in the literature on information
overload (Jacoby, Speller, and Kohn 1974) and limited capacity of information
processing (Lang 2000). According to these theories, an individual’s cognitive capacity in
processing information is limited. Thus, as demand from the “primary” task commanding
attention increases, fewer resources should be allocated to “secondary” tasks, inhibiting
elaborative processing of them. In the context of this study, it is reasoned that, as
participants are drawn to salient 3-D features and immersed into mental imagery, the
amount of capacity allocated to process information in a discursive fashion is reduced,
resulting in less cognitive responding for 3-D participants, than for 2-D participants.

Cost for such reduction in processing capacity should be greater for the utilitarian

86

product in which the consumer’s rational assessment about the frmctional utility of the
product is crucial, than for the hedonic product for which such discursive processing is of
less value. This logic is also consistent with the finding that 3-D presentation is inferior to
2-D presentation in fostering cognitive elaboration when a user approaches a web site to

gather facts rather than to be entertained (Schlosser 2003).

6.3. Moderating Role of Utilitarian and Hedonic Product Types

The final goal of this study was to delineate a boundary condition on the relative
strength of discursive processing vs. experiential processing. For that purpose, an idea of
classifying products into utilitarian and hedonic was introduced as a possible moderator.
Specifically, it was hypothesized that a stronger relationship would exit between
cognitive elaboration and accessibility, and confidence, for the utilitarian product, rather
than for the hedonic product. In contrast, a stronger relationship was expected to occur
between imagery-evoked emotions and accessibility, and confidence, for the hedonic
product, rather than the utilitarian product.

Results lend support for such moderating roles of product type. In the discursive
processing side, cognitive elaboration mediated the effect of presentation mode to
accessibility for the utilitarian product, whereas cognitive elaboration was a nonfactor for
any subsequent processing for the hedonic product. In the experiential processing side,
imagery-induced emotional response was a major determinant of attitude confidence for
the hedonic product, whereas such emotion had no bearing on attitude strength for the
utilitarian product.

It should be noted, however, that experiential processing dominated over

87

 

discursive processing, regardless of product type: mental imagery not only mediated the
effect of presentation mode to emotional response that was a strong predictor of
confidence for the hedonic product, but transmitted the effect of presentation mode to
diagnosticity that was a reliable antecedent to both accessibility and confidence across
product types. The only notable observation in discursive processing was the mediating
role of cognitive elaboration on accessibility in the utilitarian product model, but this
causal link was rather weak, despite its significance. In fact, this discursive process was
offset completely by a positive indirect effect through mental imagery and diagnosticity.
Overall, these findings highlight the importance of experiential processing of virtual

experience with respect to attitude strength, regardless of product type.

6.4. Other Significant Findings

Another major finding of this research is the robustness of information
diagnosticity as a critical mediator for the relation between mental imagery and the
strength of brand attitude, regardless of product type. It is reasoned that mental imagery
enables consumers to vicariously experience the product in imagined episodes, thus
providing an added opportunity for them to elaborate on possible consequences of using
it. This added opportunity to deliberate on pros and cons of the product in use should
render acquired product information more reliable and diagnostic as an inferential basis,
thus making the resulting attitude more accessible and more confident. The results of this
study clearly suggests that information diagnosticity is a central construct in the dual
processing model of virtual experience and should be measured and examined in virtual

experience studies.

88

Another point worth noting is the substantial difference between the two product
types in R2 values of the research model for attitude confidence. The proposed research
model explained approximately twice as much variance in attitude confidence for the
hedonic product (63 %) than for the utilitarian product (3 5%). For the hedonic product,
mental imagery worked through both diagnosticity and emotional response to influence
confidence whereas mental imagery worked through only diagnosticity to affect
confidence for the utilitarian product.

As might be expected, the presence of this additional path for the hedonic product

 

(or the absence of the path for the utilitarian product) appears responsible for the
difference observed in the explanatory power of the research model by product type. In
addition, as shown in Table 20, in the hedonic product model, attitude confidence was
significantly influenced by product category experience - a control variable of the study,
thereby, boosting an overall R2 value of the model. In the utilitarian model, none of the
control variables were significantly related to attitude confidence (Table 19). R2 values
for attitude accessibility were comparable across product types: R2 values were 35% and

32% for the utilitarian and hedonic products, respectively.

6.5. Contributions

This study has several important implications. For advertising researchers, this
study contributes to extant 3-D literature by proposing two dimensions of attitude
strength, attitude accessibility and attitude confidence in particular, as new effectiveness
measures of 3-D product visualization. Coupled with the literature endorsing these

dimensions’ impacts on attitude-behavior consistency, this research suggests the

89

possibility that virtual product experience fostered by interactive and vivid 3-D product
visualization can emulate the outcomes anticipated under direct product experience.

This study also specifies important processing variables of virtual experience with
respect to accessibility and confidence and provides an initial attempt to understand their
structural relationships. Perhaps the most important theoretical contribution of this study
is that this research proposed and empirically tested the validity of the dual processing
model of virtual experience employing products that differ in perceived functionality.
Results generally supported the existence of the modeled relationships: (i) discursive
processing and experiential processing modes indeed operate to mediate the effects of
virtual experience to the strength of attitude, despite the dominance of experiential
processing mode; and (ii) the relative influence of each processing mode on attitude
strength depends on the nature of the product being examined.

For online advertisers and marketers, this study offers a convincing argument for
the viability of this innovative technology. Apparently, inducing a positive brand attitude
is not of itself sufficient to substantially influence purchase behavior. Prior research on
attitude accessibility and confidence strongly suggests that if the goal of advertising is to
affect purchase behavior, advertisers and marketers should be also concerned with the
development of accessible and confident attitudes (Herr and Fazio 1993). Findings from
this study suggest that 3-D product visualization is one way to achieve that objective.

Another important practical implication is the suggestion that any benefits of
virtual experience may be qualified for utilitarian products. For those products, virtual
experience may function to inhibit cognitive responding of users, which eventually makes

their attitude less strong than otherwise would be the case. However, this is not to suggest

90

 

in any way that 3-D visualization needs to be completely discarded for this type of
product. Rather, considering the stronger benefits of 3-D visualization on attitude strength
through experiential processing, a better solution would be to display 3-D products via a
separate link or present them in sequential order after presentation of textual information,

so that potential distraction can be minimized.

6.6. Limitations and Future Research

This study has a few limitations and recommendations elicited from those
limitations. First, there is a possibility that the moderating effects of product type are
confounded by uncontrolled aspects of the test products other than the perceived utility
(i.e., whether the product is utilitarian or hedonic). For example, price was not controlled
across the two product conditions. Price is an important factor in product evaluation (Rao
and Sieben 1992) and in the probability of perceived risk, especially, financial risk
(Grewal, Gotlieb, and Marmorstein 1994). In fact, in this study, the IBM laptop computer
model was priced substantially higher than the Kodak digital camera model ($1399.95 vs.
$299.95) to make the stimulus materials appear more realistic. Although this made the
online ads more real, it may have caused participants to perceive the test products
different in terms of perceived risks, thus creating a possible confound. Future research
might attempt to try to control extraneous variables such as price and risk. Thereby, two
product conditions would differ only in terms of their perceived utility, thus minimizing
the potential sources of extraneous variation.

In addition, the possibility exists that the thought-listing procedure used in this

study may not be totally confound-free. Despite its popularity as a measure of cognitive

91

 

elaboration (Cacioppo and Petty 1979), sometimes whether a listed thought is derived
from a verbally driven cognitive activity or from an imagined episode is unclear
(MacInnis and Price 1987). In other words, a listed thought, although verbally
represented, still may be elicited from a mental image. And to the extent that this happens,
the measure is confounded. Future studies might attempt to try control this extraneous
variation by employing protocol analysis techniques, which can capture participants’
cognitive response concurrently with their exposure to stimuli.

Another limitation of the study is that in this study, product information was
conveyed only through a single sensory dimension — sight. This constraint with respect to
sensory inputs may have mitigated the effects of 3-D product visualization. For example,
adding sound may enhance the effects of 3-D product visualization especially for
products where sound is an important attribute (Schlosser 2003). Future research may
examine the effect of 3-D visualization with varying degrees of sensory inputs (sight only
vs. sight plus sound) on the strength of attitudes.

Another interesting avenue for future research is to examine the effects of virtual
experience in conjunction with textual information. The focus of this study was to
explore how initial differences induced by two online product presentation styles (2-D vs.
3-D) are carried over to the strength of brand attitude. So, both quantity and quality of
textual information in stimulus web sites were controlled across the conditions. However,
considering a critical role of ad copy playing in ad information processing (Petty,
Cacioppo, and Schumann 1983), future research may benefit from incorporating both
quantity and quality of textual information as experimental factors in combination with

online presentation mode. This avenue of research would help clarify possible

92

interactions between them.

In conclusion, while prior studies are concerned about either discursive or
experiential aspects of online experience, there has been a lack of integrative frameworks
that combine the various theories and findings fi'om the two research streams. This study
represents an initial attempt to develop such a framework. Specifically, this study
demonstrated the existence of two distinct processing modes in mediating the effects of
virtual experience to attitude accessibility and confidence and, more importantly,
highlighted the importance of experiential processing. Future research should be
conducted to fill the gaps of the processes left unexplained. By then we will have a more
complete understanding of what can be accomplished by this innovative but unexplored

area of virtual experience.

93

——

REFERENCES

Aaker, David A., Douglas M. Stayman, and Michael R. Hagerty (1986), "Warmth in
Advertising: Measurement, Impact and Sequence Effects," Journal of Consumer
Research, 12 (March), 365-381.

Ajzen, Icek and Martin Fishbein (1980), "Predicting and Understanding Consumer
Behavior: Attitude-Behavior Correspondence," in Understanding Attitudes and
Predicting Social Behavior, Englewood Cliffs, NJ: Prentice-Hall, 148-172.

Alba, Joseph W. and J. Wesley Hutchinson (1987), "Dimensions of Consumer Expertise,"
Journal of Consumer Research, 13 (March), 411-454.

Anderson, James C. and David W. Gerbing (1984), "The Effects of Sampling Error on
Convergence, Improper Solutions, and Goodness-of-Fit Indices for Maximum
Likelihood Confirmatory Factor Analysis," Psychometrika, 49, 155-173.

---- (1988), "Structural Equation Modeling in Practice: A Review and Recommended
Two-Step Approach," Psychological Bulletin, 103 (3), 411-423.

Ariely, Dan (2000), "Controlling the Information Flow: Effects on Consumers' Decision

Making and Preferences," Journal of Consumer Research, 27 (September), 233-
248.

Asher, Herbert B. (1983), Causal Modeling, (2nd ed.), Newbury Park: Sage.

Babin, Barry J ., William R. Darden, and Mitch Griffin (1994), "Work and/or Fun:
Measuring Hedonic and Utilitarian Shopping Value," Journal of Consumer
Research, 20 (March), 644-656.

Babin, Laurie A. and Alvin C. Burns (1997), "Effects of Print Ad Pictures and Copy

Containing Instructions to Imagine on Mental Imagery That Mediates Attitudes,"
Journal of Advertising, 26(3), 33-44.

94

Baddeley, Alan D. (1986), Working Memory, New York: Oxford University Press.

Bagozzi, Richard P. and Youjae Yr (1989), "The Degree of Intention Formation as a
Moderator of the Attitude-Behavior Relationship," Social Psychological
Quarterly, 52 (4), 266-279.

Bagozzi, Richard P., Youjae Yi, and Lynn W. Phillips (1991), "Assessing Construct
Validity in Organizational Research," Administrative Science Quarterly, 36 (3),
421-458.

Baron, Reuben M. and David A. Kenny (1986), "The Moderator-Mediator Variable
Distinction in Social Psychological Research: Conceptual, Strategic, and

Statistical Considerations," Journal of Personality and Social Psychology, 51 (6),
1173-1182.

Bassili, John N. (1993), "Response Latency versus Certainty as Indices of the Strength of
Voting Intentions in a CATI Survey," Public Opinion Quarterly, 57(1), 54-61.

Batra, Rajeev and Olli T. Ahtola (1990), "Measuring the Hedonic and Utilitarian Sources
of Consumer Attitudes," Marketing Letters, 2 (2), 159-170.

Batra, Rajeev and Michael L. Ray (1985), "How Advertising Works at Contact," in
Psychological Processes and Advertising Eflects, Linda F. Alwitt and Andrew A.
Mitchell, eds., Hillsdale, NJ: Lawrence Erlbaum Associates, 13-44.

---- (1986), "Affective Responses Mediating Acceptance of Advertising," Journal of
Consumer Research, 13 (September), 234-249.

Bennett, Peter D. and Gilbert D. Harrell (1975), "The Role of Confidence in
Understanding and Predicting Buyers' Attitudes and Purchase Intentions," Journal

of Consumer Research, 2 (September), 110-117.

Bentler, Peter M. and Chi-Ping Chou (1987), "Practical Issues in Structural Modeling,"
Sociological Methods and Research, 16 (78-117).

95

Berger, Ida E. (1992), "The Nature of Attitude Accessibility and Attitude Confidence: A
Triangulated Experiment," Journal of Consumer Psychology, 1 (2), 103-123.

Berger, Ida E. and Andrew A. Mitchell (1989), "The Effect of Advertising on Attitude
Accessibility, Attitude Confidence, and the Attitude-Behavior Relationship,"
Journal of Consumer Research, 16 (December), 269-279.

Bettman, James R. (1979), "Memory Factors in Consume Choice: A Review," Journal of
Marketing, 43, 37-53.

Betts, George Herbert (1909), The Distribution and Functions of Mental Imagery, New
York: Columbia University Teachers College.

Biocca, Frank (1992), "Communication within Virtual Reality: Creating a Space for
Research," Journal of Communication, 42 (4), 5-22.

Bollen, Kenneth A. (1989), Structural Equations Modeling with Latent Variables, New
York: John Wiley & Sons.

Bone, Paula Fitzgerald and Pam Scholder Ellen (1992), "The Generation and
Consequences of Communication-Evoked Imagery," Journal of Consumer
Research, 19 (June), 93-104.

Brock, Timothy C. and Sharon Shavitt (1983), "Cognitive-Response Analysis in
Advertising," in Advertising and Consumer Psychology, Larry Percy and Arch G
Woodside, eds., Lexington, MA: Lexington Books, 91-116.

Burke, Marian C. and Julie A. Edell (1989), "The Impact of Feelings on Ad-Based Affect
and Cognition," Journal of Marketing Research, 26 (February), 69-83.

Burns, Alvin C., Abhijit Biswas, and Laurie A. Babin (1993), "The Operation of Visual

Imagery as a Mediator of Advertising Effects," Journal of Advertising, 22 (2), 71-
85.

96

Cacioppo, John T. and Richard E. Petty (1979), "Effects of Message Repetition and
Position on Cognitive Responses, Recall, and Persuasion," Journal of Personality
and Social Psychology, 37 (1), 97-109.

---- (1981), "Social Psychological Procedures for Cognitive Response Assessment: The
Thought-Listing Technique," in Cognitive Assessment, Carol R. Glass, Thomas V.
Merluzzi, and Myles Genest, eds., New York: Guilford Press, 309-342.

Celci, Richard L. and Jerry C. Olson (1988), "The Role of Involvement in Attention and
Comprehension Processes," Journal of Consumer Research, 15 (September), 210-
224.

Chaiken, Shelly (1980), "Heuristic versus Systematic Information Processing and the Use
of Source versus Message Cues in Persuasion," Journal of Personality and Social
Psychology, 39 (5), 752-766.

Chaiken, Shelly and Alice H. Eagly (1983), "Communication Modality as a Determinant
of Message Persuasiveness and Message Comprehensibility," Journal of
Personality and Social Psychology, 45 (2), 241-256.

Chaudhuri, Arjun and Morris B. Holbrook (2001), "The Chain of Effects from Brand
Trust and Brand Affect to Brand Performance: The Role of Brand Loyalty,"
Journal of Marketing, 65 (April), 81-93.

Coyle, James R. and Esther Thorson (2001), "The Effects of Progressive Levels of
Interactivity and Vividness in Web Marketing Sites," Journal of Advertising, 30
(3), 65-77.

Dhar, Ravi and Klaus Wertenbroch (2000), "Consumer Choice between Hedonic and
Utilitarian Goods," Journal of Marketing Research, 37 (February), 60-71.

Doll, Jorg and Icek Ajzen (1992), "Accessibility and Stability of Predictors in the Theory

of Planned Behavior," Journal of Personality and Social Psychology, 63 (5), 754-
765.

97

Edell, Julie A. and Marian C. Burke (1987), "The Power of Feelings in Understanding
Advertising Efi'ects," Journal of Consumer Research, 14 (December), 421-433.

Edwards, Steven M. and Harshavardhan Gangadharbatla (2001), "The Novelty of 3D
Product Presentations Online," Journal of Interactive Advertising, 2 (1 ),
[www.iiad.org].

Edwards, Steven M. and Carrie La Ferle (2003), "Role-Taking: Enhancing the Online
Experience," Journal of Current Issues and Research in Advertising, 25 (2), 45-56.

Einhom, Hiller J. and Robin M. Hogarth (1978), "Confidence in Judgment: Persistence of
the Illusion of Validity," Psychological Review, 85 (5), 395-416.

Epstein, Seymour (1994), "Integration of the Cognitive and the Psychodynamic
Unconscious," American Psychologist, 48 (8), 709-724.

Escalas, Jennifer Edson (2004), "Imagine Yourself in the Product," Journal of Advertising,
33 (2), 37-48.

F azio, Russell H. (1986), "How Do Attitudes Guide Behavior?," in The Handbook of
Motivation and Cognitive Foundations of Social Behavior; Richard M. Sorrentino
and E. Tory Higgins, eds., New York: Guilford Press, 204-243.

---- (1990), "A Practical Guide to the Use of Response Latency in Social Psychological
Research," in Research Methods in Personality and Social Psychology, Clyde
Hendrick and Margaret S. Clark, eds., Vol. 11, Newbury Park, CA: Sage, 74-95.

---- (1995), "Attitudes as Object-Evaluation Associations: Determinants, Consequences,
and Correlates of Attitude Accessibility," in Attitude Strength: Antecedents and
Consequences, Richard E. Petty and Jon A. Krosnick, eds., Mahwah, NJ:
Lawrence Erlbaum Associates, 247-282.

Fazio, Russell H., Jeaw-Mei Chen, Elizabeth C. McDonel, and Steven J. Sherman (1982),
"Attitude Accessibility, Attitude-Behavior Consistency, and the Strength of the

Obj ect-Evaluation Association," Journal of Experimental Social Psychology, 18
(July), 339-357.

98

 

Fazio, Russell H., Martha C. Powell, and Paul M. Herr (1983), "Toward a Process Model
of the Attitude-Behavior Relation: Accessing One's Attitude Upon Mere
Observation of the Attitude Object," Journal of Personality and Social
Psychology, 44 (4), 723-735.

Fazio, Russell H., Martha C. Powell, and Carol J. Williams (1989), "The Role of Attitude
Accessibility in the Attitude-to-Behavior Process," Journal of Consumer Research,
16 (December), 280-288.

Fazio, Russell H., David M. Sanbonmatsu, Martha C. Powell, and Frank R. Kardes
(1986), "On the Automatic Activation of Attitudes," Journal of Personality and
Social Psychology, 50(2), 229-238.

Fazio, Russell H. and Carol J. Williams (1986), "Attitude Accessibility as a Moderator of
the Attitude-Perception and Attitude-Behavior Relations: An Investigation of the
1984 Presidential Election," Journal of Personality and Social Psychology, 51 (3),
505-5 14.

Fazio, Russell H. and Mark P. Zanna (1978a), "Attitude Qualities Relating to the Strength
of the Attitude-Behavior Relationship," Journal of Experimental Social
Psychology, 14, 398-408.

---- (1978b), "On the Predictive Validity of Attitudes: The Roles of Direct Experience and
Confidence," Journal of Personality, 46 (2), 228-243.

---- (1981), "Direct Experience and Attitude-Behavior Consistency," Advances in
Experimental Social Psychology, 14, 161-202.

Fazio, Russell H., Mark P. Zanna, and Joel Copper (1978), "Direct Experience and
Attitude-Behavior Consistency: An Information Processing Analysis," Personality

and Social Psychologz Bulletin, 4 (1), 48-51.

Fishbein, Martin and Icek Aj zen (1975), Belief Attitude, Intention, and Behavior: An
Introduction to Theory and Research, Reading, MA: Addison-Wesley.

99

Gardner, Meryl Paula (1985), "Does Attitude toward the Ad Affect Brand Attitude under
a Brand Evaluation Set?," Journal of Marketing Research, 22 (May), 192-198.

Gerbing, David W. and J arnes C. Anderson (1988), "Au Updated Paradigm for Scale
Development Incorporating Unidirnensionality and Its Assessment," Journal of
Marketing, 25 (May), 186-192.

Gerrig, Richard J. (1994), "Narrative Thought?," Personality and Social Psychology
Bulletin, 20 (6), 712-715.

Gibson, James J. (1966), The Senses Considered as Perceptual Systems, Boston:
Houghton Mifflin.

Green, Melanie C. and Timothy Brock (2000), "The Role of Transportation in the
Persuasiveness of Public Narratives," Journal of Personality and Social
Psychology, 79(5), 701-721.

Greenwald, Anthony G. (1968), "Cognitive Learning, Cognitive Responses to Persuasion,
and Attitude Change," in Psychological Foundations of A ttitudes, Anthony G
Greenwald, Timothy C. Brock, and Thomas M. Ostrom, eds., New York:
Academic Press, 147-170.

Grewal, Dhruv, Jerry Gotlieb, and Howard. Marmorstein (1994), "The Moderating
Effects of Message Framing and Source Credibility on the Price-Perceived Risk
Relationship," Journal of Consumer Research, 21 (1), 145-153.

Griffith, David A. and Qimei Chen (2004), "The Influence of Virtual Direct Experience
(V DE) on On-Line Ad Message Effectiveness," Journal of Advertising, 33 (1),
55-68.

Haubl, Gerald and Valerie Trifts (2000), "Consumer Decision Making in Online

Shopping Environments: The Effects of Interactive Decision Aids," Marketing
Science, 19 (1), 4-21.

100

Haugtvedt, Curtis P. and Joseph R. Priester (1997), "Conceptual and Methodological
Issues in Advertising Effectiveness: An Attitude Strength Perspective," in
Measuring Advertising Efiectiveness, William D. Wells, ed., Mahwah, NJ:
Lawrence Erlbaum Associates, 79-93.

Havlena, William J. and Morris B. Holbrook (1986), "The Varieties of Consumption
Experience: Comparing Two Typologies of Emotion in Consmner Behavior,"
Journal of Consumer Research, 13 (December), 394-404.

Heeter, Carrie (2000), "Interactivity in the Context of Designed Experience," Journal of
Interactive Advertising, 1 (1), [http://www.jiad.org].

Hirschmann, Elizabeth and Morris B. Holbrook (1982), "Hedonic Consumption:
Emerging Concepts, Methods and Propositions," Journal of Marketing, 48
(Summer), 92-101.

Hoffman, Donna L., Thomas P. Novak, and Adam Duhachek (2003), "The Influence of
Goal-Directed and Experiential Activities on Online Flow Experience," Journal of
Consumer Psychology, 13 (1/2), 3-16.

Holbert, R. Lance and Michael T. Stephenson (2002), "Structural Equation Modeling in
the Communication Sciences, 1995-2000," Human Communication Research, 28
(4), 531-551.

Holbrook, Morris B. and Rajeev Batra (1987), "Assessing the Role of Consumer
Responses to Advertising," Journal of Consumer Research, 14 (December), 404-
419.

Holbrook, Morris B. and Elizabeth C. Hirschmann (1982), "The Experiential Aspects of
Consumption: Consumer Fantasies, Feelings, and Fun," Journal of Consumer
Research, 9 (September), 132-140.

Hoque, Abeer Y. and Gerald L. Lohse (1999), "An Information Search Cost Perspective

for Designing Interfaces for Electronic Commerce," Journal of Marketing, 36
(August), 387-394.

101

 

Houston, David A. and Russell H. Fazio (1989), "Biased Processing as a Function of
Attitude Accessibility: Making Objective Judgments Subjectively," Social
Cognition, 7 (1), 51-66.

Howard, John A. (1989), Consumer Behavior in Marketing Strategy, Englewood Cliffs,
NJ: Prentice-Hall.

Howard, John A. and J. N. Sheth (1969), The Theory of Buyer Behavior, New York: John
Wiley.

 

Hu, Li-tze and Peter M. Bentler (1999), "Cutofi Criteria for Fit Indexes in Covariance

Structure Analysis: Conventional Criteria versus New Alternatives," Structural
Equation Modeling, 6 (1), 1-55.

Hunter, John E. and David W. Gerbing (1982), "Unidimensional Measurement, Second
Order Factor Analysis, and Causal Models," Research in Organizational Behavior,
4, 267-320.

Jacoby, Jacob, Donald E. Speller, and Carol A. Kohn (1974), "Brand Choice Behavior as
a Function of Information Overload," Journal of Marketing Research, 11 (1), 63-
69.

Jewell, Robert D. and H. Rao Unnava (2003), "When Competitive Interference Can Be
Beneficial," Journal of Consumer Research, 30 (2), 283-291.

Kempf, DeAnna S. (1999), "Attitude Formation from Product Trial: Distinct Roles of
Cognition and Affect for Hedonic and Functional Products," Psychology and
Marketing, 16(1), 35-50.

Kempf, DeAnna S. and Robert E. Smith (1998), "Consumer Processing of Product Trial
and the Influence of Prior Advertising: A Structural Modeling Approach," Journal
of Marketing Research, 35 (August), 325-338.

Kisielius, Jolita and Brian Stemthal (1984), "Detecting and Explaining Vividness Effects

in Attitudinal Judgments," Journal of Marketing Research, 21 (February), 54-64.

102

 

---- (1986), "Examining the Vividness Controversy: An Availability-Valence
Interpretation," Journal of Consumer Research,‘ 12 (March), 418-431.

Klein, Lisa R. (2003), "Creating Virtual Product Experiences: The Role of Telepresence,"
Journal of Interactive Marketing, 17 (1), 41-55.

Kline, Rex B. (1998), Principles and Practice of Structural Equation Modeling, New
York: Guilford Press.

Koriat, Asher, Sarah Lichtenstein, and Baruch Fischofl‘ (1980), "Reasons for
Confidence," Journal of Experimental Psychology: Human Learning and Memory,
6(2), 107-118.

Krosnick, Jon A., David S. Boninger, Yao C. Chuang, Matthew K. Berent, and Catherine
G Carnot (1993), "Attitude Strength: One Construct or Many Related
Constructs?," Journal of Personality and Social Psychology, 65 (6), 1132-1151.

Krosnick, Jon A. and Richard E. Petty (1995), "Attitude Strength: An Overview," in
Attitude Strength: Antecedents and Consequences, Richard E. Petty and Jon A.
Krosnick, eds., Mahwah, NJ: Lawrence Erlbaum Associates, 1-24.

Krosnick, Jon A. and Howard Schuman (1988), "Attitude Intensity, Importance, and
Certainty and Susceptibility to Response Effects," Journal of Personality and

Social Psychology, 54(6), 940-952.

Krugrnan, Herbert E. (1965), "The Impact of Television Advertising: Learning without
Involvement," Public Opinion Quarterly, 29 (3), 349-356.

Lang, Annie (2000), "The Limited Capacity Model of Mediated Message Processing,"
Journal of Communication, 50 (1), 46-67.

Lang, Peter J. (1979), "A Bio-Informational Theory of Emotional Imagery,"
Psychophysiology, 16 (6), 495-512.

103

Lee, Ki-Young and Hairong Li (2005), "Alternative Measures of the Effectiveness of 3-D
Advertising," paper presented at the American Academy of Advertising
Conference, (March 31 - April 3), Houston, TX.

Li, Hairong, Terry Daugherty, and Frank Biocca (2001), "Characteristics of Virtual
Experience in Electronic Commerce: A Protocol Analysis," Journal of Interactive
Marketing, 15 (3), 13-30.

Li, Hairong, Terry Daugherty, and Frank Biocca (2002), "Impact of 3-D Advertising on
Product Knowledge, Brand Attitude, and Purchase Intention: The Mediating Role
of Presence," Journal of Advertising, 31 (3), 43-57.

---- (2003), "The Role of Virtual Experience in Consumer Learning," Journal of
Consumer Psychology, 13 (4), 395-407.

Machleit, Karen A. and R. Dale Wilson (1988), "Emotional Feelings and Attitude toward
the Advertisement: The Roles of Brand Familiarity and Repetition," Journal of
Advertising, 17(3), 27-35.

MacInnis, Deborah J. and Linda L. Price (1987), "The Role of Imagery in Information
Processing: Review and Extensions," Journal of Consumer Research, 13 (March),
473-491.

MacInnis, Deborah J. and Bernard J. J aworski (1989), "Information Processing fi'om
Advertisements: Towards an Integrative Framework," Journal of Marketing, 53
(October), 1-23.

MacKenzie, Scott B. (2001), "Opportunities for Improving Consruner Research through
Latent Variable Structural Equation Modeling," Journal of Consumer Research,
28 (June), 159-166.

Mano, Haim and Richard L. Oliver (1993), "Assessing the Dimensionality and Structure

of the Consumption Experience: Evaluation, Feeling, and Satisfaction," Journal
of Consumer Research, 20 (December), 451-466.

104

Marks, Lawrence J. and Michael A. Kamins (1988), "The Use of Product Sampling and
Advertising: Effects of Sequence of Exposure and Degree of Advertising Claim
Exaggeration on Consumers' Belief Strength, Belief Confidence, and Attitudes,"
Journal of Marketing Research, 25 (August), 266-281.

Mathwick, Charla and Edward Rigdon (2004), "Play, Flow, and the Online Search
Experience," Journal of Consumer Research, 31 (September), 324-3 32.

Meyers-Levy, Joan and Prashant Malaviya (1999), "Consumers' Processing of Persuasive
Advertisements: An Integrative Framework of Persuasion Theories," Journal of
Marketing, 63 (Special Issue), 45-60.

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.

Mitchell, Andrew A. and Jerry C. Olson (1981), "Are Product Attribute Beliefs the Only
Mediator of Advertising Effects on Brand Attitude?," Journal of Marketing
Research, 18 (August), 318-332.

Novak, Thomas P., Donna L. Hoffman, and Yiu-Fai Yung (2000), "Measuring the
Customer Experience in Online Environments: A Structural Equation Approach,"
Marketing Science, 19 (1), 22-44.

Nunnally, Jum C. (1978), Psychometric Theory, (2nd ed.), New York: McGraw-Hill.

Olson, G Jerry, Daniel R. Toy, and Philip A. Dover (1982), "Do Cognitive Responses
Mediate the Effects of Advertising Content on Cognitive Structure?," Journal of
Consumer Research, 9 (December), 245-262.

Petty, Richard E. and John T. Cacioppo (1981), "Personal Involvement as a Determinant
of Argument-Based Persuasion," Journal of Personality and Social Psychology,

41 (5), 847-855.

---- (1986), Attitudes and Persuasion: Classic and Contemporary Approaches, Boulder,
CO: Westview Press.

105

Petty, Richard B, John T. Cacioppo, and David Schumann (1983), "Central and
Peripheral Routes to Advertising Effectiveness: The Moderating Role of
Involvement," Journal of Consumer Research, 10 (September), 135-145.

Petty, Richard E., Curtis P. Haugtvedt, and G A. Rennier (1995), "Elaboration as a
Determinant of Attitude Accessibility," Unpublished manuscript, Ohio State
University, Columbus [cited in Petty, Haugtvedt, and Smith (1995), "Elaboration
as a Determinant of Attitude Strength: Creating Attitudes That Are Persistent,
Resistant, and Predictive of Behavior," in Attitude Strength: Antecedents and
Consequences, Richard E. Petty and Jon A. Krosnick, eds., Mahwah, NJ:
Lawrence Erlbaum Associates, 93-130].

Petty, Richard E., Curtis P. Haugtvedt, and Stephen M. Smith (1995), "Elaboration as a
Determinant of Attitude Strength: Creating Attitudes That Are Persistent,
Resistant, and Predictive of Behavior," in Attitude Strength: Antecedents and
Consequences, Richard E. Petty and Jon A. Krosnick, eds., Mahwah, NJ:
Lawrence Erlbaum Associates, 93-130.

Priester, Joseph R., Dhananjay Nayakankuppam, Monique A. Fleming, and John Godek
(2004), "The AZSC2 Model: The Influence of Attitudes and Attitude Strength on
Consideration and Choice," Journal of Consumer Research, 30 (March), 574-587.

Raden, David (1985), "Strength-Related Attitude Dimensions," Social Psychological
Quarterly, 48 (4), 312-330.

Rao, Akshay R. and Wanda A. Sieben (1992), "The Effect of Prior Knowledge on Price
Acceptability and the Type of Information Examined," Journal of Consumer
Research, 19(2), 256-270.

Rossiter, John R. and Larry Percy (1980), "Attitude Change through Visual Imagery in
Advertising," Journal of Advertising, 9(2), 10-16.

Schlosser, Ann E. (2003), "Experiencing Products in the Virtual World: The Role of Goal

and Imagery in Influencing Attitudes versus Purchase Intention," Journal of
Consumer Research, 30 (September), 184-198.

106

Schwab, Donald P. (1980), "Construct Validity in Organization Behavior," in Research in
Organizational Behavior; Barry M. Staw and Lany L. Cumming, eds. Vol. 2,
Greenwich, CT: JAI Press, 3-43.

Sheikh, Anees A. and Charles S. Jordan (1983), "Clinical Uses of Mental Imagery," in
Imagery: Current Theory, Research, and Application, Anees A. Sheikh, ed., New
York: John Wiley & Sons, 391—435.

Sharma, S. (2000), "Managerial Interpretations and Organizational Context as Predictors
of Corporate Choice of Environmental Strategy," Academy of Management
Journal, 43 (4), 681-697.

Shavitt, Sharon and Timothy C. Brock (1986), "Self-Relevant Responses in Commercial
Persuasion: Field and Experimental Tests," in Advertising and Consumer
Psychology, Keith Sentis and Jerry Olson, eds., Vol. 3, New York: Praeger, 149-
171.

Sheehan, Peter W. (1967), "A Shortened Form of Betts's Questionnaire Upon Mental
Imagery," Journal of Clinical Psychology, 23, 386-389.

Smith, Robert E. and William R. Swinyard (1982), "Information Response Models: An
Integrated Approach," Journal of Marketing, 46 (Winter), 81-93.

---- (1983), "Attitude-Behavior Consistency: The Impact of Product Trial versus
Advertising," Journal of Marketing Research, 20 (August), 257-267.

---- (1988), "Cognitive Response to Advertising and Trial: Belief Strength, Belief
Confidence and Product Curiosity," Journal of Advertising, 17 (3), 3-14.

Steuer, Jonathan (1992), "Defining Virtual Reality: Dimensions Determining
Telepresence," Journal of Communication, 42 (4), 73-93.

Strack, Fritz (1992), "The Difierent Routes to Social Judgments: Experiential versus
Informational Strategies," in The Construction of Social Judgments, Leonard L.
Martin and Abraham Tesser, eds., Hillsdale, NJ: Lawrence Erlbaum Associates,
249-276.

107

Sundar, S. Shyam and J inhee Kim (2005), "Interactivity and Persuasion: Influencing
Attitudes with Information and Involvement," Journal of Interactive Advertising,
5 (2), [http://www.iiadcrg].

Taylor, Shelley E. and Sherry Schneider, K. (1989), "Coping and the Simulation of
Events," Social Cognition, 7 (2), 174-194.

Tremayne, Mark and Sharon Dunwoody (2001), "Interactivity, Information Processing,
and Learning on the World Wide Web," Science Communication, 23 (2), 111-134.

White, J. Dennis, Armen Tashchian, and Roobina Ohanian (1991), "An Exploration into
the Scaling of Consumer Confidence: Dimensions, Antecedents, and
Consequences," Journal of Social Behavior and Personality, 6 (3 ), 509-528.

Wicker, Allan W. (1969), "Attitudes versus Actions: The Relationship of Verbal and Overt
Behavior Response to Attitude Objects," Journal of Social Issues, 25 (4), 41-78.

Wright, Alice A. and John G Lynch, Jr. (1995), "Communication Effects of Advertising
versus Direct Experience When Both Search and Experience Attributes Are
Present," Journal of Consumer Research, 21 (March), 708-718.

Wright, Peter (1973), "The Cognitive Processes Mediating Acceptance of Advertising,"
Journal of Marketing Research, 10 (February), 53-62.

---- (1980), "Message-Evoked Thoughts: Persuasion Research Using Thought
Verbalizations," Journal of Consumer Research, 7 (September), 151-175.

Wu, Chenghuan and David R. Shaffer (1987), "Susceptibility to Persuasive Appeals as a
Function of Source Credibility and Prior Experience with the Attitude Object,"
Journal of Personality and Social Psychology, 52 (4), 677-688.

Yi, Huiuk, Joseph E. Phelps, and David Roskos-Ewoldsen (1998), "Examining the

Effectiveness of Comparative Advertising: The Role of Attitude Accessibility,"
Journal of Current Issues and Research in Advertising, 20 (1), 61-74.

108

Zajonc, Robert B. (1968), "Attitudinal Effects of Mere Exposure," Journal of Personality
and Social Psychology, 9(2), 1-27.

Zanna, Mark P. and Russell H. Fazio (1982), "The Attitude Behavior Relation: Moving
toward a Third Generation of Research," in Consistency in Social Behavior: The

Ontario Symposium, Mark P. Zanna, E. Tory Higgins, and C. Peter Herman, eds.
Vol. 2, Hillsdale, NJ: Lawrence Erlbaum Associates, 283-301.

109

APPENDIX 1

QUESTIONNAIRE FOR MAIN EXPERIMENT

110

SHORT SURVEY

The purpose of this questionnaire is to record your background and experience with digital
cameras and laptop computers in general and Kodak digital camera and IBM laptop computer

 

 

 

 

 

 

in particular.
1. Demographic information
i
Gender female male How old are you? years old i
2. Computer use.
On average how many hours a week do you spend in using the Internet? hours

 

3. Please indicate your experience with a digital camera and a laptop computer in general by

circling the most appropriate number.

 

 

How often do you use a digital camera?
Never12 3 4 5 6 7 8 9 Everyday

Do you currently own a digital camera?
Yes No

Overall, how much time have you spent searching
for digital camera?

Verylittle123456789Verymuch

 

How often do you use a laptop computer?
Never 123456789Everyday

Do you currently own a laptop computer?
Yes No

Overall, how much time have you spent searching
for laptop computer?

Verylittle12 3 4 5 6 7 89Verymuch

 

4. Please indicate your familiarity with Kodak digital camera and IBM laptop computer by

circling the most appropriate number.

 

 

Kodakdigitalmmera
Unfarniliar1234567Familiar
Inexperienced1234567Experienced

Not
knowledgeable 1 2 3 4 5 6 7 knowledgeable

 

IBM WWW

Unfamiliar1234567Familiar
Inexperienced1234567Expa'ienced

Not
knowledgeable 1 2 3 4 5 6 7 knowledgeable

 

111

 

 

 

Instructions
The aim of this survey is to determine the vividness of your mental imagery. The items of the
survey will bring certain images to your mind. You are to rate the vividness of each image by
reference to the accompanying rating scale, which is shown below. Before you turn to the items
on the next page, familiarize yourself with the different rating scale. A copy of the rating scale is
printed on the next page for your reference. Try to do each item separately independent of how
you may have done other items.

 

Rating Scale

The image aroused by an item of this survey may be:

No image present at all, you only ‘knowing’ that you are thinking Rating 1

 

of the object

So vague and dim as to be hardly discernible Rating 2
Vague and dim Rating 3
Not clear or vivid, but recognizable Rating 4
Moderately clear and vivid Rating 5

Very clear and comparable in vividness to the actual experience Rating 6

Perfectly clear and as vivid as the actual experience Rating 7

Example

An example of an item on the survey would be one that asks you to consider an image that comes
to your mind’s eye of a red apple. If your visual image is moderately clear and vivid you would
check the rating scale and mark ‘5’ in the brackets as follows:

 

Item Rating

 

A red apple ( 5 )

 

 

 

Now turn to the next page when you have understood these instructions.

112

Think of some relative or friend whom you frequently see, considering carefully the picture that
rises before your mind’s eye. Classify the images suggested by each of the following questions as

indicated by the degree of cleamess and vividness specified on the rating scale.

 

 

 

 

 

Item Rating
1. The exact contour of face, head, shoulders and body ( )
2. Characteristic poses of head, attitudes of body, etc. ( )
3. The precise carriage, length of step, etc. in walking. ( )
4. The different colors worn in some familiar costume ( )

 

 

 

 

Think of seeing the following, considering carefully the picture that comes before your mind’s
eye; and classify the image of cleamess and vividness specified on the Rating Scale.

 

5. The sun as it is sinking below the horizon. I ( ) I

 

Rating Scale

The image aroused by an item of this survey may be:

No image present at all, you only ‘knowing’ that you Rating 1
are thinking of the object

So vague and dim as to be hardly discernible Rating 2
Vague and dim Rating 3
Not clear or vivid, but recognizable Rating 4
Moderately clear and vivid Rating 5

Very clear and comparable in vividness to the actual experience Rating 6

Perfectly clear and as vivid as the actual experience Rating 7

113

MAIN SURVEY

The purpose of this questionnaire is to record your evaluation of the product web site you just
saw, your experience with the site, and the evaluation of the product in the site. Completion
of the survey will take approximately 10 minutes. Your responses will remain anonymous. Thank
you for your assistance!

 

1. Utilitarian products refer to goods whose goal of consumption is mainly practical, functional,
something that helps achieve a goal (e.g., pain killer), while hedonic products refer to goods
whose goal of consumption is mainly for fun and pleasure, something that is enjoyable and
appeals to the sense (e.g., perfume). Please try to categorize the product you just saw as
utilitarian, hedonic, or both by placing a check (V).

Utilitarian Hedonic Both

 

2. Please indicate your level of agreement with the following statements by circling the most
appropriate number.

Strongly Strongly
Disagree Agree
The web site I just saw....
Included various interactive features. 1 2 3 4 5 6 7
Included vivid product description. 1 2 3 4 5 6 7

Included a lot of information about the product. 1 2 3 4 5 6 7

Was well organized. 1 2 3 4 5 6 7

3. How would you describe the images that occurred in your mind while examining the web
site materials? Indicate your level of agreement or disagreement with the following
statements by circling the most appropriate number.

Strongly Strongly
DEM Agrc_e
The images were....
Clear 1 2 3 4 5 6 7
Vivid 1 2 3 4 5 6 7
Intense l 2 3 4 5 6 7
Lifelike l 2 3 4 5 6 7
Sharp 1 2 3 4 5 6 7

114

4. The following items concern your evaluation about product-related information obtained
during exposure to the web site. Product-related information includes any ideas, thoughts,
and images about the product you have had while examining the web site, in addition to
the product information directly obtained from the web site. Please indicate your
evaluation by circling the most appropriate number.

In judging the quality and performance of the product presented, overall, how helpful
would you rate the product-related information you obtained during exposure to the web
site?

Not helpful at all 1 2 3 4 5 6 7 Extremely helpful

In understanding different features of the product presented, overall, how helpful would
you rate the product-related information you obtained during exposure to the web site?

Not helpful at all 1 2 3 4 5 6 7 Extremely helpful

To what extent did the product-related information you obtained during exposure to the
web site enable you to accurately evaluate the product presented?

It did not It fully enabled
enable me to me to evaluate
evaluate the 1 2 3 4 5 6 the product
product at all

5. For each of the items below, circle the number that best describes how you felt while you
were viewing the web site.

Happy 1 2 3 4 5 6 7 Unhappy
Pleased l 2 3 4 5 6 7 Annoyed
Excited 1 2 3 4 5 6 7 Calm

Stimulated 1 2 3 4 5 6 7 Relaxed

Aroused 1 2 3 4 5 6 7 Unaroused

115

6. For each of the items below, circle the number that best describes your overall feelings about
the product you saw in the web site.

Bad 1 2 3 4 5 6 7 Good

Like

Drslrke 1 2 3 4 5 6 7 quitealot

quite a lot

Unpleasant 1 2 3 4 5 6 7 Pleasant

Bad quality 1 2 3 4 5 6 7 Good quality

7. How certain are you of the accuracy of the responses you gave in the previous questions
regarding your feelings toward the product?

Not confident I 2 3 4 5 6 7 Confident

Not sure 1 2 3 4 5 6 7 Sure

Uncertain 1 2 3 4 5 6 7 Certain

8. How likely would you purchase this product? Circle the number that best indicates the
likelihood, assuming that the price was within your acceptable range.

Unlikely 1 2 3 4 5 6 7 Likely

Improbably l 2 3 4 S 6 7 Probably

Impossible 1 2 3 4 5 6 7 Possible

116

THOUGHT LISTING

Please read the instructions below carefully.
Instructions

As you viewed the web site, what were the thoughts that went through your mind? Please
write down in the lines below, everything that you thought of, regardless of whether it
pertained to the product itself or anything else that went through your mind. Please write one
item per line; you do not need to fill all lines, but please do your best to remember as many as
possible. Do not worry about spelling, grammar, or punctuation.

1.

 

 

 

 

 

 

 

 

 

10.

 

ll.

 

12.

 

13.

 

l4.

 

117

APPENDIX 2

INSTRUCTION FOR EXPERIMENT

 

118

1St screen for 2-D condition

Welcome !

The purpose of this study is to collect your evaluation of two products
presented on web sites.

You will be asked to view two products’ web sites one at a time. Upon
completion of each web site viewing, you will be asked to
report your opinions and thoughts.

So you should thoroughly examine the web site to determine
how you think and feel about the product presented.

Prior to the actual experiment, you will be given a training session to
familiarize yourself with web site viewing.

When you are ready, click CONTINUE button to proceed to the training
session.

 

CONTINUE

 

 

 

119

1St screen for 3-D condition

Welcome !

The purpose of this study is to collect your evaluation of two products
presented on web sites.

You will be asked to view two products’ web sites using 30 features
one at a time. Upon completion of each web site viewing, you
will be asked to report your opinions and thoughts.

So you should thoroughly examine the web site to determine
how you think and feel about the product presented.

Prior to the actual experiment, you will be given a training session to
familiarize yourself with 30 interface.

When you are ready, click CONTINUE button to proceed to the training
session.

 

CONTINUE

 

 

 

120

APPENDIX 3

INSTRUCTION FOR REACTION TIME MEASURE

121

1St screen

PLEASE DO NOT CLICK CONTINUE UNTIL TOLD DO SO
Instructions for Reaction Time

Now we are about to test a new attitude measurement technique
using a computer program. A series of brand names from various
product categories (e.g., clothing, laptop computer, wrist watch, digital
camera) will appear on the computer screen and you need to indicate
your overall feelings about the brand by hitting one of the two keys of
your computer keyboard - “z" for “like" and “l" for “dislike."

There are two things that we want you to keep in mind as you
do this task. First, and above all, be accurate. Don’t be in such a hurry
to respond that you regret your decision. Second, while being accurate,
try to respond as quickly as possible.

Presentations of brand names will be separated by a two-second
interval. The computer will record your response and reaction time to
each trial. Prior to actual test, you will be given a training session.

Do not click CONTINUE until told do so. Please raise your hand when
you have finished reading these instructions.

 

 

ID: CONTINUE

 

 

 

 

 

 

122

2nd screen

Training Session

This is a training session for reaction time measure. When you click
CONTINUE, ten brand names from various product categories will
appear on your computer screen, and you need to indicate your overall
feelings about the product by hitting “z" of your keyboard for “like"
or “I" of your keyboard for “dislike.” To avoid mistakenly hitting a
wrong key, keep one index finger above each of the two keys during
the test. Please respond as quickly and accurately as possible.

Click CONTINUE when you are ready.

 

CONTINUE

 

 

 

3rd to 22nd screens

The test is about to begin.
Place your index fingers on "2" and "l" keys of your keyboard now.

(The above instructions were displayed for two seconds between presentations often
brand names)

123

23th screen

This is the end of a training session. If you don’t feel comfortable with
the procedure, click GO BACK TO TRAINING. It will bring you back
to the training session. Otherwise, click CONTINUE to take an actual
experiment.

 

 

GO BACK TO TRAINING CONTINUE

 

 

 

 

 

 

24th screen

Now we are about to actually record your evaluations about products
and reaction time. This time, twenty brand names, different from the
training session, will appear on your computer screen. The procedures
used in this session are exactly alike to those in the training session.
You need to indicate your overall feelings about the product by hitting
either “2” for like or “I" for dislike. To avoid mistakenly hitting a
wrong key, keep one index finger above each of the two keys during
the test. Please respond as quickly and accurately as possible.

Click CONTINUE when you are ready.

 

CONTINUE

 

 

 

124

25th to 64th screens

The test is about to begin.
Place your index fingers on "2" and "I" keys of your keyboard now.

(The above instructions were displayed for two seconds between presentations of twenty
brand names)

Last page of reaction time measure screen

This is the end of the experiment. Thank you very much for your
cooperation. In case you have any questions about this experiment,
feel free to let the researchers know.

125

APPENDD( 4

DESCRIPTIVE STATISTICS FOR MEASUREMENT ITEMS

126

Table 21. Descriptive Statistics for Measurement Items (IBM laptop computer)

 

 

 

 

 

 

 

 

 

 

 

 

 

N Minimu Maximu Mean Std. Skewness Kurtosis
Std. Std.
__ Statistic Statistic Statistic Statistic Statistic Statistic Error Statistic Error
LMIl 226 2.00 7.00 4.9779 1.17168 -.224 .162 -.335 .322
LMI2 226 1.00 7.00 4.6858 1.25201 -.196 .162 -.328 .322
LMI3 226 1.00 7.00 3.8274 1.45185 .067 .162 -.345 .322
LMI4 226 1.00 7 .00 4.5841 1.38709 -.392 . 162 -.008 .322
LMIS 226 1.00 7 .00 4.6549 1.38416 -.427 .162 .007 .322
LER] 226 3.00 7.00 4.2566 .7863! .785 .162 .956 .322
LER2 226 3.00 7.00 4.3673 .88070 .745 .162 .608 .322
LER3 226 1.00 6.00 2.9912 1.15467 .524 .162 -.189 .322
LER4 226 1.00 7 .00 3.4912 1.17091 .306 .162 -.593 .322
LDIAGl 226 1.00 7 .00 4.9690 1.09297 -.618 . 162 .239 .322
LDIAGZ 226 2.00 7.00 5.0708 1.10929 -.555 .162 .123 .322
LDIAG3 226 2.00 7.00 4.7124 1.05894 -.400 .162 -.258 .322
LACl 226 3.00 7 .00 5.7080 .97690 -.51 1 .162 -.102 .322
LAC2 226 3.00 7.00 5.7080 .97234 -.615 .162 .260 .322
LAC3 226 3.00 7.00 5.5796 1.01777 -.461 .162 -.264 .322
A11 226 1.00 7.00 5.2124 1.18473 -.532 .162 .212 .322
A12 226 1.00 7.00 4.9690 1.25217 -.160 .162 -.266 .322
A13 226 1.00 8.00 4.3053 1.30628 .251 .162 .232 .322
A14 226 1.00 7 .00 4.8407 1.47312 -.319 .162 -.483 .322
A15 226 1.00 7.00 5.5398 1.18536 -.483 .162 -.l69 .322
LPCEl 226 1.00 9.00 5.6460 3.28477 -.147 .162 -1.714 .322
LPCE3 226 1.00 9.00 3.7080 2.56535 .604 .162 -.908 .322
LBFI 226 1.00 7.00 3.3186 1.92188 .456 .162 -.896 .322
LBF2 226 1.00 7.00 3.0265 1.79114 .662 .162 -.431 .322
LBF3 226 1.00 7.00 3.1062 1.82264 .566 .162 -.677 .322
LTL 226 .00 14.00 4.6947 2.42573 .474 .162 1.060 .322
LRT 226 .371 1.892 .8788 .276363 1.133 .162 1.634 .322

 

Note: LMI (Mental Imagery for IBM laptop); LER (E
(Information Diagnosticity for IBM laptop); LAC (Atti
Imagine - common items for both products); L
(Brand Familiarity for IBM laptop); LTL (Thou

IBM laptop)

127

motional Response for IBM laptop); LDIAG
tude Confidence for IBM laptop); AI (Ability to
PCB (Product Category Experience for IBM laptop); LBF

ght Listing for IBM lapt0p); LRT (Reaction Time for

 

 

 

Table 22. Descriptive Statistics for Measurement Items (Kodak digital camera)

 

 

 

 

 

 

 

 

 

 

 

 

N Minimu Maximu Mean Std. Skewness Kurtosis
Std. Std.

‘ _ Statistic Statistic Statistic Statistic Statistic Statistic Error Statistic Error
CMIl 227 1.00 7 .00 5.3524 1.13240 -.708 .162 .484 .322
CM12 227 1.00 7.00 5.1 189 1.26524 -.795 .162 .644 .322
CM13 227 1.00 7.00 4.1718 1.39262 -.162 .162 -.500 .322
CMI4 227 1 .00 7.00 5 .0661 1.42979 -.630 .162 -.161 .322
CM15 227 1.00 7.00 5.1278 1 .34246 -.634 .162 .075 .322
CERl 227 3.00 7.00 5.110] .99611 .211 .162 -1.166 .322
CER2 227 4.00 7.00 5.1938 1.05903 .214 .162 -1.291 .322
CER3 227 1.00 7.00 4.0044 1 .26735 -.048 .162 -.604 .322
CER4 227 1.00 7 .00 4.3304 1.34068 -.055 .162 -.478 .322
CDIAGI 227 2.00 7.00 5.2379 .99368 -.629 .162 .274 .322
CDIAG2 227 2.00 7.00 5.1806 1.10828 -.599 .162 .075 .322
CDIAG3 227 2.00 7.00 5.0969 .92619 -.329 .162 .169 .322
CACl 227 3.00 7.00 5.881 1 .89687 -.469 .162 .004 .322
CAC2 227 2.00 7.00 5.8238 .97070 -.635 .162 .483 .322
CAC3 227 1.00 7.00 5.7665 1.00579 -.781 .162 1.432 .322
All 227 1.00 7.00 5.2070 1.17735 -.541 .162 .246 .322
A12 227 1.00 7.00 4.9604 1.25268 -.143 .162 -.284 .322
A13 227 1.00 8.00 4.3172 1.30547 .237 .162 .225 .322
A14 227 1.00 7.00 4.8502 1.46781 -.348 .162 -.465 .322
A15 227 1.00 7.00 5.5419 1.18686 -.477 . 162 -.183 .322
CPCEl 227 .00 9.00 4.5419 1.95339 .071 .162 -.716 .322
CPCE3 227 1.00 9.00 3.8018 2.03952 .458 .162 -.459 .322
CBF] 227 1.00 7.00 3.6123 2.00650 .099 .162 -1.220 .322
CBF2 227 1.00 7.00 3.0132 1.76365 .497 .162 -.719 .322
CBF3 227 1.00 7 .00 3.1013 1.79067 .462 .162 -.819 .322
CTL 227 .00 14.00 4.6564 2.78682 .839 .162 1.146 .322
CRT 227 .387 1.833 .80904 .224566 1.260 .162 2.516 .322

 

 

Note: CMI (Mental Imagery for Kodak digital camera); CER (Emotional Response for Kodak

digital camera); CDIAG (Information Diagnosticity for Kodak digital camera); CAC (Attrtude
Confidence for Kodak digital camera); AI (Ability to Imagine — common items for both products); CPCE
(Product Category Experience for Kodak digital camera); CBF (Brand Famrlrarrty for Kodak drgrtal
camera); CTL (Thought Listing for Kodak digital camera); CRT (Reactron Tune for Kodak digital

camera)

128

 

 
 

 

 

l“glljlij‘lgl;1lwill