rt. .V’fiq... . ‘ 6 .
hhfluéfadfifisfliniéi
Liirfiafiflg mt
$thfiL
$.21,
2;?
Hart

:9.
{3.124

17:1".
at"... «“31. n: {I
; . 2 $33.? .
. 1...?“ ,
V $3

$23. "2.5.;me om

all
5

.zrhquwmvflw

ihh.‘..l$
v5.1.3.
amwg . I! .{1
n. K“: 1.5!
In,“ ., I:
n x . ...I¢uau..o.1(
.9 1: Ltt :5...
a a ($.33?
ginflv . .
pf . diving..." L.
.. . .i i Mn.
an. :3 I}: r5935...

.. away...

1 v 9

I 59 3...?

, t: « . . ‘
5:155:3132 J I: 1...: I .3
31:: i. Wilt:

$3152.41!
2. 51-11 . av t ..
£11 £5. .. nanny-1...... lam
. i n l .4"-
d
. kg?
9

.3
t.” 3‘”?

.1. at
(13551115
rll...r.

 

Am?” LIBRARY
2 Michigan State
2 " r5 I University

This is to certify that the
dissertation entitled

APPLICATION OF THE CONVERSION AND TRACKING
MODELS IN MEASURING THE EFFECTIVENESS OF
TRAVEL MICHIGAN’S 2003 TRAVEL ADVERTISING

CAMPAIGN

presented by

KUDZAYI MAUMBE

has been accepted towards fulfillment
of the requirements for the

DOCTORAL degree in PARKS, RECREATION AND
TOURISM RESOURCES

 

 

 

‘Major Professor's Signature
97/174 9/ $50 63

Date

MSU is an Affinnative Action/Equal Opportunity Institution

_ - -.-.—.------.-.- -t- _ _ . a

_ _.-,-.-.-.-._- -

‘

o‘ffi'

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/C|RC/DaleDue.indd-p.1

APPLICATION OF THE CONVERSION AND TRACKING MODELS IN
MEASURING THE EFFECTIVENESS OF TRAVEL MICHIGAN’S 2003 TRAVEL
ADVERTISING CAMPAIGN

By

Kudzayi Maumbe

A DISSERTATION
Submitted to
Michigan State University

In partial fulfillment of the requirements
Of the degree of

DOCTOR OF PHILOSOPHY

Parks, Recreation and Tourism Resources

2006

ABSTRACT
APPLICATION OF THE CONVERSION AND TRACKING MODELS IN
MEASURING THE EFFECTIVENESS OF TRAVEL MICHIGAN ’S 2003 TRAVEL
ADVERTISING CAMPAIGN
By
Kudzayi Maumbe
This study’s objective was to determine the effectiveness of the Travel

Michigan’s 2003 travel advertising campaign by application of the conversion and
tracking models. Seven stages of data analysis were conducted in this study. The first
stage was to determine whether significant differences existed in pre and post campaign
advertising awareness. In the second stage, T-tests were conducted to identify significant
differences in Destination Awareness (DA), Intention to Visit (IV) and Actual Visitation
(AV) between those aware and those not aware of advertising. The third stage involved
principal components analysis to reduce the captured 22 attitude variables into a more
manageable data set that was used in subsequent analyses. In the fourth stage, correlation
analysis was conducted to determine the correlation of the image components with
Advertising Awareness (AA), Destination Awareness (DA) and Intention to Visit (IV). In
the fifth stage, two models (Unaided Advertising Awareness (UAA) and Destination
Awareness (DA)) were evaluated using logistic regression. In the sixth stage, the
conversion model was applied to the data to calculate conversion ratios using cross-
tabulations. Four different approaches named routes (0-3) were utilized to calculate the
four alternative conversion measures. The routes were derived from the conversion and

tracking models given by Siegel and Ziff-Levine (1990). Route zero was the direct route

 

 

 

onl}
infll
as d
mor
trac
con
Sev

ICIL

adv

we]
aw;
CIIC
imr
EX}
W e

SCI’

rou
yiel

Ovc

only considering those visitors to the state who indicated their decision to visit was
influenced by the advertisements without necessarily having gone through all the stages
as depicted by the conversion and tracking models. Route one was from the tracking
model, and it included the destination awareness stage, route two was also from the
tracking model, but it excluded the destination awareness stage. Route three was from the
conversion model which included all the stages plus the “inquiry-fulfillment” stage.
Seventh, the conversion ratios from the preceding stage were utilized to calculate the
return on advertising investment.

The results at each stage consistently show the effectiveness of Travel Michigan’s
advertising. Significant differences in advertising awareness were found between phases
1 and 2, implying the advertisements managed to gain the audience’s attention. There
were also significant differences in DA, IV and AV between those aware and those not
aware. Results showed Michigan is viewed mainly as a family destination and as a close
enough destination for short trips. Logistic regression results showed TV is the most
important in increasing one’s advertising awareness, followed by gender and experience.
Experience is the most important in increasing one’s destination awareness followed by
Web-use, intention to visit and exposure to advertising. Results from the sixth and
seventh stages of analysis show positive conversion ratios and ROI. Route zero yielded a
11.2% conversion ratio and a $10.47 ROI for every dollar spent on advertising; route one
yielded a 5.2% conversion ratio and a $4.86 ROI for every dollar spent on advertising;
route two resulted in a 8.5% conversion ratio and $7.94 ROI and finally route three
yielded a 2.2% conversion ratio with $2.06 ROI for every dollar spent on advertising.

Overall, Travel Michigan achieved a positive return on advertising investment.

DEDICATION

To my great parents Eliah and Emmah Chitiyo.

iv

 

thi:
m)
Do
this

my

all

IOL

up.

pro
(35;).
faci

for 1

line
my.
me;

Chili

ACKNOWLEDGEMENTS

There are people to whom I need to express my sincere gratitude, without whom
this degree and dissertation would not have been possible. My deepest gratitude goes to
my dissertation advisor and also the chair of my graduate academic committee Professor
Donald Holecek for his guidance and concise support throughout my study and writing of
this dissertation. I would also like to express my sincere gratitude to the other members of
my graduate committee Professors Christine Vogt and Bonnie Knutson for their guidance
and input towards the success of this study. I also feel very indebted to Mr. Mavuso and
all the library staff of the University of Fort Hare Library, South Africa for allowing me
to use their library and computer resources during the last stages of my dissertation write
up. I thank you all very much.

I am eternally grateful to the PEO and International Peace Scholarship (IPS) for
providing part of the funding for the first two years of my doctoral program. I am
especially grateful to Carol Goeman for being my required U. S. contact person and
facilitator for the IPS without which this degree would have just remained a pipe dream
for me.

I would also like to thank my friends who were there cheering me up to the finish
line. My greatest thanks go to my dearest friend Sue Tims and her husband Ted, literally
my other parents in the U. S. for their unconditional love and support and being there for
me all the time as I struggled to balance the demands of being both a parent of young

children and a doctoral student. I owe you. Thanks also go to all the members of the

University United Methodist Church community for their moral, emotional and spiritual
support during my stay in East Lansing.

I would also like to offer my deepest gratitude to my family. To my parents, I
thank you for building a strong academic foundation for me and instilling the hard work
ethic in me. Special thanks goes to my elder brother Gwinyai and elder sister Moud for
their contribution in laying the foundation to my education and to my eldest brother Onai,
younger sister Tendayi and brothers in-law Munya and Antony for their unconditional
love and all round support during my stay and academic struggle in the U. S.

Last, but not least I would like to thank my dear husband Blessing for his
encouragement and great support; putting his career on hold to help me realize my
dreams. Deepest gratitude goes to my lovely children Samantha, Mike and Victoria, for

the patience and cheer. Now you can play with the computer and bang the keyboard.

vi

TABLE OF CONTENTS

LIST OF TABLES ......................................................................................................... viii
LIST OF FIGURES ......................................................................................................... ix
LIST OF APPENDICESKEY TO ABBREVIATIONS ................................................ x
KEY TO ABBREVIATIONS ......................................................................................... xi
Chapter 1 ............................................................................................................................. 1
1.0 Introduction and Problem Statement ...................................................................... 1
1 .1 Objectives ........................................................................................................... 4
1.2 Hypotheses .......................................................................................................... 5
Chapter 2 ............................................................................................................................. 8
2.0 Literature Review .................................................................................................... 8
2.1 Theoretical Background ...................................................................................... 8
2.1.1 Persuasion Theories .................................................................................... 8
2.1.2 Selective Processes Theories (Cognitive Dissonance Theory) ................. 15
2.1.3 Communication theories ........................................................................... 15

2.2 Does Advertising Really Work? ....................................................................... 18
2.2 Travel Advertising Effectiveness Methods Review ......................................... 23
Chapter 3 ........................................................................................................................... 28
3.0 Methods ................................................................................................................. 28
3.1 Study Design ..................................................................................................... 28
3.2 Data collection .................................................................................................. 28
3.3 Data analysis techniques used ........................................................................... 33
Chapter 4 ........................................................................................................................... 36
4.0 Results and Discussion ......................................................................................... 36
4. 1 Descriptives ....................................................................................................... 36
4.2 Pre and Post Advertisement Comparisons ........................................................ 38
4.2.1 Advertising Awareness ..................................................................................... 38

4.3 Hypothesis Testing .................................................................................................. 39
4.3.1 Model Estimation ......................................................................................... 55
4.3.1.1 Advertising Awareness .............................................................................. 55

4.3.1.2 Destination Awareness .............................................................................. 60

4.4 Conversion Rates .............................................................................................. 63
4.4.1 Advertising Conversion Model ................................................................. 63
4.4.2 The Advertising Tracking Model ..................................................................... 73

4.5 Return on Advertising Investment (ROI) ......................................................... 76
Chapter 5 ........................................................................................................................... 80
5.0 Summary and Conclusion ........................................................................................... 80
5.1 Study Limitations .............................................................................................. 90
References ......................................................................................................................... 92

vii

 

 

 

 

Tat

Tat

T31

T31

Ta

T2

LIST OF TABLES

Table 1: Advertising, Related Behavioral Dimensions and Advertising Research

Related to the model ..................................................................... 22

Table 2: Variables measured and their levels of measurement ................................ 31

Table 3: Gender * Number of respondents by their residence DMA.. . . . . . . . . . . . . . ........38
Table 4: T-tests for Equality of means for DA, IV and AV between those aware

and those not aware of adverts ............................................................ 41

Table 5: T-tests for attitude towards Michigan as a tourism destination. . . . . . . . .. ...42

Table 6: Principle components analysis results showing loading of each variable on the
components ................................................................................ 45

Table 7: Principle components results showing eigenvalues for the components and %
of variation explained by each component ........................................... 47

Table 8: Correlation results between attitude factors and AA, DA and IV .................. 51
Table 9: Correlation results for Annual Income, DMA location and AA, DA, and IV ..... 53
Table 10: Variables in the Unaided Advertising Awareness (UAA) Model. . . . . . ........56

Table 11: Goodness of fit tests for the Unaided Advertising Awareness (UAA)
Model ..................................................................................... 57

Table 12: Variables included in the Destination Awareness (DA) Mode] at each step ..... 61

Table 13: Goodness of fit tests for the Destination Awareness (DA) Model ............... 61
Table 14: Cross-tab results for AA and DA with the three image factors ................... 66
Table 15: Conversion rates from Image stage to Inquiry Fulfillment stage ................. 69
Table 16: Conversion rates from the Inquiry to the Motivation stage ........................ 70
Table 17: Conversion rates for the “Conversion Behavior” stage ............................ 72
Table 18: R01 figures for the 2003 Travel Michigan advertising campaign ............... 78

viii

LIST OF FIGURES

Figure 1: H-J-K instrumental theory of communication and persuasion model. . . . . . ..9
Figure 2: An integrative framework of advertising persuasion ................................ 14
Figure 3: (A) Conversion and (B) Tracking Models ........................................... 35
Figure 4: Distribution of respondents among the three DMAs ................................. 37
Figure 5: The conversion model for Travel Michigan’s 2003 travel advertising. . . . . .......65

Figure 6: Advertising tracking model showing the different routes (0-3) used in
calculating the conversion rates and ROI ........................................................ 74

Figure 7: Advertising tracking model for Travel Michigan ................................... 75

ix

 

 

 

AP:

AP:

AP:

LIST OF APPENDICES

Appendix 1: Sample of Questionnaire used in the study ...................................... 97
Appendix 2: Description of the advertisements by channel and DMA ................... 105
Appendix 3: Detailed Descriptive Statistics by DMA ....................................... 107
Appendix 4: Test for appropriateness of data set for factor analysis ..................... 109

 

ax
CR
DA
D.\
1v:
Mi

0C
PD:
PEI
PFI
R01
TTI

UA.

AV:

CR:

DA:

DMA:

MI:

OCR:

PDI:

PEI:

PFI:

ROI:

TTRRC :

UAA:

KEY TO ABBREVIATIONS

Advertising Awareness
Actual Visitation

Conversion Rate

Destination Awareness
Designated Marketing Area
Intention to Visit

Michigan

Overall Conversion Rate
Positive Distance Image
Positive Entertainment Image
Positive Family Image
Return On Investment
Travel, Tourism and Recreation Resource Center

Unaided Advertising Awareness

xi

 

 

Th

do

1‘31

CC

Chapter 1

1.0 Introduction and Problem Statement

The economic importance of tourism to states, regions and the world has been well
documented and can’t be disputed. Within the United States alone, travel and tourism is a
US$13 trillion industry, is the leading employer in states such as Florida and Hawaii and
ranks among the top three employers in 29 of the 50 states (TIA 2005). Recognizing this
economic importance of tourism to the economy, states have agencies that promote
inbound tourism with budgets ranging from about $1.7m in Delaware to over $70m in
Hawaii (TIA, 2003). With such big tourism budgets coupled with tight economic
conditions, states’ tourism agencies are faced with increasing pressure to justify their

promotion budgets and be accountable for the planned impacts.

Conversion studies have been commonly used in most of the advertising effectiveness
research. These studies, however, have been criticized for having methodological and
design shortcomings that limit their usefulness and application (Messmer and Johnson,
1993). Some of these limitations include: overstating of the conversion ratio as a result of
including those individuals who had already decided to visit before being exposed to
advertisements, high non-response bias and inability to provide quantitative figures on
return to advertising investment. Advertising tracking studies have also been frequently
used to determine advertising effectiveness. Advertising tracking studies seek to generate
awareness among the target audience, create awareness of the destination as a place to

visit, create a positive image of the destination, motivate consumers to travel to the

 

 

dc

tr:

3‘

in

it

It

(I)

destination and influence travel behavior (Siegel and Ziff-Levine, 1990). The advertising
tracking model, like the conversion model, because it is most suitable for generating
awareness and creating positive destination image, can’t be expected to produce an all

inclusive estimate of return on investment.

Many communication theories have been used as the basis for advertising research
including the Hierarchy of Effects model (Palda, 1966) and the AIDA model: Attention,
Interest, Desire and Action (Haley, 1985). The Hierarchy of Effects model constitutes six
steps, namely Attention-Interest-Comprehension-Impact-Attitude and Sale. Howard and
Sheth (1969) developed a theory of buyer behavior that supported the Hierarchy of
Effects model of consumer behavior. In this model, Howard and Sheth (1969) employed
the following five output variables: attention, brand comprehension, attitude, intention
and purchase. Regardless of which model of communication and/or consumer behavior
being referred to, the “attention” and ‘interest/attitude’ stages both always precede the
action stage. This implies that getting the consumer’s attention, generating interest and
changing their otherwise negative, neutral or non-existent attitudes are prerequisites to
getting to the action stage. Advertising effectiveness, therefore, can be measured in more
ways than just counting the individuals that act. Research has shown that 76% of
“definites,” 31% of “probables” and 27% of “mights” actually purchase the advertised

brand (Gruber, 1970).

The statement of the problem was to determine the effectiveness of travel advertising on

 

 

us:
jus
dil

fo'

peoples’ destination awareness, attitudes towards MI as a tourism destination, motivation
(intention) to visit the state and actual visitation. The goal is to determine how effective
the advertisements are in moving individuals from a state of ‘not aware of MI as a tourist
destination’ to ‘mights,’ ‘probables’ and ‘definites.’ The study also seeks to investigate
and analyze the relationship between a number of different variables including
advertising awareness, destination awareness, attitude, intention to visit and other
demographic variables such as income and education, to mention a few. While this study
uses the same kind of procedure as in the conversion and tracking studies, it goes beyond
just determining the effectiveness of advertising. It also investigates the relationship of
different important variables as they pertain to advertising and destination awareness. The
following two main dependent variables will be evaluated in this study:

0 Advertising Awareness (AA)

0 Destination Awareness (DA
Several predictor variables will be used including;

0 Exposure to advertisements

0 Experience with leisure travel in the state

0 Advertising channel

0 Income

0 Gender

0 Web use, searching for travel information

o Toll-free phone inquiry

0 Distance of DMA from Michigan

- Attitude towards or Image/Perception of Michigan as a tourist destination

 

 

1.1

Th

1.1

Objectives

The objectives of this study are to:

1.

2.

Determine consumer awareness of Travel Michigan’s advertising.

Establish whether or not Travel Michigan’s advertising program is effective in
generating destination awareness.

Determine advertising effectiveness in changing consumers’ attitudes of Michigan
as a tourism destination.

Analyze the relationship between annual income, location of DMA and
advertising awareness (AA), destination awareness (DA) and intention to visit
(IV) the state of Michigan.

Determine advertising effectiveness of the Travel Michigan’s 2003 advertising
campaign by computing conversion ratios for the three targeted DMAs including
Chicago, Cleveland and Indianapolis.

Calculate the return to advertising investment (ROI) for Travel Michigan’ 5 2003

travel advertising campaign.

1.2 Hypotheses

The following hypotheses were addressed in this study:

1. Individuals exposed to Travel Michigan’s advertising will have greater
awareness of the state as a tourism destination than those who have not been
exposed.

Creating awareness is the first in a series of stages through which advertising
works to change people’s behaviors according to the Hierarchy of Effects Model
of communication. Advertising first creates awareness, develops knowledge about
the product, generates a liking and preference for the product leading to one’s
conviction to purchase the product resulting in the actual purchase of the product
(Lavidge and Steiner, 1961). Simply raising top-of-mind awareness of a brand
enables it to be evoked into more choice sets, thus increasing the probability of it
being selected (Nedungadi, 1990).

2. Travel Michigan’s advertising will change people’s attitudes towards the
state as a tourism destination.

Advertising theory represents advertising as indirectly influencing behavior
through its more direct effect on consumer attitudes (Butterfield et. al., 1998)

3. Those individuals aware of Travel Michigan’s advertising will have greater
motivation/intention to visit the state on a leisure trip, than those who are not
aware.

Development of motivation/intention to purchase is considered one of the
intermediate results of advertising. After being exposed to advertising individuals

develop motivation to purchase before they exhibit some conversion behavior

(Siegel and Ziff-Levine, 1990). Intentions are valid predictors of one’s

consumption of a brand (Wansik and Ray, 2000).

. Travel Michigan’s advertising is effective in generating visits to the state.

Advertising is understood to have an effect on people’s behavior or actions
through a hierarchical process which involves getting people’s attention,
generating their interest and desire resulting in some action (AIDA) (Sandage and
Fryburger, 1963). In this case, the action is taking a pleasure trip to the state of

Michigan.

. Attitudes towards Michigan as a tourism destination will be positively

correlated to Advertising Awareness (AA), Destination Awareness (DA) and
Intention to Visit the state (IV).

According to the Hovland, Janis and Kelly (H-J-K) (1953) communication
persuasion model, audience factors such as attitudes, self esteem, intelligence and
others play an important role in effective persuasion communication and affect
the way by which the stimuli (message) is perceived and received (Tan, 1985).
Attitude and belief changes are a prerequisite to and precede sales, and there
exists a positive relationship between changes in recall and attitudes (Haskins,

1964).

. Distance of a DMA from Michigan will be negatively correlated with AA,

DA, and IV.
Travel behavior has changed. People are now taking more short getaways today

than they did ten years ago (TIA, 2003).

7. Annual household income will be positively correlated with AA, DA, and IV.
Resources are not used to develop judgments, beliefs or plans of action unless
some motive exists (Feldman and Lynch, 1988). Because low income people have
no financial resources to take pleasure trips, it is assumed that they have no
motivation to pay attention to travel advertising hence, have no intention to take
pleasure trips.

8. Having taken a pleasure trip to Michigan before (Experience) will have a
positive effect on advertising awareness.

This hypothesis is based on the selective processes or cognitive dissonance theory
which states that individuals selectively attend to messages in an effort to reduce
dissonance. Post purchase doubts about the wisdom of choosing a product could
lead one to attend to that product’s advertising in order to allay such doubts
(Rotzol, 1964). Ehrenberg (2000) concluded that advertising’s role is to reinforce
feelings of satisfaction with brands already bought.

9. Advertising channel has a significant effect on advertising awareness.

The way a message is conveyed to an audience (pictorially or verbally) has been
argued to have a significant effect on the processing of the information contained
in the advertisement. These processing differences have been seen to have an
effect on the audiences’ attitudes towards the brand and purchase intentions.
Pictorial advertising stimuli can yield different results from verbal advertising

stimuli, and pictorial stimuli generally are better recalled (Edel and Staelin, 1983).

Chapter 2

2.0 Literature Review

This chapter covers the literature review on the subject of advertising effectiveness. The
theoretical background is given first, which includes a detailed discussion on persuasion,
selective processes, and connnunication theories. The chapter is concluded by a

discussion on research methods utilized in previous advertising effectiveness studies.

2.1 Theoretical Background

2.1.1 Persuasion Theories

The subject of advertising awareness is closely linked and studied using communication
and persuasion theories. Many theories have been developed over the years to try to
explain the behavior of humans when they receive messages during the communication
process. Since one of the purposes of advertising is to persuade and convince the
audience to buy the product, persuasion theories become important in advertising studies.
Hovland, Janis and Kelly (I-I-J-K) (1953) developed an instrumental theory of
communication and persuasion that defines persuasive communication as the process by
which an individual (the communicator) transmits stimuli to modify the behavior of other
individuals (the audience). Persuasive communication leads to attitude change through
changing related opinions. Opinions, according to Hovland et. al. (1953), can be changed
by exposure to persuasive communication that contains information and arguments why a
new opinion should be accepted. Bayou, Mohamed, Panitz and Eric (1993) defined

persuasion as a communication strategy designed to satisfy the parties involved by

ac?

p61

int

 

Opi

det

IOV

eff,

influencing attitudes and beliefs therefore behavior, which is what marketers hope to
achieve through their advertisements. The instrumental theory of communication and
persuasion as illustrated in Figure 1 divides the communication and persuasion process

into three sections namely stimuli, intervening process and response.

 

 

 

 

 

Stimuli Intervening Processes Response

Sources factors ——> Opinion change
Expertise,
Trustworthiness
likability \
Perception

change

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Message factors
Type Of appeal Comprehension [

\ Affect change

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Audience Factors

Initial Position A

Intelligence ‘ Acceptance I: Action change
Self esteem 7V

Personality etc

 

 

 

 

 

 

 

Figure 1: : H-J-K Instrumental theory of communication and persuasion model (Fan, 1985)

This theory has several important aspects relating to this study. First, according to the H-
J -K theory, communication success is not only measured in action change, but also in
opinion, perception and affect change. Since, one of the objectives of this study is to
determine the effectiveness of specific advertisements in changing people’s attitudes
towards Michigan as a tourist destination, according to this theory, the advert is said to be

effective if it results in a significant positive change in people’s attitudes and perceptions

 

 

10
st:
in

ad

the

Th.
ind

adx

hell
wh
this
Var
Ho
bel
ext.

lde.

towards Michigan as a vacation destination. Second, this theory includes attention as a
stage of the intervening processes. Advertising awareness, a variable also being measured
in this study, does not occur without the target audience paying some attention to the
advertisement. Third, the theory places importance on the effect of audience factors such
as initial position, intelligence, personality and self-esteem on the results. This supports
the hypothesis that individuals who have visited Michigan before, because of their
knowledge and experience, will have more positive attitudes towards Michigan, as well

as greater awareness of Michigan advertising.

The H-J-K theory also forms the basis for hypotheses 2 and 3 which state that: 1)
individuals will rank Michigan higher as a tourist destination after exposure to
advertising and: 2) that individuals who are more aware of the advertisements have
greater motivation to visit Michigan. The H-J-K theory of persuasive communication
helps us understand why individuals change or don’t change their attitudes and/or actions
when exposed to an advertisement. This theory is important, especially considering that
this study involves some measure of the effectiveness of advertising by comparing
variations in some important variables before and after exposure to advertisements.
However, the H-J-K theory of persuasive communication, like the other learning and
belief-based theories of persuasion, is based on the notion that persuasion depends on the
extent to which message recipients learn from, form beliefs about, as well as retain the

ideas conveyed by the message (Hovland, Janis and Kelly, 1953).

10

The information-processing perspective, which includes the co gnitive-response model,
directly criticize learning belief-based theories such as the H-J-K theory (Meyers- Levy
and Malaviya, 1999). The Cognitive-Response Model of persuasion states that persuasion
occurs as a result of people’s reflections on, as well as cognitive responses to, the
contents of a message (Wright, 1980). Cognitive responses are defined as the thoughts
that arise during the process of elaboration when people relate the contents of the
message to other messages or to their pre-existing knowledge and views stored in
memory. Persuasion in this case is a reflection of the net favorableness of the cognitive
responses that are evoked in people’s minds as they elaborate on the message, and a
decision is based also on the net favorableness of the information available in memory at
the time of judgment (Kisielius and Stemthal, 1986; Keller, 1987; Meyer-Levy and
Malaviya, 1999; Weilbacher, 2002). Cognitive-response theory, however, has been
criticized mainly for its failure to explain situations in which people are persuaded even
when there exists enough evidence to show that they did not elaborate on the contents of

the message.

Dual process persuasion models were prompted by the failure of the cognitive-response
models to address those situations in which individuals are persuaded even though they
don’t seem to have elaborated much on the contents of the message. The dual process
models such as the Elaboration Likelihood Model (ELM) of persuasion suggest the
existence of a systematic (central) and a heuristic (peripheral) route to persuasion (Petty
and Cacioppo, 1986). The systematic route produces judgments that are based on critical

and extensive elaboration of messages while the heuristic route produces judgments that

11

are based on simple and intuitive inferences involving scanty elaboration (Meyers-Levy
and Malaviya, 1999). Which of these two routes mediates persuasion at any given time is
affected by a number of factors influencing the amount of cognitive resources that an
individual devotes to the message elaborative process. These factors include, but are not
limited to, available time, message relevance, experience and prior knowledge. That
experience and prior knowledge affect persuasion becomes important to this study when
it comes to the differences in advertising awareness, destination awareness and attitude
towards Michigan among those individuals that have visited the state and those that
haven’t. It is also important in understanding the differences among those who have been
exposed to the advertisements more as compared to those who have been exposed less.
Individuals with prior knowledge and/or more experience with the state are likely to be
more aware of the state’s advertising than those that have neither the experience nor prior

knowledge.

Another theory, the experiential based theory of persuasion, has been suggested as a
third and alternative basis for judgment and persuasion. This theory, unlike the systematic
and the heuristic approaches which emphasize the cognitive process of persuasion, states
that judgments are mediated by the sensations and/or feelings that are triggered by the act
of engaging in an elaboration process, (Strack, 1992). Meyers—levy and Malaviya (1999)
developed an integrative framework of advertising persuasion shown in Figure 2 that
incorporates all the other theories and goes even further to include a judgment correction
phase during which individuals reconsider their decisions and make adjustments in an

effort to eliminate biases that might have occurred in the initial elaboration process. This

12

 

fran

in \x

01h:

TCIT

CUS

inte

 

 

framework was developed as a result of the earlier theories’ failure to explain situations
in which individuals change their decisions after having been persuaded and convinced
otherwise. This theory highlights to marketers the importance of continual advertising to
remind customers of products and services because even if they are initially persuaded,
customers always change their judgments and decisions with the passage of time. The
integrative framework shown in Figure 2 is divided into two main stages: judgment

formation and judgment correction.

13

 

 

 

ll

 

 

 

PERSUASION VARIABLES I MessagerecrprentContext

 

 

 

 

I

Resource Allocation for . .
Message Processing_;~

Minimal / ‘ Modest \Substantial

’E’i‘pcfimfiammmmg Hemficméfising " 3 iii-Systematic mine
Strategy Strategy 1 _ , Shategy .. .

 

JUDGMENT FORMATION

 

 

 

 

 

 

 

 

 

 

 

     
     

 

   
 

Attributions
abo

 

   

Item-Specific. f
Elaboration

  
      

Relational V
Elaboration

 

 

 

 

 

v /

EMP LOY '
DEFAULT VALUE
AS INITIAL

 

JUDGMENT

 

 

 

 

PRODUCEWIUDGMENT; ; .5: ..I

 

 

 

 

 

 

 

 

“// JUDGMENT CORRECTION STAGE
m , , . . .

 

FORM FINAL JUDGMENT

 

 

 

 

Figure 2: An Integrative Framework of Advertising Persuasion (Meyers-Levy and Malaviya, 1999)

14

 

 

2.1.

An(

sele

ind:
thei
ana
and
sele
affo.

with

2.1.3
Studi
Most
Simil;
a nutr
earlier
explai
110110
as adi
mam)

”2383 i

 

2. 1.2 Selective Processes Theories (Cognitive Dissonance Theory)

Another important theory relating to this study is that of selective processes including
selective attention, selective perception and selective retention (Haley, 1985). This theory
agrees with Festinger’s (1957) theory of cognitive dissonance, which states that
individuals deliberately attend to or avoid messages and situations that do not conform to
their beliefs, lifestyles and needs (Severin, 1979). One objective of this study is to
analyze the relationship of some demographic factors such as income with advertising
and destination awareness. Some of the variation here may be explained by the theory of
selective attention and retention. For example, individuals with low income and can not
afford vacations may not necessarily pay attention to travel advertisements while those

with higher income would be more likely to tune into travel advertisements.

2.1.3 Communication theories

Studies to evaluate the effectiveness of advertising on sales date as far back as the 19305.
Most of these studies focused on consumer products, generally single or a small group of
similar products (Butterfield et al., 1998). Advertising has been known to affect people in
a number of different ways as hypothesized by different researchers over the years. The
earlier black box theories of consumer behavior were some of the first ones to try to
explain the impact of advertising on peoples’ behavior. These models were based on the
notion that individuals show some kind of response when exposed to a set of stimuli such
as advertisements. They also assumed advertising operates in a straight-forward one way
manner in which the advertiser decides on the message, encodes it and delivers it through

mass media to a basically passive or receptive audience (Haley, 1985).

15

One of the popular earlier models is the AIDA (attention, interest, desire, action) model,
which assumes that when individuals are exposed to advertising communication, they go
through a series of stages namely Attention, Interest, Desire, and Action. Accordingly,
the purpose of advertising therefore is to catch people’s attention, leading them to
develop some interest and desire in the product resulting in some action (purchase of
product) (Haley, 1985). The Hierarchy of Effects model was developed as an upgrade of
the AIDA model. It is a straight forward model consisting of six stages namely Attention,
Interest, Comprehension, Impact, Attitude and Sale (Lavidge and Steiner, 1961). The
main weakness of these straight-forward models lies in their failure to account for
situations in which these steps are skipped and/or reversed as evidence has shown that

frequently occurs (Ray, 1973; Aaker and Day, 1974).

New models were developed which sought to address the weakneses of the previous
models. The new models, such as the Richardson-Haley model (1980), have their roots in
the two-way communication process first suggested by Bauer (1964) based on the
understanding that consumers are not inert passive targets for messages that can be
manipulated by advertisements. According to Bauer (1964), message recipients are active
participants in the communication process, screening, distorting, adding and subtracting
contents of the messages that they are being exposed to. This theory agrees with the
selective processes theories that, when individuals are exposed to a message including
advertising, they receive what they want to receive and believe what they want to believe
(Haley, 1985). The Richardson-Haley model emphasizes that the message recipients

when exposed to advertising are not in a blank state, but bring their own experiences,

16

values, interests, personality, lifestyle, moods and habits. This is especially important to
advertisers because it highlights advertising has different effects on different people and
audiences, there really isn’t a ‘one size fit all’ approach when it comes to advertising.
This theory is also important in relation to this study as it makes it worthwhile to
investigate the importance of each of the different individual characteristics such as

attitude and prior knowledge to advertising effectiveness.

Recent theories of how advertising works have their roots in two-way models of
communication based on the belief consumers are not inert, passive recipients of
advertising. What’s important when individuals make decisions is not the last advert they
were exposed to, but also what is remembered from previous advertising, past
experience, and a lot of other information about the brand that the customer holds from
other non-advertising communications out of the marketer’s control (Weilbacher, 2003).
According to Weilbacher (2003), the human brain is complex enough to process a wide
variety of incoming information, and, at any given moment, an individual perceives a
limited amount of information that has been selected by the brain from all the incoming
information fi'om the external environment. Exactly what an individual perceives depends
on what the mind remembers from the past, as well as the relevance of this information.
Weilbacher’s argument is in line with the theories of selective processes in that
individuals choose what they do or do not attend to. These new developments in
communication theory make advertising an even more complicated task for marketers.
Marketers will find it more challenging to reach their customers, especially considering

today’s consumer is bombarded with thousands of advertisements per day.

17

2.2 Does Advertising Really Work?

A lot of studies have been conducted on advertising effectiveness using different
methodologies. Most studies have been focused on consumer goods in general rather than
on tourism and or destination marketing. The questions that most advertising and/or
marketing departments and agencies have to answer are; does their expenditure on
advertising result in more sales or in increased visits in case of tourism? Does an increase
in the advertising budget, or an increase in the frequency and size of advertisements,
result in more sales or visits? Responding to these questions has become even more
important for government agencies that rely on public funds for their advertising budgets

in these times of budget crunches.

Simon and Amdt (1980) in their study on the shape of the advertising response function,
concluded that there are no increasing returns to advertising and that diminishing returns
characterize the shape of the advertising response function. Their study involved a
review of over a hundred studies examining advertising effectiveness. The advertising
response function they used referred to the quantitative relationship between advertising
inputs such as monetary expenditure, size and fi'equency of advertisements and outputs
such as sales, ad recall, attitudes and intention to buy. They proposed the advertising
response function is a concave-downward function. In the concave firnction, there are
diminishing returns based mainly in the micro-economic law of diminishing returns.
Their argument for the concave shape is that, for an individual consumer, a given

message conveys less and less information with each additional exposure. This finding

18

agrees with Krugrnan’s (1972) argument for the need for two or three exposures. He
states the first exposure creates curiosity, the second one brings recognition and the third
one clinches a decision while any extra ones have little value. This led to the theory that
there are decreasing returns to frequency of advertising. In the S—shaped function,
however, there are increasing returns first and then diminishing returns after an inflexion
point is reached. Other researchers such as Chamberlin (1962) and Rao (1970) advocate
for the S—shaped advertising response firnction. Their arguments for the S-shaped
response function are based on the notion there exists a threshold effect. In support of the
S-shaped advertising function, Chamberlin (1962) argued a consumer’s consciousness
has to be gained first, and, while it is being gained, additional expenditure in advertising

yields increasing returns.

While both these functions show there are some returns to advertising, they are more
related to consumer products, and their applicability to tourism advertising is not clear.
Tourism is unique in that once a person has visited a destination, especially long haul
destinations, they are not likely to visit the same place again, since long haul travel is
typically not a repeat visit to a familiar location, but an exploration of new and unfamiliar
parts of the world and experiences (Tourism New Zealand, 2005). Additional advertising
in this case might not result in any positive effects. With consumer goods, however, when
consumers are impressed alter the first purchase, they are likely to include that product
on their short list of preferred brands and are also more likely to repurchase the same

brand or product the next time they need it or if they run out of it (Wansink and Ray,

19

 

2000.

the cc

Advc
inclu<
(1972
such :
Irelan
effect

effect

Ehren

rather

alread

 

2000). In this case, repetitive exposure to advertising might serve a purpose of reminding

the consumer of the brand.

Advertising has been found to affect market share and firm sales in particular industries
including the beer and cigarette industries (Palda, 1964; Peles, 1971). Barry and O’Hagan
(1972) used log linear models to examine the effect of advertising and other variables
such as income and price on the number of visits and expenditures by British tourists in
Ireland. The advertising variable was found to be significant, implying advertising had an
effect. Other studies such as the one by Uysal and Crompton (1984) have also shown the

effectiveness of increased promotional expenditures on tourist arrivals and expenditures.

Ehrenberg (2000), however, does not support this assumed power of advertising, but
rather argues advertising’s role is mainly to reinforce feelings of satisfaction with brands
already bought. This argument is consistent with the theory of cognitive dissonance
which states that individuals selectively attend to advertisements of products they have
already purchased in search of information that reassures them that they made the best
decision hence reduce post-purchase dissonance. Ehrenberg (2000) argues advertising,
though effective, is not as powerful as often thought and that there is no evidence that it
works by any strong form of persuasion or manipulation. According to Ehrenberg (2000),
advertising works in an Awareness-Tnal-Reinforcement (ATR) sequence. Consumers
first gain awareness and then they make a trial purchase and finally a repeat buying habit
may develop and be reinforced if the consumer is satisfied with the product. Even though

advertising has a role in each stage of the purchase decision process, he argues repeat

20

buying is the main determinant of sales volume and advertising’s role is to reinforce
rather than persuade. Ehrenberg also criticizes the traditional Awareness-Attitude-
Behavior thinking of how advertising works. He argues peoples’ attitudes can’t be readily
changed, and even if they change, there is no evidence to support that attitude change
precedes desired purchase behavior. These arguments contrast with other researchers who
have found attitude change is an intermediate measure of advertising effectiveness and
predictor of end behavior (Lavidge and Steiner, 1961; F ishbein and Ajzen, 1975; Tan,

1985; Butterfield, Deal and Kubursi, 1998, Wansik and Ray, 2000).

Other researchers (Schmalensee, 1972; Comanor and Wilson , 1974) joined Ehrenberg
(2000) in expressing their skepticism about the effectiveness of advertising. They argue
the strong correlation between advertising and sales could be a result of the effects of
sales on advertising rather than advertising on sales as widely believed. Both
Schmalensee (1972) and Comanor and Wilson (1974) conducted studies using
simultaneous equations in which they included both effects, and they found no conclusive

evidence that supported advertising had an effect on sales.

Whether or not advertising is effective depends on one’s objectives. The important point
to note is that it is vital that one considers the initial goal of running an advert when
evaluating whether it has been effective or not. This not only ensures that the right
conclusions are drawn, but it also enables accurate selection of proper methods to use in
evaluating the effectiveness of the advert. For example, it is not fair to evaluate the

effectiveness of an advert meant only to generate awareness on the basis of increased

21

sales or visits. The advert might not have resulted in increased sales but might have
generated awareness of the product, brand or destination. In this case, if an appropriate
method to measure change in awareness is not utilized, then the advert might be
considered inefi’ective. Lavidge and Steiner (1961) gave examples as shown in Table l of

appropriate research approaches to evaluate different kinds of advertising goals.

Table 1: Advertising, Related Behavioral Dimensions and Advertising Research related to the
Model

 

 

 

Related Behavioral Movement Toward Examples of types of Examples of research
Dimensions Purchase promotion/advertising approaches related to
relevant to various steps steps of greatest
applicability
Pur hase Point of purchase Market or sales tests
CONATIVE Retail store advertisements Split run tests
-the realm of motives Deals Intention to purchase
Advertisements ‘Last chance’ offers Projective techniques
stimulate or direct sales Price appeals
Conviction Testimonials
Preference Rank order of preference
A Competitive for brands
AFFECTIVE advertisements Rating scales
-the realm of emotions. Argumentative copy Image measurements
Advertisements change including check
attitudes and feelings Image advertisements lists and semantic
Status, glamour appeals differentials
Liking Projective techniques
Knowledge Announcements Information questions
COGNITIVE Descriptive copy Play-back analyses
-the realm of thoughts Classified advertisements Brand awareness surveys
Advertisements provide Slogans, Jingles Aided recall
information and facts Sky writing
Awareness Teaser campaigns

 

Lavidge and Steiner (1961)

According to the above model, there are different methods that can be used to evaluate

the effectiveness of advertisements depending on the goal. For example, to create

22

awareness of a product, an advertiser can run a descriptive advert and then evaluate its
effectiveness through aided recall research. It is, therefore, not appropriate to conduct
market or sales tests meant to evaluate point of sale advertisements to determine the
effectiveness of a descriptive advert whose objective was to create awareness. Each stage,
according to Lavidge and Steiner, is important and relevant as it is supposed to
progressively lead to the final point which is ‘purchase.’ It is important, therefore, to
clearly state the objectives of advertising as well as to identify a suitable method to
evaluate the effectiveness of the campaign prior to launching the advertising campaign.
Doing so not only simplifies the evaluation process but also insures that the right

conclusions concerning the campaign are drawn.

2.2 Travel Advertising Effectiveness Methods Review

Many studies have been conducted on advertising effectiveness in general. However, not
many have been conducted that deal specifically with travel advertising. Several methods
have been utilized by different researchers to determine the effectiveness of travel
advertising, not without limitations, however. Conversion and tracking studies have been
the most popular approaches that have been used to determine the effectiveness of travel
advertising (Woodside and Reid, 1974; Ballman et. al., 1984; Butterfield et. al.,1998;

Zhou, 2000).

Advertising is meant to achieve a wide range of different goals including but not limited
to: creating awareness, reminding, changing attitudes and increasing sales and/or visits.

Different methods and approaches can be utilized in addition to the tracking and

23

conversion models to determine if travel advertising is effective. The advertising tracking
model evaluates advertising effectiveness through a series of stages from exposure to
advertising to conversion behavior. The advertising tracking model assumes consumers
may be converted by advertising solely on the basis of awareness and image building.
The conversion model on the other hand, measures advertising effectiveness through a
‘funnel’ process through which consumers are assumed to go from exposure to the final
goal of conversion. The conversion model process is linear, sequential,l and the group of

consumers gets smaller as they get closer to conversion.

Even though conversion and tracking studies have been the popular methods of
evaluating travel advertising effectiveness, they have been criticized for a number of
reasons. First, conversion models do not provide a quantitative figure on return on travel
advertising investment (ROI). They can provide diagnostic information to decision
makers. Therefore, the use of conversion models to produce an estimate of R01 is both
misleading and misguided (Burke and Gitelson, 1990). Second, conversion studies suffer
fi'om non-response bias, especially in mail studies (Woodside and Ronkainen, 1984;
Ballman et. al., 1987). Third, the conversion model puts emphasis on inquiry fulfillment.
According to the conversion model, inquiry is a necessary step in selling a destination
and must occur prior to visitation (Siegel and Ziff-Levine, 1990). The model, therefore,
assumes all visitors exposed to travel advertising must inquire for travel information
before visiting. This is a major shortcoming of the model for which it has been criticized
because not all advertising driven visitors inquire before visiting (Siegel and Ziff-Levine,

1990). Also, by making inquiry a requirement, the model underestimates the effects of

24

advertising. By making inquiry a necessary step, the conversion model also downplays
the effects of advertising on creating destination awareness and creating positive
destination image, the two stages that precede the inquiry fulfilhnent stage in the model.
This model, therefore, is most useful in cases where the study’s objective is to measure
the effect of travel advertising in generating inquiries and to provide diagnostic

information on advertising effectiveness.

The tracking model on the other hand does not require all advertising generated visitors
fulfill the inquiry requirement. The tracking model assumes that advertising generated
visitations do not depend on inquiry fulfillment. Visitors may or may not inquire before
visitation. The belief is that consumers can be converted by advertising as a result of
awareness and positive image (Siegel and Ziff-Levine, 1990). However, the tracking
model is also not suitable when an ROI estimate is sought. The tracking model is suitable
when the goal is to measure the effect of advertising on; generating advertising
awareness, generating destination awareness, creating positive image, motivating
consumers to visit the destination and influencing travel behavior (Siegel and Ziff-

Levine, 1990).

Linkage advertising has been suggested as one of the most effective when it comes to
measuring the effects of advertising on sales. Linkage advertising makes it easy to
evaluate the effects of advertising on sales because it links the up-front advertising to the
sale (Rapp and Collins, 1987). Rapp and Collins (1987) argued too often advertising

leaves the prospective consumer dangling without any idea what to do next, where to buy

25

or how to obtain additional information. Therefore, marketers should seek to bridge this
gap between advertising and sale by providing more information through linkage
advertising. While linkage advertising might work well for other industries its
applicability to tourism and destination marketing is complicated. State and provincial
destination marketers’ hands are tied because, as much as they are well aware of the main
attractions in their regions, they are often forbidden to mention them due to political
restrictions on highlighting individual cities and attractions. State tourism agencies,
therefore, in their efforts to lure visitors find their advertising ludicrously vague about

‘why’ (Garfield, 1994).

As much as linkage advertising is important in bridging the gap between advertising and
purchase and is certainly suitable for evaluating the effect of advertising on sales, it has a
shortcoming in that it fails to establish causal relationships between advertising and
purchase (Woodside, Trappey and Macdonald, 1997). To address this problem, the use of
true experiments is recommended. For an experiment to be considered true, it has to meet
certain criteria including, presence of a control group and random assignment of subjects
to the control and experimental groups (Trochim, 2001). Other researchers, such,as
Raymond (1974), Caples (1974) and Woodside (1990), have advocated for the use of true
experiments to determine the causal relationship between advertising and sales. The use
of experiments to determine travel advertising effectiveness is, however, not very
practical and is further complicated by the expenses associated with meeting the

requirements of true experiments (Messmer and Johnson, 1993; Woodside et al., 1997).

26

Quasi-experiments have been used as an alternative to true experiments. Quasi-
experiments are tests of the effects of changing levels of outcome variables caused by
treatment variables when random assignment has not been used to create equivalent
comparison groups (Woodside et al., 1997). Woodside et. al. (1997) used a quasi-
experiment to measure the effectiveness of linkage destination advertising on tourist
behavior and expenditures at Prince Edward Island, Canada. They chose the post-test
only non-equivalent groups design with higher order interactions to determine with
higher accuracy the cause-effect relationship between the advertisements (printed visitors
guide) and the outcome behavior (expenditure). The results of their study showed linkage
advertising had an indirect effect on expenditures through participation in more activities
at the destination. Those individuals who had acquired the guide participated in more
activities and spent more than those who had not acquired the guide. Use of the higher
order interactions made it possible to establish the cause-effect interpretation. The results
showed individuals with knowledge and experience participated in more activities at the

destination.

Regardless of the type of method utilized it is important to note all approaches have
weaknesses and/or limitations. The problem with the supposedly better and more accurate
methods is that their applicability to the travel and tourism field is complicated and costly
to apply. It is, therefore, of particular importance that the objective of the campaign be
clearly stated before initiating an assessment of advertisements so that an appropriate
evaluation method can be employed. It is important also to recognize there is no single

faultless method to evaluate advertising effectiveness.

27

Chapter 3

3.0 Methods

This chapter discusses the methods that were used in this study. The study design is
discussed first, then the data collection methods. A discussion on the techniques of data

analysis concludes the chapter.

3.1 Study Design
The study was a pre and post-test with no control group as depicted below:
R 01 X 02

Where: R = Random selection of subjects

01 = Observation 1 (phase 1)

02 = Observation 2 (phase 2)

X = Treatment (exposure to advertisements)
The study was cross-sectional not panel, implying two different groups of respondents
were randomly sampled in each wave. Since the study was measuring awareness a panel

study would have been inappropriate, as the subjects would become sensitized.

3.2 Data collection

Data for this study were collected through telephone interviews in 2003 from three
Designated Marketing Areas (DMAs) including Chicago, Cleveland, and Indianapolis.
The sampling frame included all residents of the three DMAs with a land-line telephone.

The respondents’ phone numbers, which in this case also represented the study sample,

28

were purchased from Survey Sampling International, Inc. which drew random-digit—dial
phone samples from the three DMAs. The data were collected in two waves. The first
(pre-campaign) wave was conducted between April 8 and May 1, 2003 and the second
(post-campaign) wave was conducted between July 7 and July 26 for Cleveland, between
July 27 and August 26 for Indianapolis and between July 27 and September 2 for

Chicago.

The first wave was conducted simultaneously in the three DMAs, but the second wave
was conducted at different times in each DMA depending on the time the advertisements
were placed in the media. Even though the initial target of 1,800 completed surveys (300
per DMA per wave) was not reached, a total of 1,117 surveys were completed, enough to
conduct reasonable analysis. The 900 target in the first wave was not reached mainly due
to administrative reasons. The final contract was agreed upon and signed too late to have
allowed more pre-campaign surveys to be completed. The time between the beginning of
the survey and the launching of the campaigns in the three DMAs was not long enough to
allow more surveys to be completed. However, the complete second wave target of 900

was met.

The survey instrument was a questionnaire consisting of 37 questions that were identical
across the three DMAs except for the questions that identified and described the different
specific advertisements that were run in each DMA (refer to Appendix 1 for the full
questionnaire). A complete interview lasted approximately 10 minutes and there were no

incentives offered to the respondents for participating in the study.

29

Three different communication channels were utilized for the 2003 travel advertising
campaign including Television, radio and magazine. The specific content of the
advertisements varied among the three DMAs and the channel of communication.

Appendix 3 gives a detailed description of the advertisements by channel and DMA.

Advertising Awareness was measured at two levels by measuring unaided and aided
recall of Michigan advertising. For unaided awareness respondents were asked to name
the states for which they recalled seeing or hearing any vacation or travel advertising
within the past six months. The following question was asked to prompt unaided recall,
‘For which states or provinces do you recall seeing or hearing any vacation or travel
advertising in the past six months?’ If the respondent mentioned seeing or hearing any
Michigan advertising, they were then asked to describe in general what the advertising
said and what message they thought it was conveying. A nominal measure was used to
measure unaided recall coded as; 1 if they recalled seeing or hearing Michigan
advertising and 0 if not. To measure ‘aided recall’ the interviewer would describe a
specific advert and then ask whether or not the respondent recalled seeing or hearing that
particular advert in the past six months. The following question was asked to prompt
unaided recall, ‘ Do you recall seeing or hearing any vacation or travel advertising

for. . .in the past six months?’ respondents were also asked to state whether or not they
had heard and/or seen specific television, radio and magazine advertisements that the
interviewer read and described for them. A nominal scale was also used to measure aided

recall coded as; 1 if they recalled seeing or hearing Michigan advertising and 0 if not.

30

Destination Awareness was also measured nominally, 1 if Michigan was mentioned as
one of the states that come to mind when thinking of taking a pleasure trip and 0 if not.
The following two questions were asked to determine respondents’ awareness of
Michigan as a tourism destination;
1. ‘When you are thinking of taking a pleasure trip in the US. or
Canada what cities, states or provinces come to mind?’
2. ‘Now thinking just about a pleasure trip in the Mid-west what cities
or states come to mind?
There were several other variables that were measured at different levels using different

scales as shown in Table 2.

31

 

 

basin ”25A seal :85 “8885 $8 .8 c.0803
0:80TN
0:8"— 1.880 Z 8088005 0800
anm 808 UN
anmefivm n_
vam n 8 v no 3:on 008 8 8080008 ”—0880 80809088 0885
v.0» H_ 3808 2 «men 05 8
o: Ho 0% D03 8288 3050 P033 08 8053 m8 080 8: 8 08055 ”—8802 8080385 08 9.0?
m0» H_ 888888 #028 80%
m0.» no 0868 3 08— 0:28 00b :8 90003 08 wow: 08 “on 8 08055 ”—8.882 80808005 885
“echo 088“me
3880 2.200% 88 23M .>.H 88208 20880 05 .8 8a 8988 0808080038 :2
85 awn—85 we .8 0550"— w880: 8 8000 0:308 808338 a «o: .5 8050:? 3 808308 ”108802 8080305 #0880 w8m80>v<
m0>H_
0noeno 08809 88 08808 a no :2 00:?» m8 080 8: no 05055 ”38802 8080908: 00:0t0axm
:03 boo—=8 088 Mm £808 fi 808 808308 Cr: 08%
05> 308608 53 H_ 05 8 88 08.80— m co 88m 05 28> 2 £808 080 8: co 805055 ”—88.5 0% 808039: 05 88> 3 82808—
858080 +_ 86.8260
0co=no 888080 H2 88: no 8000 m8 080 0088 .8 8388 08 .3 808808 noted 80809005 8 088%.”.—
3203800 00&<um 808090D 8288030 8889 n
3000888 00%85l manage 80.5% .8 00m 0 so HE 3588 .3 808808 ”—885 0% 8080908: 3 :2 8539 0—0832
8.: 08303 0 mg 29
Echo ”033ml Co 9388 80:3 8.8 88 888 :2 8: no 080:3 .3 808808 ”—8802 8080Q0Q mm080~03< cowgumom
A3:
Echo ”033ml 30:033.“ c0283 88 80.08 .3 808308 ”38802 8080Q0Q mm0e083< 858032.
808038—
0_a0m 80808308 .8 .054 808080Q 03mca>

 

80808808 me $32 805 98 808.808 83.88; ”m 03a.

32

3.3 Data analysis techniques used

Data analysis was conducted in stages to fiJlly address each hypothesis and meet the
objectives of the study. The first stage of data analysis included descriptive analysis
where the general characteristics of the data were described. Demographic information

for respondents was presented in tables and charts at this stage.

The second stage of analysis included some pre and post advertisement (phases 1 and 2)
comparisons to determine if there were any significant differences in advertising
awareness before and after exposure to advertising. The Mann-Whitney U test was
employed at this stage to detect any significant differences in advertising awareness
between phases 1 and 2. The Mann-Whitney U test was used as it is more appropriate for
nominal variables than the T-test. This was an important step in the analysis because it
was the first test of the effectiveness of the advertising campaign. If no significant
differences were found in advertising awareness before and after exposure, then any
further analysis would not be meaningful. This step, therefore, determined if Travel
Michigan’s travel advertising was successful in breaking through all the clutter and

getting the audience’s attention.

The third stage of analysis included testing the effectiveness of Travel Michigan’s
advertising on; creating destination awareness, improving people’s attitude towards
Michigan as a tourism destination, creating some motivation and intention to visit the
state, and increasing visitation to the state. A T-test was employed in this stage of

analysis. After having managed to capture the audience’s attention, this stage of analysis

33

sought to determine if there were any significant differences in the above mentioned

variables between those aware and those not aware of advertising.

Next, factor analysis was conducted to reduce the 22 attitude towards Mi as a tourism
destination factors to a smaller more manageable set to be used in subsequent analyses.
Principal components analysis instead of common factor analysis was used as the
extraction method because the focus was on both data reduction and total variance not

just common variance.

The next stage of analysis involved correlation analysis to determine the kind of
relationship between attitudes towards Michigan as a tourism destination, distance of the
DMA from Michigan, annual income and advertising awareness, destination awareness
and intention to visit the state. The Spearman’s rho correlation test was utilized at this
stage because most of the variables tested were either nominal or ordinal. The attitude

factors retained from the prior principal components analysis were utilized at this stage.

The sixth stage in the analytical framework that was employed was the model estimation
stage in which two models were estimated. The dependent variables for these two models
were; 1) Advertising Awareness (AA) and 2) Destination Awareness (DA). Logistic
regression was used for the AA and DA models since the variables were binary, bound by
O and 1. The forward greatest reduction in Log Likelihood value (-2LL value) method
was used to enter variables into the logistic model. The independent variables

hypothesized for the AA model included: exposure, TV, radio, magazine, attitudes,

34

experience, location, income, intention to visit (IV), age and gender. Independent
variables used in the DA model included: IV, exposure, location, web-use, experience,

and AA.

In the next stage of analysis, the real conversion ratios for the 2003 Travel Michigan
advertising campaign were calculated by use of the conversion and tracking models by
Siegel and Ziff-Levine (1990) which were presented in figure 3. The final stage of
analysis for this study involved the calculation of the return on advertising investment for
the 2003 Travel Michigan’s advertising campaign. This stage utilized the conversion
figures derived from the previous stage of analysis, 2003 advertising expenditure figures
from Travel Michigan, and tourist expenditure figures from the household survey
conducted by the Michigan Travel, Tourism and Recreation Resources Center of the

Department of Parks, Recreation and Tourism Resources at Michigan State University.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Advertesing
. _ . Destination
Target audience Advertismg Awareness
Awareness of advertising f Ar Motivation
. . ' Positive ,
Destination awareness Advertising image of x"
+ Awareness destination ,v’
Positive lmagepf destination x’
Inquiry fulfillment ,
* Inquiry C .
. . Fulfillment onverSlon
Motivation ------ p behavior
Conversion behavior
(A) (B)

 

 

 

Figure 3: (A) Conversion Model (B) Tracking model (Adapted from Siegel and Ziff-Levine, 1990)

35

Chapter 4
4.0 Results and Discussion
This chapter covers the findings of the study. The descriptives are discussed first
followed by the pre and post campaign comparison results. Hypotheses testing results are
discussed next, followed by the conversion estimation results. The chapter is concluded

by a discussion of the results on the estimation of actual return on advertising investment.

4.] Descriptives

There were a total of 1117 respondents for this study of which 209 were interviewed in
the pre-advertisement phase (phase 1) and 908 were interviewed in the post
advertisement phase (phase 2). The reason for this major discrepancy is mainly technical.
The contract for the project was not approved until very late in phase 1 and real data
collection did not start until there were only three weeks before the advertisements started
running in the three DMAs. As a result the target of at least 300 interviews per DMA per
phase originally set was not met. Since the two pre- and post campaign samples were
cross-sectional and random no adj ustments were made during the analyses to reduce the
post campaign sample size to match the pre-campaign one. Despite this limitation, the

sample was large enough to enable reasonable analyses and comparisons to be conducted.

The study was cross-sectional not panel implying the pre and post groups were different
but each was randomly selected. Since the study was measuring awareness a panel would

not have been appropriate, as the subjects would become sensitized.

36

There were more female respondents than there were male. Of the 1117 respondents,
62% were female and 38 % were male. Of the 209 phase 1 respondents, 43% were male
and 57% were female. For phase 2, 37% were male and 63% were female. The average
age of the respondents was 46 years. Distribution of respondents in phases 1 and 2
combined among the three DMAs was almost evenly distributed with each DMA

representing about a third of the respondents as can be seen in figure 4.

Indianapolis 34%

Chicago
32% —

Cleveland

 

 

34%

Figure 4: Distribution of Respondents among the three DMAs

A comparison of gender among the three DMAs showed there were more female
respondents than there were male in each DMA, and the proportion of females to males

was exactly the same for Chicago and Indianapolis and slightly different for Cleveland

37

(see Table 3). Other respondent characteristics and demographics such as income,

household size and ethnicity varied across the three DMAs (Refer to appendix 3).

Table 3: Gender * Number of respondents by their residence DMA

 

 

 

 

 

DMA of Residence Total
Chicago Cleveland Indianapolis
Gender Male 134 146 146 426
Female 222 229 238 689
Total 356 375 384 1115
Male % 38% 39% 38% 38%
Female % 62% 61% 62% 62%
Total % 100% 100% 100% 100%

 

 

 

*Gender was not recorded for two Chicago DMA respondents

4.2 Pre and Post Advertisement Comparisons

4.2.1 Advertising Awareness

The first stage of analysis was meant to determine whether Travel Michigan’s advertising
drew the attention of its intended audience, by testing whether there was any significant
difference in Advertising Awareness before and afier the advertisements were placed.
Without people being aware of the advertising in the first place, any further tests to
determine the effectiveness of the advertising would not have been meaningful. The
Mann-Whitney U non-parametric test was conducted to determine if there were any
significant differences in Advertising Awareness (AA) between phases 1 and 2 of the
advertising campaign. A non-parametric test was used because it is a more appropriate
method when working with nominal data than is the T-test. Since AA was measured
nominally (0: not aware; I: aware) a non-parametric test was appropriate. The results
showed a significant difference in Advertising Awareness (U: 72830, p< .01) between
phases 1 and 2. What this result shows is that people were aware of Travel Michigan’s

advertising after the advertisements were run. The advertising had the initial effect of

38

catching people’s attention. This is a good starting point in the evaluation of the
effectiveness of advertising. As illustrated in the models of communication in the
preceding chapter, the initial goal of effective advertising is to grab the attention of
consumers. The AIDA (Attention, Interest, Desire and Action), Hierarchy of Effects and
the Richardson-Haley models indicate getting the attention of the consumer, getting them
to even pay attention, is the first and foremost important step in effective communication,
let alone persuasion. Getting the attention of the consumer is a necessary though not
sufficient condition for effective advertising. Significant differences were also found in

Destination Awareness (DA) (U=48229; p < .01) between phases 1 and 2.

4.3 Hypothesis Testing

This study’s objective of determining the effectiveness of Travel Michigan’s advertising
follows the conversion and tracking models suggested by Siegel and Ziff-Levine (1990)
as illustrated in Figure 3 that was presented earlier. For this study, advertising
effectiveness was measured in several ways including Advertising Awareness (AA),
Destination Awareness (DA), Attitudes, Intention to Visit the state (IV) and Actual
Visitation (AV). This stage of analysis focuses on determining the effectiveness of Travel
Michigan’s advertising on creating awareness of Michigan as a tourism destination,
improving people’s attitudes/image of the state as a tourism destination, developing
motivation to visit the state among the targeted advertisement audiences and stimulating
change in consumptive behavior once motivation is developed. This stage of analysis

therefore addresses the following hypotheses:

39

 

Ha]: Individuals exposed to Travel Michigan advertising will have greater
awareness of the state as a tourism destination, than those who have not been
exposed.

Hazz Travel Michigan advertising will have a positive effect on changing people’s
attitudes towards the state as a tourism destination.

Ha3: Individuals exposed to Travel MI advertising will have greater motivation/
intention to visit the state on a leisure trip, than those who have not been exposed.

Ha4: Travel Michigan’s advertising is effective in generating visits for the state.

After determining the effectiveness of the advertisements on grabing the attention of the
consumer, the next step involved testing whether the advertisements had any effect as
measured by the four variables (DA, IV, AV and Attitudes) described above. Independent
samples T—tests were conducted to test whether there were any significant differences in
the four measures of advertising effectiveness between those aware and those not aware
of advertising. Intention to Visit (IV), Destination Awareness (DA) and Actual Visitation
(AV) were tested first because they were each measured by one variable. Destination
Awareness (DA) was measured nominally (O=unaware; 1=aware). Those considered as
being aware of Michigan as a tourism destination were the respondents who indicated
that Michigan comes to mind when they are thinking of taking a pleasure trip in the
United States and Canada. Intention to Visit (IV) was measured through a 5-point ordinal
scale (I: definitely visit; 5=very unlikely visit) and Actual Visitation (AV) was measured
on a ratio scale by the number of visits an individual has made to the state. Attitude on

the other hand was measured by a number of variables that were meant to capture the

40

different attitudes/perceptions that consumers have towards Michigan’s different tourism
industry sectors. Each variable was measured on a five-point ordinal scale (1=disagree

completely; 5=agree completely).

The results presented in Table 4 show that there are significant differences in DA, IV and

AV between those who were aware of advertising and those who were not.

Table 4: T-tests for equality of means for DA, IV and AV between those aware and those not aware
of advertisements

 

AA N Mean Mean difference Std. Deviation t Sig.

DA Not aware 377 .17 .15 .376 5.332 .000
Aware 727 .32 .466

IV Not aware 365 3.30 .41 1.661 3.909 .000
Aware 712 2.89 1.635

AV Not aware 376 .69 .61 1.512 3.713 .000
Aware 726 1.30 3.002

 

DA: 0 not aware, 1 aware; IV: 1 definitely visit, 5 very unlikely visit; AV number of times one has
visited the state within the past 12 months.

What the results imply is that Travel Michigan’s advertising had a significant effect on
generating awareness of the state as a tourism destination (DA) (t = 5.332; p < .001),
generating motivation to visit the state (IV) (t = 3.909; p < .001) as well generating more

visits to the state (AV) (t = 3.713; p < .001).

Advertising was also found to have a significant effect on changing people’s attitudes
towards Michigan as a tourism destination. All the attitude variables were individually
tested to see if there were any significant differences between those individuals that were
aware of Michigan advertising and those that were not, and significant differences were
found as can be seen in Table 5. The results clearly show that there are significant
differences in attitudes/ perceptions of people between those aware and those not aware

of Travel Michigan’s advertising. People consistently rank Michigan better afier being

41

exposed to the advertisements, than when not. Changing consumer attitudes is an

important step in advertising because advertising is understood to influence behavior

indirectly through its direct effect on customers’ attitudes (Butterfield, Deal and Kubursi,

1998).

Table 5: T-tests results for Attitudes towards Michigan as a tourism destination

 

 

Mean
Diff. Std.
AA N Mean“ A-Un Dev t. Sig.
MI has great restaurants Unaware 299 3.49 .34 1.160 4.694 .000
Aware 618 3.83
M1 has lovely small towns Unaware 302 3.66 .42 1.117 6.038 .000
Aware 652 4.08
MI has great shopping Unaware 275 3.31 .32 1.187 3.998 .000
opportunities
Aware 576 3.63
MI °ffers 3“ excellent mam" value Unaware 209 3.52 .34 1.209 4.031 .000
for the money
Aware 538 3.86
MI has many interesting historic sites Unaware 293 3.41 .41 1.177 5.392 .000
Aware 626 3 .81
MI is a good place for family vacation Unaware 319 3.65 .50 1.177 7.113 .000
Aware 686 4.15
MI offers exciting nightlife and Unaware 261 3.15 .27 1.213 3.108 .002
entertainment
Aware 526 3.41
MI is a good place to meet friendly Unaware 312 3.49 .31 1.151 4.321 .000
people
Aware 658 3.80
MI is great for winter outdoor Unaware 291 3.47 .38 1.332 4.398 .000
recreation activities
Aware 598 3.84
MI offers much scenic appeal Unaware 326 3.69 .44 1.195 6.387 .000
Aware 692 4.13
MI has a lot of high quality lodging Unaware 263 3.28 .44 1.150 5.610 .000
Aware 543 3.72

 

42

Table 5 continued

 

MI is an exciting place to visit

MI is great for summer outdoor recreation
activities

MI is close enough for a weekend
getaway

There is plenty for me to see and do on a
MI getaway tn'p

MI is a great place to relax and unwind

Vacationing in M1 is a great value for my
money

MI is great for enjoying quality time with
family or friends

MI has great beaches

MI is great for short getaway trips

MI is a great place to escape from my
daily routine

MI is a great vacation destination for the
whole family

Unaware

Aware

Unaware

Aware

Unaware

Aware

Unaware

Aware
Unaware

Aware

Unaware

Aware

Unaware

Aware
Unaware
Aware
Unaware

Aware

Unaware

Aware

Unaware

Aware

323
684

324

687

344

709

324

688
327
696

284

615

328

696
264
576
345
708

339

706

330

700

3.30
3.69

3.70

4.19

3.88

4.25

3.49

4.00
3.49
3.96

3.42

3.84

3.57

4.05
3.21
3.70
3.61
4.14

3.23

3.73

3.53

4.10

.39

.49

.37

.51

.47

.41

.49

.49

.53

.50

.58

1.243

1.173

1.212

1.248

1.177

1.144

1.192

1.345

1.253

1.268

1.193

5.107

7.310

5.218

6.902

6.556

5.310

6.767

5.369

7.264

6.340

8.080

.000

.000

.000

.000

.000

.000

.000

.000

.000

.000

.000

 

*Attitude was measured via a 5 point scale with 5 equal to ‘agree completely’ and 1 equal to ‘disagree

completely. ’

The next step in the analysis framework included factor analysis of the attitude variables

in an effort to reduce the data as well as determine if there were any similarities among

the many attitude variables. The Kaiser-Meyer-Olkin (KMO) measure of sampling

adequacy and the Bartlett’s test for sphericity were conducted to test the suitability of the

43

data set for factor analysis. The KMO value was 0.946 greater than the required
minimum Of 0.6 and the Bartlett’s test was significant at the .01 level indicating that the
data set was suitable for factor analysis (refer to Appendix 4). The R-type factor analysis
was utilized since the relationship being studied included variables not cases. The
extraction method used was principal components analysis because the objectives were
data reduction as well as determination of total variance not only common variance.
Varimax orthogonal rotation was used to make interpretation of the results easier as well
as to redistribute the variance from the earlier components in-order to produce a more
meaningful component pattern. Varimax rotation was selected mainly because it is the
best choice when the goal is to reduce the number of the original variables to a smaller
set of uncorrelated variables to be used in subsequent analyses. The factors created from
this stage of the analysis were to be used later for regression analysis, therefore making

Varimax rotation the more suitable choice among orthogonal rotation methods.

The results gave a total of 22 components with different variables loading differently on
each of the 22 components as illustrated in Table 6. The initial extraction and rotated
results varied, with the variation more evenly distributed with rotation than without as

shown in Tables 6 and 7.

44

Table 6: Principal components anal sis results showingloading of each variable on the components

 

 

 

Un-rotated Component Rotated Component Communality
Matrix Matrix
Variable /Components 1 2 3 1 2 3
MI is a great vacation destinanon for .812 ”297 .076 199 .202 .296 .753
the whole fanuly
MI is a great place to escape from my .769 -. 144 .024 $3 .309 .278 .612
daily routine
MI is a great for short getaway trips .681 .115 .492 .373 .224 ,_7_§ .719
MI has great beaches .684 -.257 —.131 £22 .257 .077 .552
MI is great for enjoying quality time .767 -.l79 -.002 £88 .297 .245 .621
with family or friends
Vacatiomng in MI 15 a great value for .777 -.O79 .188 £01 .282 .443 .645
my money
MI is a great place to relax and .835 -.093 .049 fl .369 .340 .708
unwind
There is plenty for me to see and do .817 -.001 .005 £11 .444 .320 .668
on a MI getaway trip
MI is close enough for a weekend .566 .308 .632 .144 .229 L1 .814
getaway
MI ‘5 3m" f0? 3.9mm“ ”mm” .798 -.188 .118 E .251 .359 .686
recreation act1v1t1es
MI is an exciting place to visit .804 .103 -.085 .538 $1 .267 .665
M1 has a lot of high quality lodging .747 .193 -.234 .454 .6_52_ .138 .650
Ml offers much scenic appeal .819 —.211 -.032 fig .316 .230 .716
M1 is great for winter outdoor .592 -.129 -.231 £51 .342 -.009 .420
recreation activities
MI is a good place to meet friendly .746 .078 -.045 E .483 .271 .564
people
MI offers exciting nightlife and .635 .562 -.138 .115 ._§O_2 .284 .738
entertainment
MI is a good place for family .843 -.232 .004 ,7_78_ .298 .264 .764
vacation
MI has many interesting historic sites .775 .084 -.215 .544 J; .135 .654
Ml offers an excellent vacation value .803 .067 .125 ,_5_3_Z .428 .439 .665
for the money
MI has great shopping opportunities .749 .343 -. 190 .350 £31 .221 .714
Ml has lovely small towns .775 -. 103 —.O99 £31 .398 .184 .620
M1 has great restaurants .776 .268 -.202 .422 ._'Z_(_)_§ .200 .715

 

 

 

 

45

The latent root criterion was used to select the components to retain. With the latent root
criterion, only those components with eigenvalues greater than one are retained. Only
three components were retained in this case. The eigenvalues were 12.598, 1.062 and
1.001 for components 1, 2 and 3 respectively. The rest of the components had
eigenvalues of less than one (see Table 7), therefore were not retained for further

analysis.

46

Table 7: Principal Components Analysis results showing eigenvalues for the components and % of

variation explained by each component

 

 

 

Compo Initial Eigenvalues Un-rotated matrix Loadings Rotated matrix Loadings

“em E.value % of Cumu E.value % of Cumu E.value % of Cumu
Variance lative% Variance lative% Variance lative%

1 12.598 57.265 57.265 12.598 57.265 57.265 7.414 33.702 33.702

2 1.062 4.828 62.093 1.062 4.828 62.093 4.501 20.461 54.163

3 1.001 4.552 66.645 1.001 4.552 66.645 2.746 12.482 66.645

4 .765 3.479 70.124

5 .672 3.056 73.180

6 .595 2.707 75.887

7 .532 2.420 78.307

8 .520 2.362 80.669

9 .457 2.077 82.746

10 .434 1.972 84.718

11 .381 1.733 86.451

12 .376 1.710 88.161

13 .354 1.608 89.769

14 .325 1.479 91.249

15 .302 1.373 92.621

16 .288 1.307 93.929

17 .260 1.183 95.112

18 .258 1.171 96.283

19 .238 1.083 97.366

20 .223 1.015 98.381

21 .192 .872 99.253

22 .164 .747 100.000

 

Extraction Method: Principal Component Analysis.

In the initial un-rotated extraction, component I explained about 57% of the total

variation of all the variables, and components 2 and 3 accounted for only 4.8% and 4.5%

47

each respectively. Together the three components accounted for 66.6% of the total

variation as shown in Table 7.

Examination of the individual variables and the components shows that all but one (MI is
great for winter outdoor recreation) variable had a communality value of less than the
critical value of .5. The rest of the variables have more than 50% of their variation being
explained by the 3 components. The results also clearly show that in the un-rotated
solution all the variables load significantly on the first component, and the first
component accounts for 57% of the 66.6 % total variation with the other two components
accounting for the small remainder. The rotated results, however, show an even
distribution of the variance. The total variance explained remains 66.6% but is evenly
distributed. Components 1, 2 and 3 each account for 33.7%, 20.5% and 12.5% of the
variation respectively. The variables also load heavily on only one of the three

components making it a lot easier to interpret the results.

For naming the variables only those variables with factor loadings equal to or greater than
.5 were considered under each factor. In cases where one variable had two or more
components on which it loaded greater than .5, only the greater one was considered. The
third component was the easiest to name as it had only two variables with greater than .5
factor loadings; these were ‘MI is close enough for a weekend getaway,’ and ‘MI is great
for short getaway trips,’ with loadings of .861 and .728 respectively. This component
appears to represent perceptions of people regarding how close Michigan is for short
getaways or weekend trips. This component was named; ‘MI is close enough for short

getaways.’ The second component had six variables significantly loading on it, which

48

appear to revolve around aspects of entertainment. The variable that had the highest
loading on this component was; ‘MI offers exciting nightlife and entertainment,’ with a
factor loading of .802 (refer to Table 6). The component was named; ‘MI offers great
entertainment.’ The first and most important component had sixteen variables with factor
loadings greater than .5; however, two were excluded from this component because they
loaded even higher on the second component (see Table 6). Component 1 appears to
revolve around perceptions of people towards Michigan as a fatnily destination with
family oriented activities and entertainment. This component is the most important of the
three as it accounts for over half of the total 66.6% variation. The results suggest that
people view Michigan as a good tourism and vacation destination for the family. The
variable that loaded most heavily on this component with a .79 factor loading was; ‘MI is
a great vacation destination for the whole family.’ While both components 1 and 2 have
entertainment and activities variables significantly loading on each of them, the kinds of

activities and entertainment differ.

Component 2 represents mainly the kinds of activities for the single adults or couples
without children such as night entertainment, shopping and dining while component 1
represents the activities and entertainment for adults with children such as outdoor
recreational activities. The value variables, ‘MI offers an excellent vacation value for the
money,’ and ‘Vacationing in MI is a great value for my money,’ both loaded on this
component. Price sensitivity is a trait more associated with families with children than
with adults traveling without children. Families with children most often have less

discretionary income, thus look for ways to stretch their dollars to accommodate the

49

needs of the entire family on their limited travel budgets. This component therefore, was

named; ‘MI is a great family destination.’

The next stage of analysis involved correlation analysis to address the following
hypotheses:

Has: Attitudes towards Michigan as a tourism destination will be positively
correlated to Advertising Awareness (AA), Destination Awareness (DA) and
Intention to Visit (IV).

Ha6: Distance of DMA to Michigan will be negatively correlated to AA, DA and
IV.

Ha7: Annual income will be positively correlated to AA, DA and IV.

These hypotheses are based mainly on the persuasion theories such as the H—J-K theory
of communication persuasion which state that the success of communication and
persuasion also depend on several audience factors. The initial position of the audience
affects the extent to which members comprehend the message they are being exposed to.
The initial position includes such factors as personality, socio-economic status,
intelligence and others as depicted earlier in Figure one (Tan, 1985). According to
cognitive dissonance theory and selective processes theory, people attend to incoming
messages selectively, depending on their backgrounds. People with low income or those
who reside further from Michigan might selectively avoid Travel Michigan’s advertising.
These hypotheses, therefore, intend to examine the relationship that these audience
factors have with advertising awareness (AA), destination awareness (DA), and intention

to visit (IV).

50

The non-parametric Spearman’s rho correlation test was used since most of the variables

being tested were either nominal or ordinal. The parametric tests such as the Pearson

correlation are more suited for interval and/or ratio data. The three factors retained from

the previous factor analysis procedure were used at this stage to represent the attitudes

variables.

The results showed that attitudes/perceptions towards Michigan as a tourism destination

are significantly positively correlated with both Advertising Awareness and Intention to

Visit the state as shown in Table 8. Correlation between one of the attitude variables (MI

offers great entertainment) and Destination Awareness (DA), however, was not

significant.

Table 8: Correlation Results between Attitude Factors and AA, DA and IV (Spearman’s rho

 

 

 

correlation)
Attitude factor Advertising Awareness Destination Awareness Intention to Visit
N ariable
Coefficient Sig. (2- Coefficient Sig. (2- Coefficient Sig. (2-

tailed tailed) tailed)
MI is a great family
destination .266“ .000 .204" .000 .225" .000
Ml offers great
entertainment .101 * .039 .096 .050 .210" .000
MI is close enough .l49"“'I .002 .133" .006 .157" .000

for short getaways

 

 

 

 

** Correlation is significant at the 0.011evel (2-tailed); * Correlation is significant at the 0.05 level (2-

tailed).

The results show the attitudes people have towards Michigan as a tourism destination are

positively related to their awareness of advertising, destination and intention to visit the

state. The attitudes towards Michigan as a family destination are more highly correlated

51

with all three variables (i.e., AA, DA and IV), than are the other two attitude factors. This
indicates advertising awareness increases as people’s attitudes towards Michigan as a
family destination improve and vice versa. This implies that for Travel Michigan’s
advertising to get more attention, it should most importantly focus on improving its
image as a family destination. Both Destination Awareness and Intention to travel are
also positively correlated with attitudes towards Michigan as a suitable family tourism
destination. As people’s attitudes about Michigan as a family destination improve, so

does their awareness of Michigan as a destination and intention to visit the state.

Attitude towards Michigan as a close destination for short trips was the next most highly
correlated variable with AA, DA and IV with correlation coefficients of .149, .133, and
.157 respectively. People attend more to and become more aware of advertising as they
view Michigan as a suitable destination for short or weekend getaways. This result is
consistent with other research that has been conducted which reveals that Americans now
take more weekend and short getaways, than they did ten years ago (TIA, 2003). This
may also be attributable to current and most recent world events such as 9/11 that
resulted in a reduction in long haul trips as people would rather stay closer to home and
visit destinations to which they can drive instead of flying. Therefore, destination
marketers such as Travel Michigan need to be positioned to attract those markets closer

to home.

Attitudes towards Michigan as a great destination for adults without children were the

least positively correlated with all three: Advertising Awareness (AA), Destination

52

Awareness (DA) and Intention to Visit (IV). In-fact, the correlation between Destination
Awareness (DA) and ‘MI offers great adult entertainment’ was not significant as shown
in Table 8. The results show the relationship between people’s attitudes towards
Michigan as having great adult entertainment and both Destination Awareness and
Intention to Visit, though positive is not as strong as the relationship between Advertising
Awareness and Destination Awareness with the family and distance attitude factors. The
results imply that in crafting its advertising messages Travel Michigan needs to
incorporate the suitability of the state as a destination for both the family and short
getaways in-order to increase the audience’s attention, destination awareness and

motivation to visit the state.

Some of the relationships hypothesized in Ha, and Ha7 were supported by results from

the analyses conducted, while others were not.

Table 9: Correlation results for Annual Income, DMA location and AA, DA, and IV

 

 

 

Advertising Awareness Destination Awareness Intention to Visit
Spearman Coefficient Sig. (2- Coefficient Sig. (2- Coefficient Sig. (2-
Correlation tailed) tailed) tailed)
Annual .088“ .009 .045 .184 .131" .000
Income
DMA -.056 .064 -.64* .033 -.083** .006
location

 

 

 

 

** Correlation is significant at the 0.011evel; *Correlation is significant at the 0.05 level.

The results presented in Table 9 support the hypothesized relationship that income is
positively correlated with advertising awareness, indicating that as people’s incomes
increase they tend to be more aware of vacation or travel advertising. This agrees with the

theories of cognitive dissonance and selective processes that individuals actively select

53

messages that they have the financial means to pursue and avoid those messages for
products they can’t afford to purchase. Individuals with low income who can’t afford to
take vacations, might consciously avoid travel advertising messages. On the other hand,
those with higher incomes, because they can afford to take vacations and leisure trips
might consciously attend to travel advertising in search of information to use in planning
their trips. Annual income also has a significant positive relationship with Intention to
Visit the state as was hypothesized. This relationship, however, is more obvious than not,
as people’s incomes increase so do their intentions to visit the state on a vacation or
leisure trip. It follows Engel’s law which states that as people’s incomes increase the
proportion they spend on basic commodities such as food and shelter decreases and the

proportion they spent on luxuries increases (Asimakopulos, 1978).

Results also indicate that the relationships between DMA location and both Destination
Awareness and Intention to Visit were as hypothesized. Results presented in Table 9
show that Intention to travel to the state of Michigan for leisure decreases as the distance
from Michigan increases. This again, corroborates the fact that people are more inclined
to take shorter than longer trips. People’s intentions to visit the state of Michigan
decrease as we move from nearby Cleveland to Chicago to Indianapolis. The further
away people reside from Michigan the less likely they are to take pleasure trips to
Michigan. However, no relationship was found between DMA location and awareness of
Travel Michigan’s advertising (see Table 9). Even though the hypothesis has been
disproved, this is, however, a good result for Travel Michigan because it means that its

advertising was consistently and equally received across the three DMAs.

54

4.3.1 Model Estimation

The next stage of analysis involved estimation of two models including AA and DA
through logistic regression analysis. AA and DA were estimated through logistic
regression, as they are both dichotomous variables bound between 0 and 1. This stage of
analysis sought to address the following hypotheses:

Hag: Having taken a pleasure trip to Michigan before (Experience) will have an
effect on advertising awareness.

Hag: Advertising channel has a significant effect on advertising awareness.

4.3.1.1 Advertising Awareness

The following Unaided Advertising Awareness (UAA) model was hypothesized:

UAA = f(X1,....,X,,) where:

UAA = Unaided Advertising Awareness
(X1...Xn) = (Exposure, TV, Radio, Magazine, Attitudes, Experience, Location,
Annual Income, IV, Age and Gender)

The forward greatest reduction in Log Likelihood value (-2LL value) method was used to
enter variables into the logistic regression model. The variables were entered in six steps
with Exposure, TV, Radio, Age, Experience and Attitudes entered at steps 1, 2, 3, 4, 5,
and 6, respectively. The p—value for entering into the model was set at .05. Results are
presented in Table 10 for the variables that met the minimum criterion established to
enter the model. The variables that didn’t meet this criterion were: location, IV, annual

income, magazine and the three attitude factors which included: 1) MI as a great family

55

destination, 2). MI as having great adult entertainment and 3) MI close enough for short

getaways.

Table 10: Variables in the Unaided AdvertisiniAwareness (UAA) Model

 

 

Step Variable B S.E. Wald“ Sig. Exp (B)
Step 1 Exposure .066 .006 127.727 .000 1.068
Constant 5.214 .568 84.115 .000 183.756

Step 2 Exposure .066 .006 124.026 .000 1.068
TV 1.304 .236 30.556 .000 3.682

Constant 4.660 .573 66.094 .000 105.607

Step 3 Age .030 .007 17.336 .000 1.030
Exposure .067 .006 126.528 .000 1.069

TV 1.308 .242 29.317 .000 3.698

Constant 3.348 .646 26.832 .000 28.451

Step 4 Age .030 .007 17.490 .000 1.031
Exposure .067 .006 127.448 .000 1.070
Experience .745 .245 9.267 .002 2.106

TV 1.227 .245 25.083 .000 3.410

Constant 3.096 .654 22.397 .000 22.107

Step 5 Age .029 .007 15.342 .000 1.029
Gender .764 .268 8.1 1 1 .004 2.147

Exposure .068 .006 128.731 .000 1.070
Experience .759 .247 9.419 .002 2.135

TV 1.217 .248 24.174 .000 3.379

Constant 1.974 .761 6.732 .009 7.198

Step 6 Age .031 .008 17.348 .000 1.032
Gender .761 .270 7.925 .005 2.141

Exposure .067 .006 125 .307 .000 1.069
Experience .738 .249 8.764 .003 2.093

Radio .693 .274 6.418 .01 1 1.999

TV 1.041 .258 16.248 .000 2.831

Constant 1.632 .777 4.406 .036 5.113

 

*Assesses the significance of each variable entered in the model
The final model is as follows:

UAA = 5.11 + 2.83TV + 2.14 Gender + 2.09Experience + 2.00Radio + 1.07Exposure + 1.03Age

56

According to all the Goodness of fit tests run including the R2 values, Hosmer and
Lemeshow Test and step Chi-square tests presented in Table 11 the model is a good fit.
The Nagelkerke R2 value for the model is .81, implying that the final model explains an
estimated 81% of the variation in Unaided Advertising Awareness. The Hosmer and
Lemeshow test measures the overall model fit, and the statistic indicates that there was no
statistically significant difference between the observed and predicted outcomes (Hair et.
al., 1998). The higher the p-value the better the model fit since a higher p-value implies
that there are no significant differences between the predicted and observed model. The
Omnibus test shown in Table 11 measures the significance of model coefficients at each
step, and they are all significant for this model. The Wald statistic assesses the
significance of each variable entered in the model, and each one of the six variables
entered is significant at the .01 level except for the constant. All the goodness of fit

measures support the six-variable model stated above.

Table 11: Goodness of fit Tests for the Unaided AdvertisingAwareness (UAA) Model

 

 

 

Model Summary Hosmer and Lemeshow Omnibus Test of
Test“ Model Coefficients
Step -2 Log likelihood Nagelkerke R Square Chi-square Sig. Chi-square Sig._
1 505.289 .770 1.069 .785 873.102S .000
873.102M .000
2 474.543 .787 2.220 .695 30.746S .000
903.848M .000
3 456.516 .797 14.305 .074 18.027S .000
921.876M .000
4 447.257 .802 8.559 .381 9.258S .002
931.134M .000
5 438.669 .806 10.270 .247 8.589s .003
939.722M .000
6 432.363 .810 11.629 .169 6.306s .012
946.028M .000

 

 

 

 

8: Step; M: Model
* Measures overall model fitness. The higher the p-value the better since we fail to reject the hypothesis
that ‘there is no difference between predicted and observed values.’

57

 

The results show that advertising channel has a significant effect on advertising
awareness. If a person saw the travel advertising on TV instead of the other channels
increases the odds of one recalling the advertising. TV ([3 = 2.831; p = .000) is the most
important predictor of unaided awareness of travel advertising. Being exposed to radio
advertising also significantly increases the odds of top-of-mind recall of advertising.
Radio is certainly more important than magazine advertising but not as important as
Television advertising. This result supports hypothesis 9, which states that advertising
channel has an effect on advertising awareness. This result agrees with the findings
presented by Kisielius (1982) who stated that information presented pictorially stimulates
more cognitive elaboration therefore resulting in the development of more storage
locations and pathways in memory, which in turn also increases the likelihood of the

information being retrieved when it is needed for later tasks.

The regression model results also showed that experience has a significant effect on
advertising awareness (B = 2.093; p = .003), implying that having traveled in the state on
a leisure trip prior to being exposed to advertising increases the odds of one’s awareness
of the state’s travel advertising. Those who had experience with leisure travel in the state
of Michigan were more likely to be aware of the state’s travel advertising than those who
didn’t have any experience. This result is also consistent with the theory of Cognitive
dissonance discussed in the preceding chapters which says that people selectively attend
to advertisements of products they have purchased to reinforce their decision and reduce
post-purchase dissonance (Haley, 1985). People who have traveled to the state before

might be more aware of the state’s travel advertising because the advertisements revive

58

the memories of the experience they had during their visit, or convince them that their

choice of a destination was a good one.

Exposure to advertising, as measured by the number of times an individual was exposed
to advertising was also significant ([3 = 1.069; p < .001). The more people were exposed
to travel advertising the higher the odds of them recalling the advertisements. The study
also shows that gender has a significant effect on advertising awareness ([3 = 2.141; p =
.005). Being female increases the probability of top-of-mind awareness of Travel
Michigan advertising. According to Meyers-Levy’s 1989 selectivity hypothesis, adult
males often don’t comprehensively process available information. Instead they tend to
simplify the processing task, and they focus on a single often self-related cue from the
message. Adult females on the other hand, use a comprehensive strategy to process
information. Females tend to assimilate all available cues and engage in more detailed
elaboration of message content than males (Meyers-Levy, 1989). The finding is also
consistent with the household travel decision-making dynamics. Studies have shown that
most of the travel decision-making is done by females. A study by Mottiar and Quinn
(2004) showed vacation discussion is initiated mostly by females (58%) as compared to
males (25%). Females are also most likely to choose the travel agent (34%) than males
(16%) and also make the bookings (54%) compared to 22% for males. However, there
are some decisions that are made jointly such as how much to spend, which

accommodation to use and when to travel.

59

4.3.1.2 Destination Awareness
The following Destination Awareness (DA) model was hypothesized:
DA = f ( X1 ....... X“), where:
DA = Destination Awareness

(X1. ...Xn) = ( Intention to Visit (IV), Exposure, Location, Web-use,
Experience, Advertising Awareness (AA))

Logistic regression analysis that was performed produced a four-variable model with four
of the six variables included. The cut off p-value for inclusion into the model was .05.
The AA and location variables were excluded from the final model because they failed to

meet this criterion. Details are provided in Table 12. The model that resulted was:

DA = .077 + 2.06 Experience + 1.90 Web-use + 1.49 IV + 1.01 Exposure

The entered variables were all significant at the .01 level and each variable’s effect on

reduction of Log Likelihood Value (—2LL) was also significant at the .01 level.

60

Table 12: Variables included in the Destination Awareness (DA) Model at each step.

 

 

Step/Var. B S.E. Wald" Sig. Exp( B) Model Log L Chznlg: in Sig. of Change

1 IV .523 .048 118.641 .000 1.687 -622.114 136.855 .000
Const -2.737 .188 210.881 .000 .065

2 IV .411 .053 59.970 .000 1.509 -573.253 63.212 .000
Expe. .804 .164 23.999 .000 2.235 -553.687 24.080 .000
Const -2.753 .191 207.292 .000 .064

3 Expo. .005 .002 10.162 .001 1.005 -541.647 10.365 .001
IV .407 .054 57.811 .000 1.502 -566.928 60.927 .000
Expe. .796 .165 23.338 .000 2.216 -548.172 23.414 .000
Const -2.549 .200 161.805 .000 .078

4 Expo. .005 .002 9.319 .002 1.005 -S37.625 9.489 .002
Web .642 .240 7.187 .007 1.901 -536.465 7.167 .007
IV .399 .054 55.042 .000 1.490 -561.758 57.755 .000
Expe. .722 .168 18.525 .000 2.058 -542.140 18.519 .000
Const -2.561 .201 162.012 .000 .077

 

S. E: Standard Error; *Assesses the significance of each variable in the model

The R 2 value for the model is .223 and the goodness of fit tests presented in Table 13

show that the model is a good fit.

Table 13: Goodness of fit tests for the Destination Awareness Model

 

 

 

 

 

Step Model Summary Hosmer and Lemeshow Test"
-2 Log Nagelkerke R Chi-square df Sig.
likelihood Square
1 1107.373 .174 6.565 3 .087
2 1083.294 .203 10.696 5 .058
3 1072.929 .215 11.806 8 .160
4 1065.762 .223 13.841 8 .086

 

* Measures overall model fitness. The higher the p-value the better, we fail to reject the
hypothesis that ‘there is no difference between predicted and observed values.’

61

Experience with travel in the state was the most important predictor of destination
awareness (B = 2.058; p = .000). Web use was the next most important predictor of
awareness of Michigan as a tourism destination (8 = 1.901; p = .000). Having visited the
state’s tourism website increased the odds of one being aware of the state as a tourism
destination. Destination awareness was measured by asking respondents to name
destination states or cities that come to mind when they think of taking a pleasure trip in
the United State and Canada. This implies that, Michigan comes to mind when those
people who visit the state’s tourism website are thinking of taking a pleasure trip more
often than those who don’t visit Michigan’s website. This means that the state’s tourism

web site is an important tool in generating awareness of the state as a tourism destination.

Intention to visit (IV) the state, though not as important as web use and experience, also
increases the odds of one’s awareness of the state as a tourism destination ([3 = 1.490; p <
.001). Those intending to visit the state actively search for information, therefore,
increasing their awareness of the state as a tourism destination. Exposure to advertising
also has a significant effect on destination awareness ([3 = 1.001; p = .002). The
coefficient for Exposure is very close to one though meaning that its importance in
increasing the odds of one being aware of the state as a tourism destination is very

minimal as compared to the other three independent variables in the model.

62

4.4 Conversion Rates

4.4.1 Advertising Conversion Model

The last stage of analysis involved calculation of actual conversion ratios to determine
the effectiveness of Travel Michigan’s advertising. Effectiveness of Travel Michigan
advertising was determined following the conversion and tracking models suggested by

Siegel and Ziff-Levine (1990).

An overview of the analysis performed in calculating conversion ratios is presented in
Figure 5. Out of the 908 people that were potentially exposed to Travel MI advertising in
the post campaign phase, 769 were found to be aware of the advertising, an 85%
conversion rate at that stage. To estimate the conversion rate for the proceeding stage,
cross tabs were used to determine the proportion of the people who were aware of
advertising, and who also had destination awareness. The analysis showed that 225 (29%)
of the peOple who were aware of Travel Michigan’s advertising were also aware of the

state as a tourism destination.

63

 

Advertising Travel MI

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1—1 Advertising
Target Audience * 903
Aware of Advertising * 769; 85% CR
Unaided Awareness of 8’
Destination 7 225: 29% CR
Positive Image of ppl p51 PD]
destma‘lon _, 214: 95% CR 149: 67% CR 205: 91% CR
Inquiry
Fulfillment ' 18: 8.4% CR 10: 6.7% CR 16: 7.8% CR
M t' t' ‘
0 rva ‘0" ' 16: 89% CR 9: 90% CR 15: 94% CR
Conver§i0n > 8: 50% CR 4: 44% CR 8: 54% CR
Behavror
Overall Conversion Rate: 2.2% (20/908" 100)

 

 

 

PFY: Positive Family Image; PEI: Positive
Entertainment Image; PDI: Positive Distance
Image; CR; Conversion Rate

Figure 5: The Conversion Model for Travel Michigan’s 2003 Travel Advertising (Adapted from
Siegel and Ziff-Levine, 1990)

To determine the proportion of the target group that hold a positive image of the state as a
tourism destination, three attitude variables were used to represent the three attitude
factors. The three variables used were those that had the highest loading values on each
component in the factor analysis. These attitude variables were:

1. Michigan is a great vacation destination for the whole family

2. Michigan offers great nightlife and entertainment

3. MI is close enough for short getaways
To perform this analysis, the three variables were each re-coded. Ranks 1-2 were recoded

0 representing negative attitude and ranks 3-5 were re-coded 1 representing positive

64

attitude towards MI as a tourism destination. A new variable was also created to represent
the 225 individuals who had both advertising and destination awareness. Out of the 225
respondents who had both advertising and destination awareness, 214 (95%) had a
positive image of the state as a suitable tourism destination for the whole family, 149
(67%) had a positive image of the state as a destination suitable for adults seeking
exciting nightlife and 205 (91%) had a positive image of the state as being a close enough

destination for short get aways as shown in Table 14.

The high conversion rates at this stage showed that, once people are aware of the
destination, then their image or attitudes of the state as a tourism destination improve
substantially. Combining the three image factors gives an average conversion rate of
84%. The challenge, however, is to improve the content of advertisements to highlight
the important attributes of the state’s tourism destinations because of the very low
conversion rate (29%) from the Advertising Awareness stage to the Destination
Awareness stage. Michigan is not high on the list of destinations that people consider
when they are thinking of taking a pleasure trip. While the advertisements reach a lot of
people, they are not as effective in informing people what the state has to offer to

potential tourists.

65

owns: 3:68 .5: 033qu 556: was 0:3 2333.6:— 4

 

 

mNN mON 3 N mmm 3; ON om mNN EN 5 «4 o~m3<
owns: owns: owns: owns: owns: owns:
30,—. airmen gamma Z Became—25... =38. 398m o>cmwoz 3:63.83... .83. o>£mom o>umwoz Became—0:3...

 

 

 

 

 

 

 

 

 

 

owes: 8:890

 

owns: EoEEetoam

 

owes: nouaeumon 388m

 

 

386$ owes: web: 2: firs <Q 6% << 8m 3:58 93-890 ”3 2an

66

The conversion ratios from the Image stage to the Inquiry fulfillment stage, however,
were very low. Only 18 (8.4%) of the 214 people who had a positive family image (PF 1)
of the state inquired, 10 (6.7%) of the 149 individuals who had a positive entertainment
image (PEI) inquired and only 16 (7.8%) of the total 205 people who had a positive
distance image (PDI) inquired. This result supports the claim made by Siegel and Ziff-
Levine (1990) that the inquiry fulfillment stage while it may facilitate conversion is
certainly not a necessary condition for advertising-induced visitation. The results showed
that only 18 out of the 214 individuals who had positive image of the state as a family
fiiendly tourism destination used the toll-free telephone number to inquire about travel
information. The other two Image (Youthful and Distance) factors had conversion rates
of 6.7% and 7.8%, respectively (refer to Table 15). The average conversion rate for this
stage was 7.6%. Telephone inquiries have become less important given that people are
increasingly using the web for travel information. As a result, the ‘telephone inquiry
fulfillment’ requirement of the conversion model makes it less effective in measuring
advertising effectiveness. Travel Michigan has deemphasized the toll free telephone

number in its advertisements.

To determine the conversion rates from the Inquiry-fulfillment to the Motivation stage,
the Intention to Visit (IV) variable was used. The variable was re-coded to have only
two classes representing those who had motivation to travel and those who didn’t. The
original variable was an ordinal level variable coded from 1-5: 1 being “will definitely
visit,” 2: “very likely visit,” 3: “somewhat likely visit,” 4: “somewhat unlikely” and 5

“will very unlikely visit.” The variable was recoded O and 1, 0 representing no

67

motivation to travel and 1 representing motivation to travel. All those who gave
responses fiom 1 to 3 were considered as having some motivation to visit and those
with responses from 4 to 5 were considered as not having any intention/ motivation to

visit the state.

68

22626 35 2 2539 9 @338 no .305. ..cEP .35 85 @8865 9.3 «EEECE 9 waged...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

owes:

ms mom mom 3 m $5.0 o3 m2 3 v o\ov.m 3m 3— E m 957.9”
3 I. n. G n a O I. H. G n a 3 I. n... G n a
O 0 a l. S n O O a l. S n O O 9 I. S n
M n W. w a n w. w ..s. n w. m
m. a .l. m. m. x ..l w. m. «as ..l w.
u 1 H. W. u 1 no w; ..u 1 n. W.
ml... % D. W a 0. ml... % P
omeH 3534.5 owns: EoEEetoEm owns: nouaccmon Egan

 

 

 

owflm “coca—52%???“ 2 uwflm owns: Boa mowed 855280 ”2 033.

69

The results of this stage show very high conversion rates from the Inquiry stage to the

Motivation stage. Out of the 18 individuals who passed through all the preceding stages,

16 generated motivation to visit the state, an 89% conversion rate from the preceding

stage. The other two groups, those who had a positive image of the state being an exciting

destination and those who had a positive distance image, each had 94% and 90%

conversion rates respectively as shown in Table 16. The average conversion rate for this

stage from the preceding Inquiry-Fulfillment stage was 91%. What this result means is

that people who use the toll free telephone number to inquire and/ or ask for the state

travel information have great motivation to take a pleasure trip to the state of Michigan.

Table 16: Conversion rates from the Inquiry to the Motivation stage

 

 

 

AA, DA, positive AA, DA, positive AA, DA, positive
family image and Youthful image and Distance image and
inquired inquired inquired
No motivation to visit 2 1 1
Motivation to visit 16 9 15
Total 18 10 16
Conversion rate 89% 90% 94%

 

 

 

 

The results also show that the distance factor generates greater motivation to visit the
state than any of the other two image factors. However, even though the family image
factor has a lower conversion rate from the inquiry to the motivation stages, it still drives

more people through the stages than any of the other image factors.

The last stage on the conversion model involves determining the conversion rate from the

Motivation to the Actual Visitation (AV) stage. The original AV variable was measured

70

at a ratio level representing the number of times a respondent had visited the state. This
variable was re-coded to enable cross-tab analysis. The new variable was a nominal level
one with 0 representing no travel and 1 representing those that had traveled in the state on
a pleasure trip. Also, not all the people who took a pleasure trip to the state were
influenced by the advertising, therefore, only those individuals who indicated that their
travel was influenced by the advertisements were included in the calculation of the final

‘conversion behavior’ stage.

The results show that exactly 20 (50%) of the total 40 people who filtered through all the
preceding stages traveled to the state and that their travel was influenced by their
exposure to the state’s travel advertising. However, the results presented in Table 17 also
show that a total of 102 (11.2%) out of the 908 phase 2 respondents took a pleasure trip
in the state of Michigan and also indicated that advertising primarily influenced their trip.
These individuals didn’t, however, all pass through the conversion process proposed by
the Conversion Model. This, therefore, implies that by making inquiry a necessary

condition, the Conversion Model underestimates the effects of advertising.

71

0:055:00 53m: 05 :0 0:0 :80— :a 8 58:8: 0: 5850: :0 .308. :.:05. was 020:8: 08:3 53535:: 085 0: 80:0: .6...
.8508: 05 0: 5:0:8: 0: 500.50: :0 0:3 :0: 0:03 :05 :05 5985:: 0:3 53555:: 0005 0: £0.23. 2..
owns 5050502. 95000:: 05 50:: 0.8:: 8:88:00 0

 

 

 

 

0:0:

$0: 4.8% .52: .80: 4.53. $4.2 .50: 4.5% $4.2 8:82:60

so

we: 2 mm: m :8 m 8: 0: :8 0: E: 0: : :

$1.585:

.3 58:26::

8: w 8 a. S: 4 0: N: 8: w a m 0.6::

3:05:05:

.3 58:25::

: E m an m : :m m we... a. :m m 8:. R :8 025

0 mmmfi DU

8: N 8: : 3: N NE a 3: m 0mm 0 :0 a. : a
as: 2 2: 38:58:80 :88. 2 2: 4:08:38: :86: E Z: ...BEaeSD
5:02:02 5:: :050>.:0_2 8:03:02

.0365 .085: 00:80.5 03:60: .<Q .<<
5:025:00 wasps—0.: 05 :0 =0 36 :88
:0 :25 :88 :.:05 0:3 2:00: :0 85:52

 

5:: .0555 .088: ::0E::8:0::m 0358:
.3 .<< .8888 0:523: 2:: :6 =a :6
80:: :0 AEQV :02: :.:05 055 :0 85:32

 

5:: >563 .088: 3:80: 05:00:

.<: .5. .8388 2:: :6 EW :20 :88
:0 AEQV 80:: :.:05 055 2:00: :0 5.58:2

 

 

0W8m .:0:>0:0m 8:88:00. 05 :0: 00:0: 888500 ”2 030:.

72

By combining all the individuals who took a pleasure trip to the state of Michigan whose
trip was influenced by the state’s advertising gives an overall conversion rate of 2.2% ((8
+ 4 + 8/ 908)* 100). The overall conversion rate including those individuals who were not

captured by the conversion model comes to 11.2%.

4.4.2 The Advertising Tracking Model

For two reasons, the tracking model conversion rates are higher and a lot more realistic
than those calculated using the conversion model. First, the tracking model does not
consider inquiry as a necessary step to conversion, and second the tracking model
assumes that one can develop a positive image of a destination just by being exposed to
advertising which can produce motivation to visit without necessarily going through the
Destination Awareness stage. There are three routes (1-3) through which individuals can
end up exhibiting conversion behavior through the tracking model as compared to the
conversion model. An additional route (0) was added to capture those individuals who
didn’t go through either the conversion model or tracking model stages, but indicated that

their decision to travel was influenced by the advertising. This is illustrated in figure 6.

73

 

 

 

Unaided

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 Awareness of
Destination \ ' Motivation
\ To visit
. . . . 2 "
. ' Advertlsmg Posmve Image / 3
Advertismg —1> Awareness —I> of Destination / 1 & 2
Inquiry

Fulfillment .
~- 3 _- Conversmn

Route 0 — Behavior

 

 

 

 

 

The dotted line to and from the inquiry fulfillment
stage means that the stage is not required for conversion to occur.

Siegel and Ziff-Levine (1990)

 

Figure 6: Advertising Tracking Model showing the different routes (0-3) used to calculate the
conversion rates and ROI.

Travel Michigan’s advertising effectiveness was also calculated using the tracking model.
Only routes 1 and 2 above were used to calculate the conversion rates at this stage. Recall

that only the third route requiring inquiry fulfillment was used in the conversion model.

The results from the tracking model illustrated in Figure 7 show that route 1 has an
overall conversion rate of 5.2%. This route, though not as conservative as the conversion
model, also requires that one goes through the Unaided Destination Awareness stage
before developing motivation to visit. Route 2, however, gave the highest conversion
rate simply because the route excludes both the Destination Awareness and Inquiry
Fulfillment stages. The route presumes that one can develop motivation to travel and
eventually visit the state only after having developed a positive image of the destination

from exposure to advertisements. For this route, the three image variables were used and

74

 

three different conversion percentages were obtained. The analysis resulted in a different

tracking model for the state’s travel advertising as shown in Figure 7 below.

 

 

 

Advertising
audience
908

/

 

 

 

 

 

Advertising
Awareness
769: 85% CR

 

 

414: 61% CR

V

 

Destination
Awareness
225: 29% CR

 

 

 

 

I78: 79% CR

 

 

 

PFI ’
683: 89%CR

 

\ 279: 63% CR

‘
V

 

 

Route 0
Conversion
behavior

 

102: 11.2% CR

 

 

CR; Conversion Ratio; OCR Overall Conversion Ratio
Route 1: Includes ‘Destination Awareness’ stage
Route 2: Excludes ‘Destination Awareness’ stage

PFI: Positive Family Image
PEI: Positive Entertainment Image
PDI: Positive Distance Image

PEI
440: 57% CR

 

 

\.

 

 

 

 

 

 

40 . 61% CR

Motivation
To visit

 

 

 

 

 

 

 

 

 

 

 

 

 

Route 1
PDI Conversion
behavior
. o
671. 237/. CR 47. 26% CR
k 5.2% OCR
v ‘
PFI Conversion PEI Conversion PDI Conversion
Behavior Behavior Behavior
86: 21%CR 61: 15%CR 85:21%CR
9.5% OCR 6.7% OCR 9.4;/o OCR
Route 2
Average OCR
8.5%

 

 

Figure 7: Advertising Tracking Model for Travel M1(Siege| and Ziff-Levine, 1990)

The results show the conversion rates for Travel Michigan advertising range from 2.2%

using the conservative conversion model to 8.2% through route 2 of the tracking model

and 11.2% through the straightforward route (0). While these results show that the

percentage of people who were brought to the state as a result of the state’s travel

advertising, they do not say anything about the return on investment (ROI) of the

75

 

 

 

advertising campaign since neither accounts for sales in dollars or advertising costs in

this analysis at this juncture. Determination of the ROI is the subject of the next stage of

the analysis.

4.5

Return on Advertising Investment (ROI)

Return on advertising investment (ROI) was calculated for all the four routes illustrated

earlier in Figures 5 and 6. ROI in this study refers only to government tax revenues

captured from tourist activity. Calculating these figures was made possible by making use

of several assumptions given below:

It was assumed that the visitors from the three DMAs spend the same amount of
money once they cross the MI borders into the state. The researcher was
interested in the visitors’ expenditures once they are in the state not before. Data
on visitor expenditures and travel party size were obtained from the Household
Surveys conducted by Michigan TTRRC.

It was assumed that the advertising target population is the same as the
population of the greater metropolitan areas of the three DMAs since 97% of all
US households have cable and access to television, and an even bigger
percentage have access to some form of media, be it radio magazine or
newspaper (N eilsen Media Research, 2002). Only that proportion of the
population over the age of 18 was used since it was the targeted population
group.

Population data were based on the 2000 census.

76

o The state gets approximately 8% of revenue through sales tax, lodging room tax
and income taxes from tourism businesses and their employees. According to the
research department of The Travel Industry Association of America (TIA) (2004)
the state collects 5 cents out of every traveler spending dollar in taxes. Carr and
Holecek (2003) estimated the state collects 10% of tourist expenditure in taxes.
This study used 8%, an average of these two figures (5% and 10%), which
translates to 6% from sales and room taxes and 2% from tourism businesses and
their employees. The 8% figure was assumed to be an appropriate estimate of
state tax collections in the absence of any better data.

The population figures for the greater metropolitan Chicago, Indianapolis and Cleveland
are 9,650,137; 1,607,486 and 2,900,000 respectively. Since only those individuals above
the age of 18 were considered only 7,044,600 (73%) of the Chicago population,
1,194,362 (74.3%) of the Indianapolis population and 2,073,500 (71.5%) of the
Cleveland population belonged to the defined age category. The sum of the populations
of the three DMAs was used for the calculation of the ROI. The calculations and results

of the ROI analysis are presented in Table 18.

77

Table 18: ROI figures for the 2003 Travel Michigan Advertising Campaign (4 routes used)

 

 

 

Item Route 0 Route 1 Route 2 Route 3
POPUIaIiOH 10,312,462 10,312,462 10,312,462 10,312,462
Travel Party size 3.5 3.5 3.5 3.5
Ave. Expenditures/Party US$947 US$947 US$947 US$947

Ave. Expend/individual

 

(US$947/3.5) US$271 US$271 US$271 US$271
Conversion Ratio (%) 11.2 5.2 8.5 2.2
Number of people
convened (cwpopulafion) 1,154,996 536,248 876,559 226,874
Revenue USS (90868“ 313,003,847 145,323,215 237,547,562 61,482,898
people*Expend/md1v1dual)

0
State Revenue USS (8 /" °f 25,040,308 11,625,857 19,003,805 4,918,632
Revenue)
Tm"? MI advfmsmg ”“5 2,392,062 2,392,062 2,392,062 2,392,062
for th1s campaign (US$)
ROI 1: 10.47 1: 4.86 1: 7.94 1: 2.06

 

Route 0: Straightforward route; Route 1: Tracking model route including ‘Destination Awareness’ stage;
Route 2: Tracking model route excluding ‘Destination Awareness’ stage; Route 3: Conversion model route
including ‘Inquiry Fulfillment’ stage. ’

These results show a positive return on advertising investment even for the conservative
conversion model (route 3) which includes the inquiry fulfillment requirement. Through
this route, Travel Michigan more than breaks even getting about US$2.06 for every dollar
spent on advertising. Route 2 is the one that assumes that individuals can develop
motivation to visit and visit after being exposed to the advert without necessarily having
gone through the destination and inquiry fiilfillment stages. This route gives an ROI value
of 7.94 implying that there is a $7.94 return for every dollar spent on advertising. The
ROI values range from 1: 2.06 for the most conservative route 3 to 1: 10.47 for the most
liberal one (route 0). Either way, the state is getting more from advertising than it is
spending, meaning that travel advertising is a worthwhile investment. It is also important
to note that these analyses only include direct expenditures and net revenues, factoring in
the multiplier effect or considering gross revenues would certainly give even higher
conversion and ROI ratios. Such further analyses, however, are beyond the scope of this

study. These results tend to justify spending of tax dollars on travel advertising and the

78

results are consistent with what other researchers have found. Mok (1991) found gross
tourism advertising returns to range from 1:10 to 1:56 and concluded that there is

considerable justification for using public resources in tourism advertising.

79

Chapter 5

5.0 Summary and Conclusion

The main object of this study was to determine the effectiveness of Travel Michigan’s
travel advertising campaign for the 2003 season in three DMAs: Chicago, Indianapolis
and Cleveland. The study utilized different analyses in an effort to establish the
effectiveness of the travel advertising campaign. First, comparisons were made between
the pre and post campaign phases to note if there were any differences in advertising
awareness between those interviewed before and after the campaign. The significant
difference in advertising awareness found between these two groups of respondents was a
good starting point for the study. It meant that the campaign had made that initial
breakthrough of getting the attention of the audience that is required for advertising to be
effective. Further analyses to determine the effectiveness of the campaign would not have
been necessary had there been no differences in advertising awareness between
respondents interviewed pre and post campaign. Getting attention and creating awareness
are the first and certainly important stages in most of the communication models
including both the old and newer models of communication. Today’s consumers are
exposed to not less than 1000 advertisements daily (Kotler, 1997), therefore, breaking
through such clutter to gain the consumer’s attention is a challenge for any advertiser.
Just getting the consumer’s attention is, however, not sufficient by itself. It is hoped that
this stage will result in some more positive responses including, but not limited to:
opinion change, attitude change, perception change, affect change and/or action change

(Tan 1985).

80

Second, having established that the campaign was effective in getting the intended
audience’s attention, the analyses that followed were intended to determine if the
campaign had any further effects beyond just getting attention. Effectiveness in this study
was measured in several ways including: generating destination awareness, creating
positive attitude toward the state as a tourism destination, generating motivation to visit
the state and generating visits to the state. These measures were used based on the
assumption that success in communication is shown in many ways as illustrated by the H-
J -K Communication Persuasion Model introduced in Figure 1. It is especially important
in tourism because it is a unique industry which relies much on word of mouth
(Opperman, 2000); therefore, by generating destination awareness, positive perceptions
and attitude towards a destination, an advert can be said to be effective since such people
are likely to recommend the destination to others even if they have not visited it
themselves. Furthermore, having positive attitudes and product awareness have been
shown to be predecessors of action (Lavidge and Steiner, 1961; Haskins, 1964; Fishbein
and Azjen, 1974; Tan, 1985; and Butterfield et. al., 1998). Those who are aware of
advertising have greater levels of destination awareness, greater intention to visit the
state, better attitudes towards the state as a tourism destination and are more likely to take
a pleasure trip to the state than those who are not aware of advertising. Significant
differences were found in visits to the state among those who were aware of advertising

and those who were not.

81

While proponents of experiments and quasi-experiments might argue that this does not
establish a causal relationship between advertising and results, it is important to note that
the results drawn from the analyses reported herein were all highly and consistently
significant at the p < .001 level. Also, Starch (1961) argues that when divided into those
that recall and those who do not recall an advert, the differences between these two
groups can be attributed to advertising. He also argues that typically those who recall
advertisements buy more than those who do not. This argument, however, is not meant
to undermine the importance of experiments in establishing true causal relationships,
rather it is meant to offer support for the findings of this study. Those individuals who
were aware of Michigan’s travel advertising campaign consistently ranked Michigan
higher as a good tourism destination than those who were not aware of the advertising.
Similarly, those who were aware of the state’s advertising had greater
intention/motivation to visit the state and actually visited the state more than those who
were not aware. Even though this study does not meet the requirements of an experiment,
the consistency of the results can only imply that some credit belongs to the Travel

Michigan advertising campaign.

Next, principal components analysis was conducted to reduce the attitude data as well as
examine the grouping pattern of the different attitude variables. Three components
resulted which represented three attitudes towards Michigan as a tourism destination. The
first and most important attitude component was the Family Image factor which accounts
for 34% of the variance in the rotated matrix. Next most important was the

Entertainment Image factor representing 20% of the variance, followed by the Distance

82

Image factor which represented 12% of the variance. This result means that Michigan is
viewed by consumers most importantly as a family destination and also as a close enough
destination for short get away trips. This information is of particular importance to the
state’s destination marketers because it gives them a good idea about what the visitors are
looking for in the destination, and therefore, helps them to craft their advertising
messages to communicate those messages as well as develop destination attributes to

meet the customers’ needs.

Correlation analysis showed a positive relationship between attitudes of people towards
the state as a tourism destination and advertising awareness, destination awareness and
intention to visit the state. These results also show Michigan is most importantly viewed
as a family tourism destination. Attitudes towards Michigan as a family destination were
found to be highly and positively correlated to advertising awareness, destination
awareness and intention to visit the state. As people’s ‘family destination’ attitudes
towards Michigan improve, so does their intention to visit the state. Also, as people’s
distance attitudes towards the state improve, so do their advertising awareness and
intention to visit. Also, the relationship between DMA location and both destination
awareness and intention to visit was negative. This means that people’s awareness of
Michigan as a tourism destination as well as their intention to visit the state both decrease
as the DMA’s distance from Michigan increases. Again, this highlights that the state of
Michigan needs to also focus its attention on markets closer to home as people’s
preferences for short holidays increase. Recent world events such as 911 and the war

have left people more inclined to take short trips to which they can drive rather than fly

83

and also which they feel they are secure. Travel Michigan, therefore, needs to take

advantage of the situation and get short haul travelers to its destinations.

The fourth stage of analysis involved use of logistic regression to determine the factors
affecting unaided advertising awareness (UAA) and destination awareness (DA). The
results showed exposure to television advertising was the single most important predictor
of unaided advertising awareness, followed in descending order by gender, experience,
radio, level of exposure and age. Being female increased the probability of one recalling
the advertisements, implying that females were more likely to recall the advertisements
than their male counterparts. This is an important finding for destination marketers as it
tells them that in designing their advertisements they need to include material that
appeals to females since they are more likely to recall the advertisements when it comes
to making decisions on which destinations to visit. Results from the previous stages of
analysis show Michigan is a state that is ranked very highly as a family destination, so in
its effort to attract two adult (male, female) families to the state Travel Michigan should
mostly consider including advertisements that appeal to females. Women play a dominant

role in the vacation destination selection process in the family (Mottiar and Quinn, 2004).

Having traveled to the state on a pleasure trip before was also a significant factor
affecting advertising awareness. This finding collaborates Ehrenberg’s (2000) claims that
advertising’s main role is to reinforce feelings of satisfaction for brands already
purchased and that consumers mostly ignore advertising for brands they are not already

using. This also agrees with the theory of cognitive dissonance which states that

84

individuals consciously avoid or attend to messages in an effort to reduce dissonance. By
attending to advertisements of brands and/or products they have already purchased,
consumers reinforce their decision and convince themselves that their decision was the
right one, thereby reducing post-purchase dissonance. Having been exposed to radio
advertisements, even though not as important as TV, also increased the probability of an
individual recalling the advertisements. Also, the more an individual was exposed to

advertising the more likely they were to recall the advertisements.

Experience was the single most important factor affecting destination awareness in the
DA model, followed by web use, intention to visit and finally exposure to advertisements
in that order. Having traveled in the state on a pleasure trip before increased the
probability of an individual being aware of the state as a tourism destination. This is an
important result for the state DMO because it also suggests that repeat visitors are
important to the state’s tourism industry. DA was measured by asking the respondents to
name states and/or provinces that come to mind when they are thinking of taking a
pleasure trip in the U. S. and Canada. Individuals who have experience traveling in the
state on pleasure trips keep considering it for their future vacations, illustrates the
importance of repeat business to the state’s tourism industry. Michigan leisure and
tourism providers, therefore, need to pay attention to their repeat visitors to the point of
developing relationships and loyalty because repeat visitors are an especially important

target market.

85

Web use was also an important factor affecting destination awareness. Those individuals
who had visited the state’s official travel and tourism web site were more aware of the
state as a tourism destination than those who had not. This implies the state’s travel and
tourism web site is an important resource in generating awareness of the state as a
tourism destination among potential visitors. Intention to visit the state was also
significant, maybe because those with the intention to visit the state on a pleasure trip
actively seek information resulting in them being more aware than those with no

intention to take a pleasure trip to the state.

Exposure to advertisements was also a significant factor affecting destination awareness.
The more individuals were exposed to advertisements the more they were aware of the
state as a tourism destination. Assuming destination awareness was the goal of the
advertising campaign, this result show there are positive returns to advertising. This
study, however, does not show at what point this ceases to be true (i.e. when does

increasing exposure stop increasing destination awareness?)

The fifth stage of analysis included calculations of real conversion ratios using four
different approaches involving named routes (0-3) developed in this study. The first route
was from the tracking model which includes the destination awareness stage. The second
route was also from the tracking model, but excluded the destination awareness stage.
The third route was from the conversion model which included the inquiry fulfillment
stage, and the fourth route was the ‘straightforward’ route called route 0. Under the

straightforward route (route 0) all respondents who indicated that they had visited the

86

state and that their visit was influenced by advertising were considered. These individuals
needed not to have gone through all the other steps as depicted by the tracking and
conversion models. The results ranged from a conversion ratio of 2.2% from the most
conservative conversion model to 8.5% for the liberal tracking model route that excludes
the destination awareness stage and finally 11.2% for the straightforward route. The
straightforward route conversion rate was arrived at by simply considering those
individuals who had traveled to the state on a pleasure trip and indicated that advertising
influenced their decision to travel without necessarily having gone through the stages
given by both the tracking and conversion models. The important result from the sixth
and final ROI stage of analysis is that there is a positive return on advertising investment
taking all the routes into consideration. The straightforward route yielded a liberal
$10.47 return in state tax revenue collections on every dollar spent on advertising, route 1
from the tracking model including the destination awareness stage yielded a $4.86 return
on every dollar spent on advertising, route 2 yielded a $7.94 ROI and finally route 3 from
the conservative conversion model yielded a $2.06 ROI. It is important to note these
figures are actually underestimates as they include only estimated tax revenue collections

associated with projected direct expenditures.

In summary, the findings of this study reveal the positive effect of travel advertising on a
number of important factors. The main weakness of the study that is most likely to be
raised by the proponents of experiments is that, it fails to establish and ascertain causal
relationships. However, the consistency of results given in the different stages of analysis

conducted in this study stand to support that the advertising campaign had substantial

87

positive impacts. Results clearly showed that there were significant differences in
destination awareness, intention to visit and actual visitation between those who were
aware and those not aware of advertising. There were also significant differences among
those aware and those not aware of advertising on the attitudes towards the state as a
tourism destination. Those aware of advertising consistently ranked Michigan more
highly as a desirable tourism destination than those not aware of advertising. Considering
attitude change and destination awareness as prerequisites to visitation, Travel
Michigan’s advertising can be considered highly effective. The DA logistic regression
model also shows that exposure to advertising is important in increasing one’s chances of
being aware of Michigan as a tourism destination. Finally, the conversion ratios and the
ROI figures all show the effectiveness of the advertising campaign. So, even though the
study is not an experiment, the consistency of the results clearly illustrate the importance

of the travel advertising campaign.

Woodside (1990) advocates the use of experiments such as A-B or A-B-C splits in which
two groups get a different treatment each and the other becomes a control group; and then
compare the results from these two or three groups. As much as these are capable of
producing the causal relationships among different variables, they are easier to
contemplate than conduct, especially in our field of travel and tourism that relies heavily
on word of mouth. For instance, how possible is it to run different television travel
promotional programs targeting different groups in the same population without the
possibility of treatment spillover from one treatment group to the next and even to the

control (no treatment) group? This might be easy in a laboratory setting, but it is not so

88

simple to implement in real life situations. Quasi-experiments on the other hand, that
have been advocated for and used by a number of researchers (Mok, 1990; Woodside et.
al., 1997), though they are more practical than true experiments, have their own
shortcomings. Quasi-experiments do not utilize random assignment of subjects between
the treatment and control groups implying that the results can’t entirely be accurately
attributed to advertising alone. The resulting differences between the groups might result
from initial differences before the treatment was administered. Given these shortcomings
associated with the supposedly better methods, the important question remains: Is it
really worth it to invest in more expensive and laborious methods when there is no
guarantee that the outcome is any better than those obtained from cheaper and more
practical methods? The answer partly lies in Messmer’s and Johnson’s (1993) conclusion
that more testing of both conversion and alternative approaches is needed before

conversion studies can justifiably be abandoned.

Future research, therefore, should focus more on comparing the results from the different
methods including conversion studies, experiments and quasi-experiments to see if there
are any great differences that warrant the complete abandonment of one method for
another. Also, more work needs to be done in refining experiments and quasi-
experiments to make them more practical and applicable to the travel and tourism field.
This is important because for most DMOs, the problem is not about which is the best
method to utilize, but is about whether or not there are enough funds to conduct any
accountability research at all, especially considering the tight budget circumstances under

which most of them operate. Investing in more expensive impractical methods with no

89

guaranteed clean results is not an option for them when there are other cheaper more
practical alternatives. Unless other cheaper and more practical approaches than the
current experiments are developed and tested, conversion studies remain that best viable

alternative available.

5.1 Study Limitations

The main limitation of the study is the short lag time between the campaign and the post
campaign phase. Pre-campaign data collection ended on the 1St of May 2003 and post-
campaign data collection started on July 7. There was only two months in between the
two phases during which the campaign was run. The period might not have been long
enough for exposed individuals to respond to the advertisements. In the future, such
studies should allow more time for full campaign effects to be recorded. The campaign
effects delivered in this study, therefore, might be under-estimates of actual impacts.
Underestimates are also probable since neither the possibility of the advertisements
generating positive word of mouth or repeat visits over multiple years were included in

the projections.

Another limitation of this study is that it did not filter out the effects of other independent
campaigns that might have been running simultaneously with the state’s campaign. Even
though caution was taken to ensure that no other major campaigns were running in the
three DMAS during that time, other Michigan tourism providers were not prohibited from
running advertisements to promote their businesses, thereby creating an element of

‘noise’ for this particular study. There were no major advertising events taking place

90

during the time for this study in the three DMAs. Future studies of this sort should,
therefore, account for and filter out the impact of other advertising. Future studies should

also account for cumulative effect of advertising exposure.

The use of secondary data was also a limitation to this study. There are things that could
have been done differently, had it been a primary research study. Placement of
advertisement campaigns in the media would have been delayed to allow more pre-
campaign interviews to be completed. Also, the timing of post campaign interviews
would have been different. Consumers would have been allowed more reaction time,

between exposure to advertisements and post campaign interviews.

91

References

Aaker, D. A. and Day, G. S. 1974. A Dynamic Model of Relationships Among
Advertising, Consumer Awareness, Attitudes and Behavior. J oumal of Applied
Psychology, 59: 281-286.

Asimakopulos, A._1978. An Introduction to Economic Theory: Microeconomics.
Oxford. Oxford University Press.

Ballman, G., J. Burke, U. Blank, and D. Korte. 1984. Toward Higher Quality Conversion
Studies: Refining the Numbers Game. Journal of Travel research, 22 (4): 28-33.

Barry, K. and J. O’Hagan. 1972. An Econometric Study of British Tourist Expenditure in
Ireland. Economic and Social Review, 3(2): 143-161.

Bauer. 1964. The obstinate audience: The influence process from the point of view of
social communication. American Psychologist, 19 (5): 319-328.

Bayou, Mahomed, E., Panitz and Eric. 1993. Definition and Content of Persuasion in
Accounting. Journal of Applied Business Research, 9 (3): 44-51.

Buterfield, D. W., K. R. Deal and A. A. Kubursi. 1998. Measuring the Returns to
Tourism Advertising. Journal of Travel Resea_rgh, 37 (August): 12-20.

Burke, J. F. and R. Gitelson. 1990. Conversion Studies: Assumptions, Applications,
Accuracy and Abuse. Journal of Travel Reseaih, 28 (3), 46-51.

Caples, J. 1974. Tested Advertising Methods. Engelwood Cliffs, NJ. Prentice-Hall.

Chamberlin, E. H. 1962. The Theory of Mongolistic Competition. Cambridge. Harvard
University Press.

Comanor, W., and T. Wilson. 1974. Advertising and Market Power. Cambridge, MA.
Harvard University Press.

Ehrenberg, A. S. C. 2000. Repetitive Advertising and the Consumer. Journal of
AdvertisingResearch, (November): 39-48.

Edell, J. A. and A. Staelin. 1983. The Information Processing of Pictures in Print
Advertisments. Journirl of Consumer Research. 10 (June): 45-61.

F estinger, L. 1957. A Theory of Cognitive Dissonance. Stanford, CA. Stanford
University Press.

 

Fishbein, M. and Azjen, I. 1975. Belief, Attitude, Intention and Behavior. Reading, MA.
Addison-Wesley.

92

Garfield, B. June 27 1994. There’s Catch to States’ ‘Come Hither’ Approach. Advertising
Aga, 65 (27): p. 32.

Gruber, A. 1970. Purchase Intent and Purchase Probability. Journal of Advertising
Research 10 (1), 23-27.

 

Haley, R. I. 1985. Developing Effective Communication Strategy: A Benefit
Segmentaaion Approach New York. John Wiley and Sons.

Haskins, J. B. 1964. Factual Recall as a Measure of Advertising Effectiveness. Journal of
Advertising Research, 4 (March): 2-8.

Feldman, J. M. and J. G. Lynch. 1998. Self-Generated Validity and Other Effects of
Measurement on Belief, Attitude, Intention, and Behavior. Journal of Applied
Psychology, 73 (3): 421-35.

Hovland, C. I., I. L. Janis and H. H. Kelly. 1953. Cited in: Tan, A. S. 1985. Mass
Communication Theories and Research. New York. Wiley.

Howard, J. A. and J. N. Sheth. 1969. The Theory of Buyer Behavior. NY. John Wiley
and Sons.

Keller, K. L. 1987. Memory Factors in Advertising: The Effect of Advertising Retrieval
Cues on Brand Evaluations. Journal of Consumer Research, 14 (December), 316-
33.

Keller, K. L. 1996. Brand Equity and Integrated Communication. In: Thorson, E. and J.
Moore (eds) Integrated Communication: Synergy of Persuasive Voices. Mahwah,

NJ. Lawrence Erlbaum Associates, Inc.

Kisielis, J. and B. Stemthal. 1986. Examining the Vividness Controversy: An
Availability-Valence Intepretation. Journal of Consumer Research 12: 48-431.

Kisielis. 1982. Detecting and Explaining Vividness Effects in Attitudinal Judgement.
Unpublished Doctoral Dissertation. Northwestern University. Evanston, IL.

Kotler, P. 1997. Marketing Management: Analysis, Flaming, Implementation, and
Control. Englewood Cliffs, NJ: Prentice Hall.

Krugrnan, H. E. 1972. Why Three Exposures May Be Enough. J ournal_of Advertising
Research, 12 (6): 11-14.

Lavidge, R. C. and Steiner, G. A. 1961. A Model for Predictive Measurements of
Advertising Effectiveness. Journal of Marketing, 25(October): 59-62.

93

Messmer, J. D. and Johnson, R. R. 1993. Inquiry Conversion and Travel Advertising
Effectiveness. Journal of Travel Research, 31(Spring): 14-21

Meyers-Levy, J. 1989. Gender differences in information processing: A selectivity
interpretation. In: P. Cafferata and A. Tybout (eds). Cognitive and affective
responses to advertising. Lexington, MA. Lexington.

Meyers-Levy, J. and Malaviya P. 1999. Consumers’ Processing of Persuasive
Advertisements: An Integrative Framework of Persuasion Theories. Journal of
Marketing, 64 (special issue), 45-60.

Mottiar, Z and Quinn, D. 2004. Couple Dynamics in Household Tourism Decision
Making: Women as the Gatekeepers? Journal of Vacation Marketing, 10 (2): 149-
160.

Nedungadi, P. 1990. Recall and Consumer Consideration sets: Influencing Choice
Without Altering Brand Evaluations. Journal of Consumer Reseach, 17 (4): 263-
76.

Opperman, M. 2000. Tourism Destination Loyalty. Journal of Travel Research 39 (1):
78-84.

 

Palda, K. 1964. The Measurement of Cumulative Advertising Effects. Engelwood Cliffs,
NJ. Prentice Hall.

Palda, S. K. 1966. The Hypothesis of a Hierarchy of Effects: A Partial Evaluation.
Journal of MarketingResearcl'i, 3(February): 13-24.

Peles, Y. 1971. Economics of Scale in Advertising Beer and Cigarettes. Journal of
Business 44 (January): 32-37.

 

Petty, R. E. and J. T. Cacioppo 1986. Communication and Persuasion: Central and
Perjrherafl Routes to Attitude Change. New York. Springer.

Rao, A. G. 1970. Quantitative Theories in Advertising. New York. Wiley

Rapp, S., and T. Collins. 1987. Maximarketing. New York. McGraw-Hill.

Ray, M. L. 1973. Marketing Communications and the Hierarchy of Effects: In Clarke, P.
(ed.). New Models for Mass Communication Research, vol. 11. Sage Annual
Review of Communication Research. Beverly Hills. Sage Publications.

Raymond, C. 1974. The Art of Using Science in Marfiketing. New York. Harper and Row.

Rotzol, K. B. The Starch and Ted Bates Correlative Measures of Advertising

94

 

Effectiveness. Journal of Advertising Research, (March): 22-24.

Sandage, C. H. and V. Fryburger. 1963. Advertising'l’heory and Practice. Homewood,
IL. Richard D. Irwin.

Severin, W. J ., and J. W. Tankard Jr. 1979. Communication Theories: Origins,
Methods, Uses. New York. Hastings House.

 

Siegel, W., and W. Ziff-Levine. 1990. Evaluating Tourism Advertising Campaigns:
Conversion vs. Advertising Tracking Studies. Journal of Travel Research, 28 (3):
51-55.

Simon, J. L. and J. Amdt. 1980. The Shape of the Advertising Response Function.
Journal of AdvertisingResearch, 20 (4): 11-28.

Starch, D. 1961. M_e_asuring Product Sales Made by Advertising. Mamaroneck, N. Y.
Starch and Staff.

Strack, R. 1992. The Different Routes to Social Judgments: Experiential Versus
Informational Strategies. In Martin, L. L., and Abraham Tesser, (eds). m
Construction of Social Judgments. Hillsdale, NJ. Lawrence Erlbaum Associates.

Schmalensee, R. 1972. The Economics of Advertising. Amsterdam, Netherlands. North
Holland.

Tan, A. S. 1985. Mass Communication Theories and Research. New York. Wiley.

Tourism New Zealand, 2005. Understanding the Japanese Long Haul Market.
Unpublished Manuscript.

Travel Industry Association of America (TIA), 2003. 2002-2003 Survey of US. Stag and
Territory Tourism Office Budgets.

Trochim, W. M. K. 2001. The Research Methods Knowledge Basa. 2nd Edition. Atomic
Dog Publishing.

Uysal, M. and J. Crompton. 1984. Determinants of Demand for International Tourist
Flows to Turkey. Tourism Management, 5 (4): 288-97.

Wansik, B. and M. L. Ray. 2000. Estimating an Advertising’s Impact on One’s
Consumption of a Brand. Journal of AdvertisLng Research, (November): 106-13.

Weilbacher, W. M. 2003. How Advertising Works. Journal of Advertising Research,
43(2): 230-234.

Woodside, A. G. and D. M. Reid. 1974. Tourism Profiles Versus Audience Profiles: Are

95

 

Upscale Magazines Really Upscale? Journal of Travel Research. 12 (Spring):17-
23.

Woodside, A. G. and I. A. Ronkainen. 1989. How Serious is Non-response Bias in
Advertising Conversion Research, J oumal of Travel Research, 26 (Spring): 34-37.

Woodside, A. G. 1990. Measuring Advertising Effectiveness In Destination Marketing
Strategies. Journal of gavel Research, 29 (Fall): 3-8.

Woodside, A. G., R. J. Trappey III, and R. MacDonald. 1997. Measuring Linkage-
Advertising Effects on Customer Behavior and Net Revenue: Using Quasi-
Experiments of Advertising Treatments with Novice and Experienced Product-

Service Users. Revue Canadienne des Sciences de I’Administration, 14(2): 214-
228.

Wright, P. 1980. Message-Evoked Thoughts: Persuasion Research Using Thought
Verbalization. Journal of Consumer Research, 7 (September), 151-75.

Zhou, Z. 2000. The Impact of Memory on Expenditures Recall in Tourism Conversion
Studies. Journal of Travel Research. 38 (February): 304-308.

96

Appendix 1: Sample Questionnaire

Date: Survey Number: 03-CL—2-
Time: Interviewer:

 

 

 

 

Travel Michigan Market Survey

CLEVELAND DMA Awareness Study
(Do not read the list)

1. When you are thinking about taking a pleasure trip in the UNITED STATES or
CANADA what cities, states or provinces come to mind? Any others?

 

First mention Others

[
[

 

California (Los Angeles, San Francisco, San Diego)

 

Florida (Orlando, Tampa, Jacksonville, Miami)

 

Hawaii

 

Illinois (Chicago)

 

Indiana (Indianapolis)

 

 

 

 

 

 

 

 

 

 

 

l ]
l l
l ]
l l
] l
] ] Michigan (Detroit, Traverse City, Mackinac)
I l l
l ]
l I
l l
l ]
l ]

 

 

 

1
Minnesota (Minneapolis/ St. Paul, Bloomington [Mall of America])

[ New York
[ Ohio (Cleveland, Columbus, Cincinnati, Sandusky [Cedar Point])
[ [ Ontario (Toronto)
[ [ Wisconsin (Milwaukee, Green Bay, The Dells, Madison)
[ [ Other: , 3

(Do not read the list)

2. Now thinking just about a pleasure trip in the MIDWEST, what cities or states
come to mind? Any others?
(Do not read the list. Record answers in the first column below)

 

First mention Others

[I
[l
I]
]
l
]

 

Illinois (Chicago)

 

Indiana (Indianapolis)

 

Michigan (Detroit, Traverse City, Mackinac) f

 

Minnesota (Minneapolis/St. Paul, Bloomington [Mall of America])

 

Ohio (Cleveland, Columbus, Cincinnati, Sandusky [Cedar Point])

 

Wisconsin (Milwaukee, Green Bay, The Dells, Madison)

 

Other: , .
97

[ ]

 

 

 

 

 

 

 

3. For which states or provinces do you recall seeing or hearing any vacation or travel
advertising in the past 6 months?

(Read each state NOT checked in the first column. Record responses in the columns on
the right column)

4. Do you recall seeing or hearing any vacation or travel advertising for in the past 6
months?

 

 

 

 

 

 

 

 

 

 

 

 

Do not read the states Read the remaining states
Illinois [ ] Yes No
Indiana [ ] Yes No
Michigan 8 [ 1 Yes N6
Minnesota [ ] Yes No
Ohio [ ] Yes No
Ontario [ ] Yes No
Wisconsin [ ] Yes No
Other (specify) [ ] Do not ask “other”
None [ ]

 

 

 

 

If a respondent DID NOT MENTION any MICHIGAN advertising in question 3 or 4,
GO TO QUESTION 10.

5. You mentioned that you saw or heard vacation or travel advertising for the State
of Michigan in the past 6 months. Please describe that advertising to me. What did it say
or show? [ ] —(-99) (Do not read) Don’t know

 

6. In your opinion, what ideas or messages was Michigan trying to communicate in
this advertising? [ ] —(-99) (Do
not read) Don’t know

 

98

 

7. Approximately, how many times have you seen or heard Michigan travel
advertising in the past 6 months?

# times [ ] —(-99) (Do not read)
Don’t know

 

[If“zero” go to question 10. ]

 

(Read the list)
Did the information you saw or heard on Michigan have

.°°

[ ] —1 a primary influence on your decision to travel to Michigan

] ——2 a partial influence on your decision
] -3 no influence on your decision

] —4 (Do not read) Did not visit Michigan

] -(-99) (Do not read) Don’t know

Hf—1f—17—1

(Read the list)
Compared with other states and provinces would you say you have seen
] —1 Much more advertising for Michigan
] —2 Somewhat more advertising for Michigan
] —3 About the same
] —4 Somewhat less advertising for Michigan
] —5 Much less advertising for Michigan?
] —(-99) (Do not read) Don’t know

f—‘Il—If—1HI—1f—W \o
C

10. What do you think is the Internet address for the state of Michigan’s official

 

 

tourism web site? [ ] —(-99) Don’t know
11. Do you have access to the Internet?
[ ] —1 Yes

 

[ ] —2 No => [Go to question 17.]

 

12. Have you used the Internet for travel planning purposes in the past 12 months?

 

 

[ ] —1 Yes

I l -2 No => [Go to question 16.]

13. Have you made a travel-related purchase over the Internet in the past 12 months?
[l—lYes []—2N0

99

14. In the past 12 months, how often, if at all, have you visited the state of Michigan’s
official tourism web site to obtain Michigan travel information?

# times [ ] —(-99) Don’t know
=> I Go to question 16.]
[If “zero” => Go to question 16.]

 

 

 

(Do not read the list, unless respondent is unable to give a specific answer).
How did you access it?

[ ] -1 Typed in the web site address

[ ] —2 Used a search engine (search program) to find it
[ ] —3 Used the bookmark from previous use(s)

[ ] —4 Clicked on the link on another web site

[ ] —(-99) Don’t know

(Read the list)
16. In the next 12 months, how likely is it that you will visit the state of Michigan’s
official tourism web site to obtain Michigan travel information?

[ ]—1 Very likely

[ ] —2 Somewhat likely

[ ] —3 Somewhat unlikely

[ ] —4 Very unlikely

[ ] —(-99) (Do not read) Don’t know

17. Have you used a State of Michigan toll-free phone number to inquire about travel
information?

[ ]—1 Yes
[ ]—2 No
[ ]—(-99) Don’t know

ADVERTISING BLOCK

Now I’d like to ask you some specific questions about vacation or travel advertising for
the State of Michigan that you MAY or MAY NOT have seen or heard in the past 6
months. For each description I read, please tell me whether or not you have seen or heard

that advertising.

(Read ALL the descriptions). TV Advertising

100

 

18. Have you seen advertising on TELEVISION for the State of Michigan in

the past 6 months that... .?

 

showed a girl talking about her vacation trip to Detroit and attractions she visited

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

there. The ad ended with the “Great Lakes. Great Times.” logo, and gave the Don’t
Michiganorg web address for more information. Yes No Know
(Read ALL the descriptions). Magazine Advertising
19.
Have you seen similar advertising for the State of Michigan in the past 6 months in a Don’t
MAGAZINE? Yes No Know
(Read ALL the descriptions). Radio Advertising
2 1 .
Have you heard a similar ad on RADIO? Don’t
Yes No Know
Have you heard RADIO advertising for the State of Michigan in the past 6 months
that... .?
... had a boy describing his vacation in Frankenmuth swimming in an indoor pool,
visiting a Christmas land and having his tooth come out. The ad ended with the “Great Don’t
Lakes. Great Times.” logo, and gave the Michigan web site for more information. Yes No Know
had a boy talking about his fun family vacation in Michigan, swimming, fishing,
going on a glass-bottom boat. The ad ended with the “Great Lakes. Great Times.” Don’t
_logo, and gave the Michigflweb site for more information. Yes No Know
had a child during a show-and-tell presentation for his classmates showing
memorabilia that his parents brought back from Michigan. The ad ended with the
“Great Lakes. Great Times.” logo, and gave the Michigan web site for more Don’t
information. Yes No Know

 

 

 

 

 

22. Direct Mail Advertising.
23. Internet Advertising.
(Read the list)

101

 

 

 

‘61”

24. Using a 5-point scale where means you “disagree completely”, and “5” means

you “agree completely”, please tell me how well, you think, each statement describes

MICHIGAN as a vacation destination.

Michigan is a great place to escape from my daily routine
Michigan is great for short getaway trips

Michigan has great beaches

Michigan is great for enjoying quality-time with family or friends

Michigan is a great place to relax and unwind
There is plenty for me to see and do on a Michigan getaway trip
Is close enough for a weekend getaway

Is great for summer outdoor recreation activities
Is an exciting place, to visit

Has a lot of high quality lodging

Offers much scenic appeal

Is great for winter outdoor recreation activities
15 a good place to meet friendly people

Offers exciting nightlife and entertainment

Is a great place for a family vacation

Has many interesting historic sites

Offers an excellent vacation value for the money
Has great shopping opportunities

Has lovely small towns

Has great restaurants

25. In the past 12 months, how many times, if any, have you traveled in Michigan on a
pleasure trip?
# times 0 — None

 

26. Do you have specific plans to take a pleasure trip in Michigan in the next 12 months?

 

[ ] —1 Yes => ]Go to question 28.]
[ ] —2 No

 

(Read the list)
27. How likely is it that you will take a pleasure trip in Michigan during the next 12
months?

[ ]—1 Very likely

[ ]-—2 Somewhat likely

[ ]—3 Somewhat unlikely

[ ] —4 Very unlikely

[ ] —(-99) (Do not read) Don’t know

102

To conclude, we'd like to ask just a few questions to help us classify your answers.

28. What is your age? [ ] —(-55) Refused

29. (DO NOT ask) Gender: [ ] -—1 Male [ ] —2
Female

(Do not read the list)

30. What is the highest level of school that you have completed?

] —-1 Elementary or less

] —2 Some high school

] —3 High school graduate

] —4 Some college

] —5 College graduate

] —6 Graduate / professional

] —7 Vocational or technical school
] —(-55) Refused

] —(-99) Don’t know

f—1f—1Hf_|f—‘f—If—If—!f—1

31. Including yourself, how many people live in your household? #
32. How many children under the age of 18 live in your household? #
(Do not read the list. Enter exactly what the respondent says.)

33. What is your current occupation?

 

 

1=Professional/technical
2=Managerial

3=Sales

4=Clerical
5=Construction
6=Craftsman
7=Operative
8=Transportation perative
9=Labor

10=Personal service

1 1=Retired
12=Unemployed
-55=Refused

~99=DK

103

34. What is your zip code?
[ l-(-55) Refused [ ]—(-99) Don’t know

 

35. What racial or ethnic group do you belong to?
i I 4'55) Refiised [ ] —(-99) Don’t know

36. The median household income is $42,000. Would you say your total
household income before taxes in 2001 was above or below the median?

[ ] —1 Above the median

[ ] —2 Below or at the median => ]Go to question 38.]
I

[

 

 

] —(-55) Refused => ] Go to question 38.]
] —(-99) Don’t know => I Go to question 38]

 

 

37. Was your total household income above $75,000?

[ l

[ ]

[ ]—(-55) Refused

[ ]—(-99) Don’t know

38. Thank you for your time and have a great day / evening.

104

 

 

 

00:88:08: 0:08 :0: 0:0 :03 00:50:: 0:: 02:»

05 .006. .0808. .020 .30.: 86:0: 2:: 0:8 06000 0.. 2:. .08.: 2:: =6 .5: :05: :6

 

 

 

 

 

 

 

 

 

0::N0w02 0w0020 500500: 0 80:0 ”80:0: 50:30:: 5:07.: 0005.002 0: 0.005 8:0: 0:: :0 0.2:» 0.5: 5030:: ....0
80:08:08: 0:08 :0: 0:0 :03 83:03.:
83:08:55 05 80» 5:0 6:0: ...882. 080 .83 00:0: 0:: 5:3 505:0 50 0:.:. .800585: 5:0 8:00.:
0:502 owaoEU 5:0 0:050:03 0:0: 8565: 055.88: :0::D 0.00:5:05: 8 :05000> 0:: 80:0 wax—0: >0: 0 50: ....m
80:08:88 0:08 :0: 0:5: :03 88:02): 0:: 0>0w 5:0
83:08:55 .0w0_ ...88:. :00:0 .000—0.: :80: 0:: 5:3 505:0 50 0:2. 0:55.: 5:0 85:05: 5:0 .80: 0 08: 85:80.:
0:502 OWNOEU .05: 5:00 05 :305 888:: .885 :00m w:::00_m 05:0 8888 80:0 $5050: :0: 0 50: .. :5
80:08:08: 0:08 :0: 0:0 :03 :0u::0=>:
83:08:55 0:: 02:» 5:0 .0u0_ 8.882. :080 .000—0.— 80:08 0:: 5:3 505:0 50 0:.—. 8000: 00:30:): 0 :0
0::N0w02 .>:. 0w0QEU 850:: :0: 5:3 5:0: 0:: 0: "8.0.: 5:0 0580 5:00 0 ”85:0: 80:0 85:0: Eu 0 5030:: ....m
80:08:88 0:08 :0: 0:0 :03 88:22 05 80: 5:0 .0w0_ .2885. :00:0 80:0: :00:0: 05
0:50: 008:5 5:3 0608 0: 60.: .80: 60886 0:8 .600 :5: .83 2:: 8 00%: a 32:6 08:8 :0 ... 00:. .. .0
00:08:08: 0:08 :0: 08.550 :03 3 05 80:
83:08:55 5:0 6:0: ...88::. :00:0 80:5 :00:0.. 05 5:3 505:0 50 0::. 00:05: 0:: 555 5:00 0:: 8 was: 5:0
QEwawz .>.H ommoEU 8:000: 5:0 8:0— .8805:w= 8088:): 8 :0::000> 3:80: :0: 0:: 80:0 85:0: :0: 0 5030:: .—
.EE
an; wflUfl—OwEQrfia on: aflosommuho>fia
:0:—:3 :w:0::: .0::0:0 0:: :0 <29 0058885 ::080m:::0>5<

 

 

<55 5:0 .000—0:0 : 3008005850 0:: :0 :0:: :88: ”N 0:500 <

 

105

 

 

6003088: 08:08.: 58.0 >:00:.:. 088896;: .80: 880:0": .> 5:0 .:: .0 .0m05:0m

 

80:88:08: 0:08 :0: 0::: :03 00:80:):

0:: 08» 5:0 .80: 0.885. 80:0 .884 80:03 0:: :::3 505:0 50 0::. .80: 80::0:-::0_u 0

 

 

 

 

 

 

850M: 5:0—95:0 :0 :80: .883: 888853: 80080:: 8 00:800.» 3:80: 8:: ::: 80:0 005:0: m0: 0 50: ...N:

80:88:08: 0:08 :0: 0::: :03 008:0:2 0:: 030: 5:0 .000: 0.882.

80:0 :0:—0‘: 80:03 0:: :::3 505:0 50 0::. .80 0800 :80: :8 «0:8: 5:0 5:0: :08:::::0

350M: 58:0>0:U 0 M8383 .80.: :0058 :0 0: ”888:3: 50880088.: 8 00:800.» :8 :0::8085 :0: 0 50: ...:

80:88:08:

588.820 ”0800902 .>.:. 5:0:0>0:U 0:08 :0: 0::: :03 00:80:): 0:: 08: 5:0 0:0: :.:08:.:. 80:0 :0:—0.: 80:03 0:: :::3 505:0 50
80:08:05 6:00: 2:808:08 2.: .22.: 02:... 2.: 2.2.252. .2: 0.2.0: 3 ...: 8.2.2; :2. 2.2.0 05...: ...: a ..2. ......
5:0—26:0 80:88:08: 0:08 :0: 0::: :03 00:80:: 0:: 08» 5:0 0:0: ...:08:.:.

83:08:55 80:0 80:01—80:03 0:: :::3 505:0 50 0: .H 80:80:): 80:: 0:00: 8000:: 800:0.— ::: 8::

850M ow0oEU 0::::0:0808 0830:: 888880 ::: :0: 8:880:08 :0:-5:0..30:: 0 ”8:05 5:80 0 50: ....0

80:88:08: 0:08 :0: 0:::

:03 00:80:: 0:: 08» 5:0 0:2 0.883. 80:0 :0:—0‘: 80:0: 0:: :::3 505:0 50 0::. .885

0::N0w02 ow00£U :00: «8.005 80:0 800:: 0 5008808 5:0 :000: 0:: :0 8:30: 8 9:0: 03: 5030:: ....0
00:88:08: 0:08

:0: 0::: :03 88:00): 0:: 0.30» 5:0 0:0: :.:08:.:. 80:0 80:0.— 80:0.. 0:: :::3 505:0 50

250% ow00£U 0: :. .:0:::: :8 :::3 880: «85:: 5:0 ::0:0: 00:80:): 0 :0 088:: 80:0 ”80:0: :0: 0 50: .55

 

 

 

 

50.8:800 N 85:08:.

106

Appendix 3: Descriptive Statistics by DMA

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Chicago Cleveland Indianapolis Total
Number % Number % Number %

Age 45 - 47 - 47 - 46
Gender: Female 222 32.2 229 33.2 238 34.5 689

Male 134 31.5 146 34.3 146 34.3 426
Education
Elementary 2 4O 2 40 1 20 5
Some High School 6 42.9 2 14.3 6 42.9 14
High School Graduate 66 24.7 99 37.1 102 38.2 267
Some college 77 32 83 34.4 81 33.6 241
College Graduate 127 36.2 111 31.6 113 32.2 351
Graduate/Professional 69 34.7 64 3 2.2 66 33 .2 199
Technical school 4 18.2 10 45.5 8 36.4 22
Number of people in H. Hold
1 57 32.9 57 32.9 59 34.1 173
2 118 29.8 136 34.4 141 35.7 395
3 57 28.7 69 34.8 72 36.4 198
4 69 36.9 61 32.6 57 30.5 187
5 40 37.7 33 31.1 33 31.1 106
6 4 16 10 4O 11 44 25
7 2 22.2 3 33.3 4 44.4 9
8 l 50 0 O 1 50 2
9 0 0 l 50 l 50 2

 

 

 

 

 

 

 

 

107

 

 

Appendix 3 contd.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 l 100 0 0 0 0 1
Number of children under 18
0 229 33.3 225 32.8 233 33.9 687
1 53 30.5 58 33.3 63 36.2 174
2 48 31.6 55 36.2 49 32.2 152
3 18 28.1 21 32.8 25 39.1 64
4 3 23.1 6 46.2 4 30.8 13
5 0 0 0 O 4 100 4
6 0 0 1 50 1 50 2
Annual H. Hold income
</= median ($42K) 62 23.5 224 38.2 101 38.3 264
> $45K but < $75K 172 32.3 101 33.2 184 34.5 533
> $75K 124 38.8 325 30.3 99 30.9 320
Racial/ethnic group
Asian/Pacific Islander 11 78.6 1 7.1 2 14.3 14
Black/A. American 23 33.8 23 33.8 22 32.4 68
Hispanic 15 83.3 3 16.7 0 0 18
White/Caucasian 278 29.8 321 34.4 333 35.7 932
Multi-racial 3 33.3 3 33.3 3 33.3 9
Other 5 25 6 30 9 45 20

 

 

 

 

 

 

 

 

108

 

 

Appendix 4: Test for Appropriateness of data set for factor analysis

KMO and Bartlett's Test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy.
.968

 

 

 

 

Bartlett's Test of Sphericity Approx. Chi-Square 7020.503
df 23 1
Sig. .000

 

 

109

CHIGA STAYE UNIV SJTV LIBRAR S
1

111111111 1

 
      

llWill/Hill!Hill/(111111111711!

3 02845 4068