 

 

 

 

 

.40.,“ .u... L

 

 

 

 

THESIS

2004-

- 1'1?
tr 5 N"? ‘ U

 

 

 

LIBRARY
Michigar. State This is to certify that the
U nive rsity dissertation entitled

 

THE ROLE OF MOTORSPORTS INVOLVEMENT ON
MOTORSPORTS MEDIA USAGE: AN ANALYSIS OF
INDUSTRY-DEVELOPED MEASURES

presented by
Molly Catherine Ziske

has been accepted towards fulfillment
of the requirements for the

Ph. D. degree in Mass Media

Major Professor's /S#t ature

Afar/03

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 Cz/CIRC/DaleDue.p65-p.15

 

 

THE ROLE OF MOTORSPORTS INVOLVEMENT 0N MOTORSPORTS MEDIA

USAGE: AN ANALYSIS OF INDUSTRY-DEVELOPED MEASURES
By

Molly Catherine Ziske

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Advertising

2003

ABSTRACT

THE ROLE OF MOTORSPORTS INVOLVEMENT ON MOTORSPORTS MEDIA

USAGE: AN ANALYSIS OF INDUSTRY-DEVELOPED MEASURES
By

Molly Catherine Ziske

The first area of inquiry of this study was to assess the reliability and validity of
an involvement scale developed for industry research by an advertising agency. The
second area of inquiry (using data from the same industry study) was to assess the impact
of involvement and demographics on media usage. The data used was from a study of
NASCAR fans and is NASCAR-speciﬁc in its scope.

Results indicated that the original 30-item scale, used to measure involvement, in
fact measured a variety of concepts. However, six items from this original scale did
measure the concept of involvement as put forth by the literature. Other concepts
measured by the scale (as indicated by factor analyses) included technology transfer and
social involvement. Since several variables measured involvement, social involvement,
and technology transfer, new (composite) variables were created in an effort to have one
variable for each of the concepts. These new variables were used in regression analyses
to assess the effect of each concept on media usage. Involvement with racing had a small
and positive inﬂuence on media usage. Social involvement had a mixed impact on media
usage with a slight and negative impact on radio and cable television usage in particular.

Technology transfer did not have a signiﬁcant impact on respondents’ media usage.

Demographic variables included gender, income, age, employment status, and
occupation. The impact of these variables on media usage was assessed using Analysis of
Variance. Age appears to be the only demographic variable that inﬂuenced radio
listenership, with the highest radio listenership among 35- to 44-year-Olds. Age also had
an inﬂuence in that as age increases, respondents are less likely to look to cable television
and specialty magazines and newspapers for motorsports information. Men tend to be
more likely to use most of the media, including newspapers, cable television, and
specialty magazines and newspapers, as well as the Internet.

The inﬂuence of income on motorsports media usage was interesting as well. In
particular, respondents who fell into the higher income categories (i.e., more than
$60,000 annually) had the highest average media usage for all motorsports media.
Interestingly, in all of the media categories for which employment status was a signiﬁcant
differentiator, those respondents who fell into the category of not employed had the
highest media usage. For example, the average Internet usage rate for those unemployed
was 29 Internet “hits” over the past 5 months—considerably higher than students, who
had the lowest average rate of .13 hits over the past 5 months.

The present study points to the need for better understanding of involvement and
what it means to both consumers and marketers, as well as a continuing understanding of

how consumers use media and what inﬂuences their media choices.

Cepyright by
MOLLY CATHERINE ZISKE
2003

DEDICATION

To Emerson, Carter, and Claire. Thanks for helping me get to 100 pages.

For Matt. He’d have been proud.

ACKNOWLEDGMENTS

Completing a dissertation requires, more than anything else, a great support
system. I have been fortunate to have such a support system, composed of all the people
named here (and, I’m sure, some I’ve forgotten over the years). First and foremost I am
eternally grateful to my dissertation chairman and committee members. In particular, Dr.
Bruce Vanden Bergh motivated me in so many ways. It is difﬁcult to ﬁnd words to
describe how Dr. Vanden Bergh was the perfect chairperson, giving guidance when
needed, backing off when needed, and motivating me along the way. It would not be an
overstatement to say that without his guidance, I would not have ﬁnished.

Just as crucial was the support of Dr. Bonnie Reece, who gave me statistical
advice and guidance (as well as a ream of paper) in an effort to help me. Without
knowing it, she has reminded me of why I got into this in the ﬁrst place. Dr. Keith Adler
and Sue Carter have both provided great insight into things I had not considered
including in my dissertation and—more importantly—believed I could ﬁnish.

Aside from my committee, there are many others who have made completing my
dissertation much easier. Brian and Katy Wagner gave many Monday nights so that I
could work on the many iterations of the draft—they are truly a gift from God. Many of
my colleagues at Campbell-Ewald Advertising supported me: Ruth Dunn, Chris Schim,
Judy Peitsch, and David Hudson among them. Each helped by telling me “You can do
this!” when I felt I could not or by allowing me time off to complete my ﬁnal draft. More

importantly, most of them shared my joy over ﬁnishing. For making it at all possible, I

vi

must thank Tara Seilkop and Terry Dolan for allowing me to use THE DATA! Dave
Pagnucco and Dayton Brown also deserve thanks for their help with data and statistics.

The wonderful women at MSU saw me through my entire program—Nancy
Steck, Gloria Baird, and Connie Williams were there through it all. Drs. Erika Butler and
Karan Wyre both believed in me and shared their many keys for ﬁnishing.

Elizabeth PuleO-Tague and Karen Van Antwerp believed I could ﬁnish and were,
quite possibly, the happiest of all when I did. They are fantastic friends.

So many people tied to academe were key in my ﬁnishing as well and deserve
thanks. Dr. Rich Semenik always wants what is best for me and it shows. Dr. Tom
O’Guinn always provided direction, even when I called after neglecting to for so long.
Drs. Terry Pavia and Janeen Costa deserve thanks for their years of friendship and
support as well. Finally, Dr. Al Muniz helped by listening and never telling me just how
daunting a dissertation can really be.

My family, of course, was also a source of inspiration when I had none. My sister,
Lisa, and mom, Theresa, encouraged me—sometimes with guilt! My sister-in-law,
Cheryl, shared in my excitement about ﬁnishing, which means so much to me.

Finally, I must thank the four people who have provided me the most
encouragement and motivation of all. My three wonderful daughters, Emerson, Carter,
and Claire, motivated me to ﬁnish what I started—a good lesson for them. They also
helped by going to bed quietly and letting Mom work on her dissertation.

In addition to my children, I have been blessed with the friendship of Jim
Harrington, who helped me through the dissertation process more than he could ever

imagine. Jim was there for all of the real work and motivated me consistently with his

Vii

encouraging words and unfailing support. He brought me back to the reality that I could
ﬁnish when I was convinced I could not - there are not enough words to thank him for

that.

viii

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................ xii
CHAPTER 1 - Introduction ................................................................................................ 1
Background ............................................................................................................. 1
Purpose of the Present Study ................................................................................... 6
CHAPTER 2 - Literature Review ....................................................................................... 9
Involvement ............................................................................................................. 9
Involvement in the Communication Literature ......................................... 15

Involvement in Recreation Literature ....................................................... 22

Specialization Theory ................................................................................ 26

Involvement Summary and Implications .............................................................. 32
Changing Media Usage ......................................................................................... 33

Sports Growth in the United States ....................................................................... 34
NASCAR ............................................................................................................... 35
Research Questions ............................................................................................... 39
CHAPTER 3 - Methodology ............................................................................................ 41
Background ........................................................................................................... 41
Sample ................................................................................................................... 42
Original Analysis .................................................................................................. 43
Independent Variables ............................................................................... 44

Dependent Variables ................................................................................. 46

Current Study Variables and Methodology ........................................................... 46

ix

Reliability Analysis ................................................................................... 46

Factor Analysis .......................................................................................... 47
Regression Analysis .................................................................................. 47
Analysis of Variance (ANOVA) ............................................................... 49
CHAPTER 4 — Results ...................................................................................................... 50
Reliability Analysis ............................................................................................... 51
Validity Analysis ................................................................................................... 57
NEWTECH and NEWSOC Reliability and Validity ................................ 66
Regression Analyses ............................................................................................. 68
Motorsports Newspaper Article Usage ..................................................... 7O
Motorsports Cable Usage .......................................................................... 74
Motorsports Magazine and Newspaper Usage .......................................... 77
Motorsports Television Usage .................................................................. 78
Motorsports Internet Usage ....................................................................... 8O
Motorsports Radio Usage .......................................................................... 82
Demographic Dependent Variables ...................................................................... 84
Occupation and Media Usage ................................................................... 85
Gender and Media Usage ...................................... » .................................... 8 6
Income and Media Usage .......................................................................... 88
Employment Status and Media Usage ...................................................... 89

Age and Media Usage ............................................................................... 90
CHAPTER 5 - Discussion and Conclusions .................................................................... 92
Overview of Findings and Implications ................................................................ 92

Did the “Involvement Scale” Measure Involvement? ............................... 93

Do Involvement, Social Involvement, Technology Transfer, and

Demographics Affect Media Usage? ........................................................ 95

What Next? Study Implications ............................................................................ 99

Study Limitations ................................................................................................ 102
APPENDIX - Survey Booklet: MOTORSPORTS INTEREST STUDY ....................... 103
REFERENCES ................................................................................................................ 117

LIST OF TABLES

TABLE 2.1: Zaichowsky’s Personal Involvement Inventory ........................................... 18
TABLE 2.2: Overview Of NASCAR Divisions ................................................................ 36
TABLE 2.3: NASCAR Fan Characteristics ...................................................................... 38
TABLE 4.1: Media Usage (Dependent) Variables ........................................................... 51
TABLE 4.2: Original Involvement Variables ................................................................... 52
TABLE 4.3: Reliability Analysis ...................................................................................... 54
TABLE 4.4: Exploratory Factor Analysis Results ............................................................ 59
TABLE 4.5: Exploratory Factor Analysis (partial results) ............................................... 60
TABLE 4.6: Final Variables Used to Measure Involvement ............................................ 64
TABLE 4.7: Reliability Analysis of Final Involvement Items ......................................... 64
TABLE 4.8: NEWTECH Factor Analysis ........................................................................ 67
TABLE 4.9: NEWSOC Factor Analysis ........................................................................... 68
TABLE 4.10: Summary Regression Output: Newspaper Article Readership .................. 71
TABLE 4.11: Summary Regression Output: Cable Viewership ....................................... 75
TABLE 4.12: Summary Regression Output: Magazine and Newspaper Readership ....... 78
TABLE 4.13: Summary Regression Output: Television Viewership ............................... 79
TABLE 4.14: Summary Regression Output: Internet Usage ............................................ 81
TABLE 4.15: Summary Regression Output: Auto and Truck Racing

Radio Listenership ............................................................................................................ 83
TABLE 4.16: Occupation and Media Usage .................................................................... 86
TABLE 4.17: Gender and Media Usage ........................................................................... 87
TABLE 4.18: Income and Media Usage ........................................................................... 88

xii

TABLE 4.19: Employment Status and Media Usage ....................................................... 90

TABLE 4.20: Age and Media Usage ................................................................................ 91

xiii

CHAPTER 1

INTRODUCTION

Background

Each year, advertising agencies attempt to create communications that will target
the right audiences with the right message. In an effort to reach the right audience, major
advertisers spend millions of dollars1 researching questions relevant to the demographics,
psychographics, and involvement with the product, and the media usage habits of the
intended audience.

This means that for any given product, many basic questions have to be answered
about the audience and its intended media usage. For example, the following questions
illustrate only some of those that must be answered by an advertising department or
agency to begin the process of creating and disseminating effective advertising. First,
who is the audience for the product or service to be advertised—can we create a proﬁle of
the “average” person in this audience demographically, psychographically, and so forth?
Secondly, what media do these consumers use and what types of variables affect the
target audience’s media usage? Finally, are there promotional efforts in which we can
participate that will increase our product’s Visibility? This last question is important
because it assumes we know the target audience well enough at this point to begin to

understand the particular lifestyles of our target audience, implying a deep understanding

 

' The exact ﬁgure of what is spent on research is extremely difﬁcult to ascertain. However, Nayyar

(2001) cites the ﬁgure as $5 .4 billion annually for 1999 market research spending.

of the consumer of the particular product. This suggests that the advertiser’s
understanding of the consumer goes beyond demographic information and allows a more
complete picture of the audience’s hobbies and interests, for example, making the
audience more of a “person” to whom the advertiser can talk.

Regardless of the product, in most instances, advertising is crucial for getting
even the simplest message about one’s product before the public. This is especially true
for nationally distributed products whose audiences are widely dispersed and might vary
in proﬁles. For example, how does a national advertiser communicate a beneﬁt to West
Coast car buyers who might be more inclined to purchase import vehicles while also
appealing to Midwestern buyers who might have more of a propensity to purchase
domestic vehicles? What is the common thread that ties the West Coast domestic buyer
and the Midwest domestic buyer together?

To make the advertisers’ job more difﬁcult, the landscape of advertising is
becoming increasingly cluttered. This is true not only because there are more actual ads
in traditional media venues, but also because there are increasing opportunities for
advertising placement. For example, a greater ratio of television air time is devoted to ads
in most dayparts (with the exception of prime time) than was true even 5 years ago
(American Association of Advertising Agencies/Association of National Advertisers,
2002). (Television dayparts are the segments of time broadcasters use for setting
programming and advertising pricing, for example, morning, prime time, and evening.)
Imagine a trip to the grocery store or to a ballpark only 10 years ago. It would have been

difﬁcult to imagine then what actually exists today: ads placed on virtually every

available space, including store ﬂoors and public bathroom doors. It would not be an
overstatement to argue that advertising is virtually everywhere (Stossel, 2001).

Given the proliferation of advertisements, then, how does an advertiser get a
message across without irritating the consumer or allowing the consumer to block out the
message? Some feel that one increasingly feasible way is to sponsor events of interest to
the target audience, thereby becoming part of the environment these people have come to
enjoy.

When the product to be advertised is one such as tires or oil ﬁlters (where the
association with automobiles is integral to the product beneﬁt) or an overnight delivery
service (where speed is a relevant differentiating beneﬁt), advertising is as crucial as for
any other consumer product or service. However, these products and services also have
the opportunity to tie their product in with the very relevant sport of auto racing. Auto
racing allows corporations to not only have their names seen by audiences, both live and
television, but to build an association between their product and a winning race team. For
example, auto manufacturers and auto parts manufacturers might rely on their alliances
with racing to promote technology transfer. (Technology transfer is the notion that
participating in sporting events such as racing helps manufacturers build better products
for consumers.)

While these promotional tie-ins are beneﬁcial, advertisers must remember that
sports consumers today are faced with what appears to be an almost limitless choice of
media delivering professional sports such as baseball, basketball, golf, auto racing, tennis,
and even bull riding in one day (Smith & Street, 2001). In addition, consumers appear to

be taking advantage of the sports media choices. Interests are becoming so diversiﬁed

that less popular sports such as tennis and soccer are seeing gains in fan support while
more popular sports such as professional hockey, basketball, and football continue to '
have a healthy fan base. It would appear, then, that fans are not switching from
consuming traditional sports to less traditional sports, but are consuming more of all
sporting events. That is, with the proliferation of choices, consumers are taking advantage
of the Opportunity to View relatively new sporting events (e.g., the WNBA) in addition to
the traditional events (e.g., the NBA).

Many consumers are broadening their sports viewing habits to include less
traditional sporting events. It is becoming more acceptable and feasible for sports fans to
split viewing time across a variety of sports. This is obvious when we see cable channels
dedicated solely to golf or racing.

Team spOnsors and marketers must understand, in this fragmented sports
viewership scenario, how various levels of involvement with sporting events translates
into media usage. If a consumer is involved with a sporting event or type of sport overall,
does this mean that he or she is inherently more likely to View this sport in mass media?
In person? Will the involved consumer be more likely to read or View specialty media
dedicated to the sport or to seek general information related to the sport? Even more basic
questions have to do with levels of involvement: Would those sports fans who might not
classify themselves as diehard fans, for example, still consume sports-related media?
Would those people who are reluctant attendees of sporting events still be consumers of
sports-related media? Each of these questions returns to the fundamental notion that it is
important to understand and measure involvement. Is involvement merely participating

by Viewing (either in person or via media)? Or does involvement include more, such as

following the personal lives of the sport’s players, talking about the sport with friends, or
purchasing sports-related materials? For example, one might theorize that since the
Internet offers easy accessibility to many people, reluctant fans of certain sports might be
likely to search Web sites for information on particular sports professionals while having
very little interest or involvement in the sport itself. The original study using the present
data found that people who reported themselves as reluctant fans of auto racing tended to
still seek information on the sport Via the Internet. Of particular importance to advertisers
and advertising agencies then, are some of the questions addressed in the present study.
Perhaps foremost is how to best measure involvement. This is crucial when placing your
product in the center of a speciﬁc event. The audience needs to have some level of
involvement to even pay attention to the event and, thus, your product or message.
Advertising departments and agencies often ﬁnd themselves in the situation of
having to create questionnaires and surveys that measure concepts such as recall,
attention to an advertisement, and even involvement (either with the product or the
communication). Very rarely, if ever, are these concepts based on theoretical
underpinnings in academic literature. In addition, industry-created survey questions are
typically based on intuitive knowledge of the concepts (for example, recall of an ad is
often based on whether the consumer can play back the sponsor for the ad, whereas recall
of creative elements is not counted as recall). Again, it is very rare for these concepts to
be tested statistically for reliability: Do the items consistently measure what they are
purported to measure? Or for validity: Do the items measure the concept they are

supposed to measure and not some other concept?

Just as important is the question of whether involvement is a good predictor of
speciﬁc media usage. Manufacturers measure concepts such as involvement in an effort
to best reach consumers (e. g., the assumption of those creating the 30-item involvement
scale used for this study was that highly involved translated automatically into high media
usage). This is, like so many aspects of consumer behavior, an assumption not to be taken
for granted.

Finally, it is important for marketers to realize the differences between different
media. Do people use certain media differently based on their involvement with the
product, message, or event? Are visual media used differently than auditory media? How
is the Internet used differently from other media? Also, what are the implications of

demographic variables on media usage, if any?

Purpose of the Present Study
Given the factors and questions described above, the goal of the study reported
here was to determine if:
I An intuitively based involvement scale, developed for use by a major advertiser, did
in fact measure involvement.
- Involvement measures, when valid and reliable, can act as predictors for speciﬁc
media usage.
I Demographic variables inﬂuence speciﬁc media usage.
The original involvement scale items from the survey were analyzed in the present
study to assess if these items did in fact measure involvement according to the academic

literature and statistical analysis. Once analysis conﬁrmed the involvement items, these

items (and others) were used to assess the role of involvement in the respondents’ media
usage. The involvement variables and others of interest were employed in the regression
analysis as the independent (or predictor) variables.

The impact of involvement on media usage is analyzed by looking at sports fans
using a data set that was originally gathered to analyze a variety of attributes of
motorsports (e.g., the NASCAR, Indy, and Lemans circuits) fans. In particular, the
original study measured involvement and media usage, but never the effect of one on the
other.

The importance of involvement as a concept in consumer behavior is apparent in
the literature and in the attempts made over time to deﬁne what involvement truly
consists of (see for example Krugman, 1965; Sherif & Cantril, 1947). Though still
somewhat nebulous to researchers, the concept is important to advertising practitioners
hoping to ﬁnd key concepts that will drive interest in their products.

The leisure studies literature provides a good basis for looking at sports
participation and involvement. In particular, specialization theory (based in recreation
and leisure literature) has been developed in the leisure literature to analyze participants’
involvement in certain leisure activities. Using specialization theory for a theoretical
background and underpinning, the data were analyzed to assess if there is indeed a
correlation or association between peoples’ reported interest levels and their media usage.

Media usage variables were taken directly from the survey as straightforward self-
reports of respondents’ motorsports media usage (for example, respondents were asked
“How many auto or truck races, if any, have you watched on TV in the past 5 months?”).

Each of these media usage variables was treated as a dependent variable in a regression

and AN OVA to assess what factors contribute to the prediction of each specific type of
media usage.

Chapter 2 presents an overview of involvement as it has evolved in the literature,
outlining the difﬁculty in deﬁning and measuring the concept. The concept is looked at
from the perspective of psychology, consumer behavior, and recreation literature. In
addition, chapter 2 gives a brief overview of NASCAR (the major motorsports circuits
addressed in the study) as well as media usage.

Chapter 3 provides the methodology used for this study. Chapter 4 presents the
results of analysis and Chapter 5 includes a discussion of the study’s ﬁndings,

conclusions, and next steps.

CHAPTER 2

LITERATURE REVIEW

This chapter reviews the literature relevant to the two main areas of interest in this
study. First, the literature relevant to involvement is presented. Speciﬁcally, literature
regarding involvement in consumer behavior and recreation are discussed at length. In
addition, some background on NASCAR and a review of media usage and involvement

are presented.

Involvement

The concept of involvement has evolved from the ﬁelds of psychology, social
psychology, and consumer behavior. The concept is one that has been examined (albeit
relatively infrequently) in the sports management and leisure ﬁelds, too. For purposes of
this study, the discussion of involvement centers primarily on its introduction and
evolution in the areas of psychology and consumer behavior, as well as its applicability in
leisure studies (which applies more directly to the topic at hand).

Unfortunately, there still appears to be no agreed-upon deﬁnition of involvement
across ﬁelds (Day, Stafford, & Camacho, 1995; Anti], 1984; Bloch, 1981; Beatty, Homer,
& Kahle, 1988; Greenwald & Leavitt, 1984; Arora, 1982; Hupfer & Gardner, 1971;
Havitz & Dimanche, 1997). Involvement can be looked at in a variety of ways, especially
when one realizes that it is applicable to a variety of ﬁelds. Psychologists, for example,
might think of involvement as something more introspective or thought—driven than

consumer behaviorists. Consumer behaviorists might think of the concept as one of

action—what a person buys or talks about. In particular, marketers hope to create an
involvement with a product or service and even involvement with an advertisement, the
idea being that this will lead to a purchase and eventually to product loyalty. Leisure
theorists also tend to put into operation the construct of involvement in behavioral terms,
perhaps even more so than consumer behaviorists (e.g., Bryan, 1979; Donnelly, Vaske, &
Graefe, 1986; McFarlane, 1994; McIntyre & Pigram, 1992; Schreyer & Beaulieu, 1986;
Scott & Godbey, 1994; Williams, 1984). Involvement tends to be of interest to leisure
studies researchers in particular when the focus is on time spent on the particular leisure
activity or equipment owned.

Speciﬁcally, involvement can be thought of as “an individual, internal state of
arousal with intensity, direction, and persistence properties” that affects the manner in
which a consumer responds to some stimulus (Andrews, Durvasula, & Akhter, 1990, p.
28) or a process whereby there are distinct, discemable stages of analysis of the involved
object (Greenwald & Leavitt, 1984). In particular, Andrews, Durvasula, and Akhter
(1990) posited that the state of involvement is one that, once reached, is resistant to
contrary communication. They further stated that those highly involved are more likely to
seek out inﬂuences that are in agreement with their feelings regarding the object of
involvement.

An alternate View is that involvement is a process whereby there are distinct,
discemable stages of analysis of the involved object (Greenwald & Leavitt, 1984).
Greenwald and Leavitt focused on the amount of lower to higher levels of processing of
elaboration to discern the level of involvement with an Object. (Processing of elaboration

is the creation of links between the object in question and already-known concepts.) The

10

levels each have increasingly stronger effects on the viewer or listener with regard to
involvement. One might think of their View, then, as involvement being deﬁned by
processing levels. That is, higher involvement equals higher processing of information.
However, Greenwald and Leavitt acknowledged that, while the impact of communication
at low levels of involvement is different than that at high levels of involvement, there are
still communication inﬂuences. In particular, the authors note that the four levels of
involvement (preattention, focal attention, comprehension, and elaboration) coincide with
levels of processing. For example, communications processed at the preattention level
might be thought of as background noise—communications that receive little, if any,
attention. As the levels of involvement increase, the cognitive effects and attitudinal
effects also increase or become more detailed.

Greenwald and Leavitt’s explication of involvement into four stages might be
thought of as an expansion of the ideas put forth by some communications researchers,
for example, Krugman and Zaichowsky, who stressed the importance of relevance in
their deﬁnition of involvement. However, the effect is the same—when a consumer ﬁnds
something highly relevant, the intensity of attention is affected, as was implied by
Andrews, Durvasula, and Akhter (1990). In particular, Greenwald and Leavitt posited
that at the highest level of involvement, the message is seen as important to personal
goals and elaboration of the message is based on its importance to the listener.

As noted above, involvement can be deﬁned by relevance (Zaichowsky, 1985) or
importance and interest (Antil, 1984). Anti] and others discussed the idea of involvement
being enduring or situational, the implication of this being that involvement need not

always be high to be effective (especially with regard to communication). For example,

11

after a recent purchase or newfound interest in an object or activity, involvement is high,
but ﬂeeting.

Obviously the discussion of involvement in various literature implies different
deﬁnitions. This is illustrated in the following discussion of the concept.

An unobservable state of motivation, arousal or interest toward a

recreational activity or associated product. In other words, involvement

refers to how we think about our leisure and recreation, and it affects our

behavior as well. (Havitz & Dimanche, 1999, p. 122)

Both deﬁnitions offer insight into how involvement can be multifaceted (e.g.,
pertaining to arousal and relevance) and can easily affect behaviors. Involvement, for
present purposes then, is not simply interest in a particular product, service, or event, but
also the notion that this interest (or relevance) can be a direct inﬂuencer or cause of
subsequent behavior.

This, of course, is problematic when one hopes to measure involvement. If
relevance is the deﬁning quality of involvement, how does one accurately deﬁne and
measure relevance? The same might be said of perception—accurate measures of how
one perceives different products or activities might be difﬁcult to create.

With varying deﬁnitions of involvement come various measures of the concept.
Part of the difﬁculty in deﬁning and measuring involvement stems, perhaps, from the
various ﬁelds to which the concept is being applied (e.g., Hupfer & Gardner, 1971;

Laurent & Kapferer, 1985; McQuarrie & Munson, 1987; Sherif & Cantril, 1947). Again,

12

psychological involvement will most likely mean a very different thing than consumer or
product involvement, implying that each will have different measures.

Academic research on the topic of involvement began in earnest from a
psychological standpoint (Sherif & Cantril, 1947) and has since received treatment in a
variety of ﬁelds. Sherif and Cantril’s treatment of the concept is often cited (regardless of
the ﬁeld of study) as the basis for examination of involvement.

Sherif and Cantril’s exhaustive look at ego-involvement discussed many
interesting areas of involvement, including, for example, the potential changes individual
involvement undergoes in a group setting. Perhaps the most important contribution of the
authors is their discussion of (ego-involved) attitudes in which they stated that “[a]ll
attitudes that deﬁne a person’s status or that give him some relative role with respect to
other individuals, groups, or institutions are ego-involved” (1947, p. 96). This
involvement ultimately shapes our attitudes and helps to deﬁne who we are and where we
belong.

The implication of this assertion is that involvement is at least partially comprised
of factors that deﬁne us to others. It might be read then that ego-involvement is ﬂexible,
based on what status or outward appearances one hopes to project.

Sherif and Nebergall (1965) further reﬁned the notion of involvement in their
social judgment theory, which posits that highly involved consumers have a relatively
narrow range of acceptable positions with respect to the issue being considered. They are
likely to reject most communication messages as these messages “are more likely to fall
within the unacceptable range of a person’s implicit attitude continuum” (Petty, Cacioppo

& Schumann, 1983, p. 137). This resistance to persuasive communication messages

13

implies a somewhat active communication consumer and one who is not easily
inﬂuenced by mass communication. This implies an involvement that is based more on
central values than on outward appearances.

Other researchers have taken this notion a step further by more explicitly linking
involvement to a behavior. For example, Greenwald (1965, p. 3) deﬁned position
involvement as “adherence to a prior behavior,” implying that the behavior will be
repeated in the future.

Ostrom and Brock (1968) later extended the social judgment theory posited by
Sherif and Nebergall (1965). The authors examined the idea that high involvement
creates resistance to even high levels of discrepant communication. In other words,
“Resistance to change is a direct function of the magnitude of ego involvement” (p. 379).
More speciﬁcally, the authors argued that ego involvement stems from value centrality,
relatedness (relevance) to a value and number of values engaged by the attitude in
question. These components of the authors’ model (based on cognitive theory beyond the
scope of this paper) purported that the creation of high involvement in an experimental
setting would bear out the idea of resistance to attitude change. It is interesting to note
that their experiment was “effective in altering resistance to attitude change” (p. 382).
Ostrom and Brock (1968) expanded the foundation of the idea of involvement to be more

action oriented and experimentally testable.

14

Involvement in the Communication Literature

The research of Ostrom and Brock (1968) and Sherif and Nebergall (1965) was a
Virtual contradiction to the assertions made by Krugman (discussed below) in his analysis
of communication and involvement, reiterating the differences in the role of involvement.

Perhaps the most inﬂuential authors in the area of involvement, from a
communication perspective, are Krugman and Zaichowsky, both of whom contributed to
the development and measurement of involvement. Krugman and Zaichowsky might be
thought of as having created the foundation for the discussion and evolution of the
concept of involvement as it pertains to communication. However, a variety of other
communication and marketing researchers have worked to deﬁne or better measure
involvement (see for example Arora, 1985; Beatty & Smith, 1987; Celsi & Olson, 1988;
Bloch, 1981; Lastovicka & Gardner, 1979; Petty & Cacioppo, 1981; Wright, 1974).

Krugman began work in the 19608 developing the notion that it is essentially
inappropriate to base consumer response to advertising (or packaging or new product) on
a single exposure. He argued that, in fact, people are exposed to advertisements again and
again, so to test response to one exposure was unrealistic.2 Krugman also argued that
learning is different for passive exposure versus active exposure to advertisements or
products by consumers. He argued that, while learning can take place for those who are
passively receiving messages, those highly involved will retain the message longer

(Krugman, 1965). When the message is involving, attitudes and perceptions are more

 

2 Interestingly, testing advertising after one exposure is still commonplace today. However, some
copy testing companies have realized the need to test ads not only in groups but also after repeated

exposure, thus creating a more realistic testing environment.

15

likely to change. This makes sense if one thinks of a product category in which one is
interested. If a consumer is in the beginning stages of new car shopping, for example, he
or she pays more attention to automobile ads and most likely retains the details of those
ads longer.

Later work by Krugman discussed how the role of involvement dictates at which
exposure a consumer will actually pay attention to the message of an advertisement. In a
1972 article, Krugman noted that basically all consumers need are three exposures to an
ad to learn its message. The fourth, ﬁfth, sixth, and subsequent exposures are simply
repetitions of the third exposure. The ﬁrst exposure is curiosity (“What is it?”), the
second exposure is recognition (“What of it?”) and the third exposure is decision or
reminder. What dictates when a person will go from one to two or from two to three
working (or relevant) exposures is the point at which he or she becomes involved. This is
the point at which the product or product category becomes important to the person—he
or she is in shopping mode or is helping someone else with a shopping decision, for
example.

Krugman argued that most commercial messages are ﬁltered out in such a way
that the ﬁrst exposure (the “What is it?”) is the end point. That is, even if the consumer
sees the same commercial 25 times over the next few weeks, he or she will not get to the
second working or relevant exposure (“What of it?”) until he or she is involved with a
(potential) purchase decision on that product category. Krugman noted, then, that
advertising is only powerful when the consumer is involved—something “largely outside

the control of television or advertising” (1972, p. 13).

16

While Krugman contributed greatly to the understanding of involvement
(especially with regard to advertising), consumer behavior researchers spent many years
measuring involvement in different ways. Zaichowsky noted that this might be due, at
least in part, to “different applications of the term ‘involvement’ ” (1985, p. 341). She
proposed a scale that would be applicable to involvement with advertising, product,
packaging, and so forth. Obviously, such a scale would need to incorporate psychological
and behavioral aspects of involvement.

After extensive reliability and validity testing (see Zaichowsky, 1985), the
resulting scale contained 20 semantic differential items. These items are, at their core,
based on the notion of involvement deﬁned by personal relevance. For example, in
measuring involvement with an advertisement, the measures are linked to factors that
make an ad relevant to the receiver (and hence motivating). For product research,
involvement is measured by the relevance of the product to the consumers’ wants and
needs. This ties back nicely to Krugman’s notion that people will become involved with
an advertisement only when it is relevant to them—when they can make a connection
with it.

The Zaichowsky involvement scale came to be known as the Personal
Involvement Inventory (PII). The PII scale can easily be adapted to assess involvement
with products, packaging, advertising, or purchase decisions. All that is required are
slight changes in the instructions to the respondents. However, the items remain the same,
regardless of what is being measured.

The PH shown in Table 2.1 consists of the complete scale used in the P11.

17

Table 2.1

Zaichowsky ’s Personal Involvement Inventory

 

[object or action to be judged inserted here]*

 

Important _______ Unimportant
Of no concern _______ Of concern to me
Irrelevant _______ Relevant
Means a lot to me _______ Unimportant
Useless _______ Useful
Valuable _______ Worthless
Trivial _______ Fundamental
Beneﬁcial _______ Not beneﬁcial
Matters to me _______ Doesn’t matter
Uninterested _______ Interested
Signiﬁcant _______ Insigniﬁcant
Vital _______ Superﬂuous
Boring _______ Interesting
Unexciting _______ Exciting
Appealing _______ Unappealing
Mundane _______ Fascinating
Essential _______ Nonessential
Undesirable _______ Desirable
Wanted _______ Unwanted
Not needed _______ Needed

 

* Note: As noted earlier, Zaichowsky envisioned a scale that would measure not only
involvement with advertising but also the product category or product itself, as well as a
purchase decision.

Ultimately, this scale measures involvement as a unidimensional construct by
summing all 20 items to produce a single score. Since the P11 is a 7-point scale, the
highest score attainable is 140, while the lowest is 20. Zaichowsky (1985), in addition to
validity and reliability testing, had respondents rate a variety of products to assess mean
PII scores for scale distribution. Products that fell on the short end (low involvement)

included instant coffee and bubble bath. Automobiles, laundry detergent, and calculators

fell on the upper end of the scale as products with high consumer involvement.

18

At roughly the same time Zaichowsky was developing the Personal Involvement
Inventory, Laurent and Kapferer (1985) were developing a multifaceted proﬁle of
involvement. The authors “argue that a consumer’s involvement can not be expressed in a
single score, because the type of involvement is as important as is its level” (McQuarrie
& Munson, 1987). Laurent and Kapferer’s Involvement Proﬁle (IP) is also comprised of
20 items, but measures four dimensions of involvement, which include risk, pleasure,
sign, and perceived importance. These Likert scale items underwent trait and
discriminant validity testing as well as reliability testing. In the end, Laurent and
Kapferer’s regression analyses appeared to prove that all of the items contributed to the
prediction of behaviors but that not all items inﬂuenced all types of behaviors. The
authors noted that this gives a much richer look than previous research provided at
involvement and hints at notions such as situational and enduring involvement. The
implication of their discussion was that perhaps the theoretical notion of involvement had
been oversimpliﬁed to that point. Perhaps not only are there different kinds of
involvement but that involvement tends to inﬂuence consumers differently based on the
situation.

The concept of situational and enduring involvement was discussed theoretically
(e.g., Rothschild, 1979; Houston & Rothschild, 1978) and experimentally (Richins &
Bloch, 1986). Enduring involvement changes only over long periods of time while
situational involvement is ﬂeeting, usually based on the period within which a purchase
decision needs to be made. Higie and Feick (1989) attempted to measure enduring
involvement with a reliable and valid scale. The authors’ upfront discussion of enduring

involvement is most interesting because it put the concept in proper perspective. For

19

example, they pointed out that product involvement should not be confused with product
importance. They proposed instead that the concept should be deﬁned mostly by the
hedonic quality of a product, for example. They agreed that relevance is important but
argued that other researchers have failed to measure relevance through the important
notion of hedonism. “Speciﬁcally, with enduring involvement, personal relevance occurs
because the individual relates the product to his self-image and attributes some hedonic
qualities to the product” (p. 690). They found that their Enduring Involvement Scale
(EIS) sufﬁciently measures enduring involvement among product categories. For
example, they found that the hedonism component of enduring involvement was more
important than self-expression in the category of personal computers.

Most authors agree that high enduring involvement is rare (Richins & Bloch,
1991). What is most common is a period of high situational involvement (SI) or concern
for the product usually brought on by a purchase decision. Of course, there are other
situations in which SI can be heightened, but to most researchers the purchase process is
of interest (Houston & Rothschild, 1978; Parkinson & Schenk, 1980; Richins & Bloch,
1991; Rothschild, 1979). Situational involvement wanes after the purchase has been made
and interest in the new product has subsided. Enduring involvement, on the other hand “is
a stable characteristic and doesn’t change much over time” (Richins & Bloch, 1991, p.
147). Of course, one must consider the context of involvement. Product involvement will
certainly have a different life cycle than sports fans’ involvement. Enduring involvement
in N ASCAR, for example, might be more enduring for fans who are members of fan
clubs, attend events more Often, or own a variety of NASCAR paraphernalia than for fans

who do none of these things.

20

Richins and Bloch (1991) examined levels of involvement and satisfaction,
hypothesizing that consumers with high EI tend to be more satisﬁed with their durable
goods purchase over time than low EI consumers. Upon analysis of car purchasers
longitudinally, the authors found support for this hypothesis. This could be due to the
inherent risk involved with being E1 and making a bad purchase decision or it could
simply be that high EI consumers ultimately make better purchase decisions. This
conﬁrmed the ﬁndings of Oliver and Bearden (1983) who examined satisfaction and
involvement as it related to an over-the-counter drug item.

The above review of involvement in academic research literature makes clear
some important points. First, there seem to be varying deﬁnitions of involvement
(relevance, interest, hedonism, etc.) that have been applied to a variety of different
research situations (e.g., attention to an ad, purchase decisions, satisfaction with a
purchase). Thought of another way, involvement has been used to mean a variety of
things to a variety of researchers. Secondly, measurement of the concept remains
uncertain. Part of the uncertainty regarding measurement comes back to the ﬁrst point—
how do we measure across disciplines what we cannot deﬁne? However, a great deal of
work has been done to both deﬁne and measure involvement. One could argue that the
concept has come a long way in its evolution. Finally, there seems to be a movement
away from interest in involvement, at least academically (as illustrated by the lack of
current literature assessing involvement). Perhaps this is due to the difﬁculty in
measuring and deﬁning it or perhaps due to the waxing and waning of interest in topics in

academe.

21

Taken together, these three points are crucial from the standpoint of the
advertising practitioner. Advertising agencies and their clients still maintain interest in
how involved consumers are with their advertising, products (both existing and
conceptual), and communication strategies. It is no surprise then that advertisers and their
agencies create, based on intuition, scales to measure involvement. Unlike concepts such
as attention to the ad, involvement was probably never clearly explicated for most
advertising practitioners (even those in advertising research). That leaves those creating
questionnaires to deﬁne involvement based on their own set of experiences and what, for

them, constitutes involvement with the object at hand.

Involvement in Recreation Literature

The concept of involvement has been explicated in many contexts. In marketing
and communication contexts, involvement with the advertising and the product are
probably the most notable treatments. To date, relatively little literature exists that
discusses spectators’ involvement with sports or sporting events. The majority of eXisting
literature regarding recreation involvement examines actual participation in a sport versus
spectator involvement (see for example, Havitz & Howard, 1995; Snyder & Spriezter,
1974). Interestingly, none of the leisure involvement literature reviewed for this study
discussed the idea of involvement as it affects media usage; that is, how involved with a
sport or leisure activity one needs to be to use media for the purpose of learning more
about that particular sport or activity.

This is an increasingly interesting area of review when one considers the plethora

of television and radio stations dedicated to sports. For example, how involved does the

22

consumer need to be to watch tennis on a regular basis? A less-than-regular basis?
Never? Are there intuitive differences in the level of media consumed based on
involvement?

Involvement is examined for its effect on information search (exploring various
sources for new knowledge on the topic of interest) for sports participants (see Havitz &
Dimanche, 1999, or Shamir & Ruskin, 1984, for speciﬁc examples); event attendance
(Ferrand & Pages, 1996; Laverie & Amett, 2000); and even approaches to product
consumption (Redden & Steiner, 2000). For example, Shamir and Ruskin looked at
differences in involvement in sports participation and spectatorship. They found that the
motivations (including levels of motivation) are different for sports participants and
spectators. They conﬁrmed many “intuitive expectations” (p. 19) when they ﬁnd that
involvement with participation in sports is related to health and wellness concerns, for
example.

Laverie and Amett (2000) perhaps provided one of the most comprehensive looks
at involvement in their analysis of fan attendance. For example, they noted that leisure
researchers tend to focus on either enduring or situational involvement, but not
necessarily the idea that both might occur with one person. The authors also made
apparent their knowledge of how involvement is deﬁned in consumer behavior literature
without mentioning directly the notion of relevance. In particular, they acknowledged
that involvement might consist of “motivation, arousal, or interest toward a recreational
activity or associated product” (p. 228). As noted previously, most recreation literature
discusses involvement in participation or even spectatorship, but rarely in the associated

products.

23

The general concepts of involvement discussed in consumer behavior literature
certainly provide a viable foundation for this exploratory research and are obviously
applicable to situations of leisure and recreation involvement studies.

As is the case with the concept of involvement in the communication literature,
there seem to be a variety of deﬁnitions and uses—none of which are agreed on (Havitz
& Dimanche, 1990). For example, Ellis and Witt (1994) measured depth of leisure
involvement, concentrating on respondents’ participation in a leisure activity. Selin and
Howard (1988) used ego involvement to explain attachments to leisure activities.
Mannell (1980) discussed involvement based on the level of interest one has in a product
when looking at information-seeking in leisure pursuits. Laverie and Amett (2000) noted
that enduring involvement and situational involvement appear to be the ways
involvement is described in leisure literature.

Havitz and Dimanche (1990) summarized best when they noted that “Attempts to
measure involvement in the recreation and tourism literature have generally been limited
to the antecedents or to the behavioral consequences of involvement” (p. 181). For
example, they noted that Gunter and Gunter (1980) discussed involvement in terms of
engagement (high involvement) and disengagement (low or nonexistent involvement).
This is only one example of cases where involvement essentially equals participation or
activity. In fact, Havitz and Dimanche noted that the concept of involvement, when
studied in the leisure literature, is rarely explicated but somehow vaguely referred to and
assumed to be understood by the reader.

Perhaps the broadest and most inclusive approach to involvement in the leisure

literature comes from Bryan’s (1977) attempt to understand involvement based on things

24

such as time and money investment as well as kinds of equipment used to participate in
sporting events. Bryan’s Specialization Theory was used to better understand the
differences in involvement among sports participants (versus viewers or fans), but still
provides an excellent basis for analyzing how even the sports fan can be involved.

Bryan did not explicitly deﬁne involvement, but rather gave the reader an
understanding of how involvement essentially comes to equal specialization. High
specialization (involvement) is characterized by commitment, motivation to perform
better, interest in the sport, and relevance, as noted by the pride in and knowledge of
equipment, investment in time and equipment, and association with others who were also
highly involved with the sport.

Given Bryan’s discussion of specialization and involvement, as well as the
deﬁnitions provided by consumer behavior literature, for purposes of this study,
involvement has been deﬁned as:

A product, activity, or event that is highly relevant and meaningful and

often motivates one to invest time, money, and effort in pursuing the

activity, event, or product. In particular, high involvement creates a desire

to invest more money, time, and effort in the pursuit of keeping current

with the product, activity, or event.

For example, if a person is highly involved with a particular product he or she will
invest the time to ﬁnd a store that carries that product, invest the money necessary to
acquire the product, and the effort to stay informed about the product’s changes,

speciﬁcations, and so forth. The same can be said for events (such as sporting events) or

25

sports participation. Interestingly, Bryan included not only money but also time as

relevant investments in his discussion of anglers.

Specialization Theory

Specialization theory was ﬁrst discussed in publication by Bryan in 1977 and
expanded in various recreation publications (e.g., Backman & Crompton, 1991; Ditton,
Loomis, & Choi, 1992). The theory states that recreation participants each have a degree
of specialization based on their involvement in the activity. Specialization might be
thought of as level of expertise (for example amateur to professional), which means not
just how well someone performs at a given sport, but the degree to which he or she takes
the sport seriously. Each level of specialization carries “distinctive behaviors and
orientations [which] include equipment preference, type of experience sought . . .
preferred social context, [etc.]” (Bryan, 2000, p. 18). Specialization theory does not
strictly deﬁne involvement, but more loosely bases it on things such as equipment used
and expertise in the activity. The implication of this is that as specialization in a leisure
activity increases, so does the time spent with that activity, money spent on it, and grade
of equipment used in it. In particular, it is deﬁned as “a continuum of behavior from
general to particular reﬂected by equipment and skills used in the sport and activity
setting preferences” (Bryan, 1977, p.175).

Bryan proposed and developed specialization theory in the 1970s in an effort to
explain what he had observed in sports enthusiasts, speciﬁcally anglers. Bryan began to
notice the way various sports enthusiasts went about their chosen sport and the conﬂicts

that arose. He noted that often the anglers would be in conﬂict with, for example, boaters.

26

Each group was participating in their sport of choice and wanted an area for their sport.
The anglers, to continue the example, would desire an area free of distractions—to
themselves and the ﬁsh—while boaters wanted an area free of hazards. This conﬂict was,
to Bryan, one that could have been expected.

The conﬂict that Bryan did not expect was the one among the groups of sports
enthusiasts themselves. He noted, “I was especially struck by conﬂicts within these
groups, particularly trout ﬁshermen, as they clashed among themselves over the
appropriateness of ‘catch and release’ and ‘ﬂy-ﬁshing only’ regulations, stream etiquette,
and a host of other issues” (Bryan, 2000, p.19). Bryan also noted that the differences
seemed to stem from commitment to or involvement in the sport. Highly involved anglers
differed greatly when compared with the less involved anglers. Perhaps the most obvious
examples were the investments in equipment and observation of rules (stated and
unstated). Bryan summarized the theory by saying that participants “can be placed on a
continuum from general interest and low involvement to specialized interest and high
involvement” (Bryan, 2000, p.18). General interest might be thought Of as an interest and
participation in ﬁshing of any kind on a nearby lake, while specialized interest would be
catch-and-release ﬁshing on certain waterways, using more professional grade equipment
and following more exacting rules. Of course, the notion of specialization can be applied
to almost any sport. The more specialized the participant becomes, the more particular he
or she becomes about the equipment and venues used to participate in that sport.

There are many implications of the theory. Most important of these is the idea that
those highly involved are more likely to invest in equipment and time spent participating

in the sport. For example, Buchanan (1985) and Backman and Crompton (1991)

27

characterized the concept of investment in recreation activity as “side bets.” Interestingly,
this term most commonly refers to activities held simultaneously with other competitive
activities. In golf, the idea of “skins” between players of a game is often a monetary
wager related to which person or team will perform better from hole to hole. This is a
wager aside from the actual game of golf being played, hence a side bet.

However, in recreation literature, side bets are regarded as slightly different than
the wagers between friends during a friendly game. These bets are the investments of
one’s own time or money in the activity.

Side bets are an important part of involvement in an activity, whether it be
recreational or purely product or marketing related. For example, involvement with
exercising on a regular basis and involvement with certain products or marketing
communication (e. g., advertising) require certain investments. Side bets are the
investments made in relation to participation in an activity or a product. The investment
made when one is involved with exercise might be not only the investment of time, but of
money (for gym membership, equipment, etc.) as well as the psychological investment of
trying to make oneself better (e.g., motivation, positive thinking). The investments or side
bets involved with communication are, again, the time and energy spent devoted to the
message versus time spent doing something else or, in some cases, nothing at all.

These side bets have implications with regard to many things, the most interesting
for present purposes being investment in the actual sport or activity and investment in
media associated with the sport or activity. That is, how much of an investment does one
make in watching, listening to, or reading about the particular sport of interest. The

investment made can be that Of actually attending a sporting event or of taking the time

28

(and perhaps money) to ﬁnd the appropriate media outlets with information of interest on
that particular sport.

Backman and Crompton (1991) analyzed involvement in recreation activities and
the resulting loyalty. They looked speciﬁcally at attachment and consistency in behaviors
related to the activity. They focused on the following side bets or investments when
looking at loyalty and participation in leisure activities:

I Equipment

I Friendship

I Magazine subscriptions

I Money spent on the activity

Their ﬁndings were interesting, albeit somewhat intuitive. They found that there
was a strong relationship between side bets and loyalty. In particular, high side bets
related to high involvement. The authors’ thinking on this point was that people were
encouraged to keep participating in the sport (i.e., golf and tennis, in this particular study)
because of their investment in things such as equipment and friendships.

‘ Buchanan (1985) measured these investments based on scales developed by Bloch
and Bruce (1984) and Bryan (1977). This scale, with an alpha of .86, measured items
such as money spent on the game, equipment, magazine subscriptions, and time spent
watching the game on television. These items were found to relate to involvement in the
game as well.

Ditton, Loomis and Choi (1992) further developed Bryan’s theory of recreation
specialization (as well as that of side bets) in an attempt to provide support for some of

the propositions set forth in the theory. The resulting article was an interesting look at

29

how various sporting participants are grouped into recreation social worlds and
subworlds based on recreation activity. This is somewhat of an explication of Bryan’s
work in that it expands the notion of low specialization and high specialization into four
categories (strangers, tourists, regulars, and insiders) outlining each group’s orientation,
experiences, relationships, and commitments. Perhaps the most interesting and relevant
notion put forth by the authors is the idea that highly specialized sports participants
(speciﬁcally anglers) exhibited a higher media interaction than did those who were less
specialized. In particular, the authors looked at speciﬁc radio shows, magazines, and
television shows and found that those who were categorized as highly specialized
interacted with (i.e., used) these media more often than those who were less specialized.
In addition, the authors found that specialized anglers exhibited higher resource
dependency. In particular, they found that specialized anglers tended to agree that the
more and bigger ﬁsh they caught, the better.

Ditton, Loomis and Choi (1992) noted that for many recreation activities, on-site
spectatorship or participation is not feasible for everyone. They pointed out that as people
become more widely dispersed, the need for information about a particular recreation
activity will increase. This idea is not new (see for example, Crane, 1972; Goode, 1957;
Katz, 1958; Mullins, 1973; and Shibutani, 1961) but it bears repeating as the notion of
fans (or even sports participants) is discussed in the current fragmented media
environment. McIntyre (1989) went a step further in linking enduring involvement and
specialization theory. He noted the need for enduring involvement to be put into
operation in recreation research to try to more fully understand recreation enthusiasts’

levels of specialization. McIntyre stated that “levels of enduring involvement would be

30

indicative of degrees of specialization” (1989, p. 170). His analysis of enduring
involvement being comprised of enjoyment, importance, centrality, and self-expression
introduced recreation researchers to the notion of involvement being multifaceted.
McIntyre used items directly from the Involvement Proﬁle (IP), as well as others. He was
able, through his analysis of beachside campers, to validate such items for recreational
involvement measurement.

Dimanche, Havitz, and Howard (1991) further developed the notion of
involvement in recreation settings (versus consumer behavior settings). In addition, they
noted the multidimensionality of involvement. It is this multidimensionality that they
examined closely by using the IP in a recreational research setting. The authors looked at
a variety of recreation and tourism situations and were able to measure involvement using
Laurent and Kapferer’s English translation of the IP. Not only did they support
application of the IP to recreation literature, they also supported the opinion of other
researchers (Havitz & Dimanche, 1990 and 1999; Higie & Feick, 1989; Iwasaki &
Havitz, 1998; McIntyre, 1989; McQuarrie & Munson, 1987; Zaichowsky, 1985) that
involvement is a multidimensional construct.

The implications of specialization theory are that:

I Fans can be placed in categories according to their sports specialization behavior and
subsequent purchase and media usage behavior.

I High-interest recreation participants or fans will be the group consuming the mOst
live sports events, mass-mediated sporting events, and sports-related products

indicating more investment in their sports of interest.

31

Such implications are being tested by racing fans today, for example, who defy
the traditional stereotypes about NASCAR fans (e.g., that they are less educated and have
lower incomes than the general population). In addition, such implications assume that
high-interest fans are those who consume the most sports-related media. With the
diversity of fans emerging in motorsports, this might be a faulty assumption. More
moderate fans might use certain race-related media more than diehard fans. For example,
moderate fans might not watch races as religiously as diehard fans but might consume

more follow-up stories in the media (e.g., newspaper, Internet) to learn race results.

Involvement Summary and Implications

The above review of involvement in research literature makes clear some
important points. First, varying deﬁnitions of involvement (e.g., relevance, interest,
hedonism) have been applied to a variety of different research situations (e.g., attention to
an ad, purchase decisions, satisfaction with a purchase). Secondly, there is a lack of
consensus in how to measure the concept. Third, some literature does not make clear
whether involvement is enduring or situational, nor does it address the issue of whether
the two need to be delineated. Finally, there seems to be decreasing interest in the topic of
involvement, at least academically (as illustrated by the lack of current literature
assessing involvement).

Taken together, these three points are crucial to the advertising practitioner.
Advertising agencies and their clients still maintain a great deal of interest in how

involved consumers are with their advertising, products (both existing and conceptual),

32

and communication strategies. It is no surprise, then, that advertisers and their agencies

create, with little or no academic input, scales to measure involvement as they deﬁne it.

Changing Media Usage

Media choices have been growing and evolving rapidly over the past 25 years.
Cable television, the Internet, niche magazines, and even increased programming choices
within these media have created a more fragmented media landscape. This means more
choices for consumers and a more difﬁcult time for advertisers attempting to reach target
markets in traditionally targeted media vehicles (Picard, 1999).

One might argue, then, that audiences are becoming less homogeneous as they
become dispersed. For example, women might not only be watching daytime dramas but
tuning in to sports specialty channels, creating a diverse audience for these channels.
Consumers are more likely to be taking advantage of the media choices available to them
and, hence, be consuming more media types in the time they have for such endeavors. In
addition, they are likely to be almost constantly consuming (different) media with the
availability of personal computers, personal digital assistants (PDAS), cellular telephones
with Internet access, and so forth. So, the amount of time devoted to media consumption
has increased, but so have the attractive media choices. Thought of another way,
audiences have become more fragmented but people want to be constantly connected to
others and to things in which they are interested.

How does this affect sports viewership? One example is the Super Bowl which,
while still a great audience draw, is “showing signs of audience fragmentation”

(McAdams 2001, p. 100). In 20 years, ratings for the Super Bowl have dropped 20%,

33

which could be partly due to the increased alternative offerings. Most of the alternative
media vehicles are targeted to women (for example, see McAdams, 2001) and include
other sporting options, such as ﬁgure skating. The result of audience fragmentation,
greater time spent consuming media, and greater media offerings is more variety of all
types of programming. Reality shows, for example, can coexist with traditional situation
comedies just as golf and motorsports can each have their own channel. There is an

audience for each and, sometimes, these audiences overlap.

Sports Growth in the United States

Sporting events are prime vehicles for the new media channels. For example, the
Internet allows consumers to catch up on, or watch in their entirety, certain sporting
events. According to Pastore (1999), the Internet challenges the traditional television
media for presentation of sporting events. The Internet is making obvious inroads by
offering league and team Web sites, although viewing an entire sporting event is also a
possibility. The implication for marketers who work for sports with growing fan bases is
that they must examine new media opportunities carefully and use them wisely. In
addition, the traditional fan of such sports might not be the only kind of person to whom
viewership is appealing. That is, the casual fan or observer of a sport might be just as
important a target—considering how much of the particular sport’s media is consumed
and whether such a person has the potential to become an even bigger fan of the sport.
With so many sports viewing options, consumers might easily tune to sports they had not
had easy access to before. In addition, fans might ﬁnd tuning into sporting events less

expensive and more enjoyable than attending in person. This is especially true for sports

34

such as golf, tennis, and racing where the cost is high or a regional venue is not
immediately accessible. These sports, in particular, have a history of having a very select
audience. The audience for these sports is increasing, as is media coverage.

NASCAR is an excellent modern example of a sporting option that is
experiencing growth in both popularity and fan base. What is more intriguing to
marketers is the fact that with fan base growth comes fan base diversity. This is
illustrated below in the discussion about the fans themselves. First, however, the basics of

the sport itself are outlined to give the reader an idea of what NASCAR racing really is.

NASCAR

The National Association of Stock Car Racing (NASCAR) was founded more
than 50 years ago. NASCAR is comprised of many divisions and series (see Table 2.2)
and sanctions roughly 2,000 events each year. Divisions include major series such as
Winston Cup, Busch Grand National, Craftsman Truck, and a variety of local and
regional series. The major series differ by vehicles that are allowed to compete in them.
For example, in the Winston Cup series, slightly modiﬁed street vehicles compete, with
only the three most recent years’ models allowed to race. Busch Grand National Division
series cars are similar to Winston Cup cars but are lighter (by 100 pounds), shorter (by 5
inches) and less powerful. The Craftsman Truck series allows pickup trucks to race with
modiﬁcations similar to those of the Winston Cup cars. The remaining series are typically
local or regional and feature cars similar to Winston Cup, Busch Grand National, or

Craftsman competitors, depending on the venue (Hagstrom, 1998).

35

Table 2.2

Overview of NASCAR Divisions

 

National Series—Premier Division

 

NASCAR Winston Cup Series

* N ASCAR’s premier division with events at 23 tracks nationwide

* The top racing series in the United States, with more than 6 million fans in attendance a
year

NASCAR Busch Series Grand National Division
* The second most popular series with 32 events nationwide

NASCAR Craftsman Truck Series
* Races with full-sized pickup trucks
* 23 tracks nationwide

 

Local Series

 

NASCAR Weekly Racing Series
* Almost 100 short races in 10 regions

 

Regional Series - Touring Division

 

* NASCAR Winston West Series

* Busch North Series

* Featherlite Modiﬁed Series

* Gatorade All—Pro Series

* Goody’s Headache Powder Dash Series
* Featherlite Southwest Series

* Raybestos Brakes Northwest Series

* REIMAX Challenge Series

* O’Reilly Auto Parts All-Star Series

 

NASCAR primarily differs from other organized racing in its car types. For
example, while Indy racing uses the long, lean, bullet-shaped “open wheel” vehicles,
NASCAR uses stock cars, which are modiﬁed street cars. A Chevrolet Monte Carlo

driven on the NASCAR circuit might look like your neighbor’s Monte Carlo, but it

36

differs in its internal equipment and its sponsorship decals.3 While the original intention
of NASCAR was to race “absolute showroom models like the ones folks drove to work
every day” (Cotter, 2001), by the early 19608 these cars were being modiﬁed in the
factory and included safety features in addition to speed-enhancing features.

N ASCAR is the fastest growing professional sport in the United States (Parnell,
1999). This is evidenced by its growing fan base, and made more impressive when one
looks at fan characteristics (see Table 2.3). Contrary to popularly held conceptions,
NASCAR fans as a group are reﬂective of the general population. For example,
NASCAR fans are just as likely as the general population to drink wine, own and use
cellular phones, travel, and invest (Harris Interactive, 2001). They are also slightly more
afﬂuent than the US. population as a whole (NASCAR New York Corporate Marketing
Ofﬁce, 2001).

Weekly racing series fans are more avid fans and more likely to support sponsors,
and feel that the sport is part of a family experience for them (NASCAR New York
Corporate Marketing Ofﬁce, 2001) when compared to moderate or reluctant fans. When
one looks at fans more closely, other differences in types of fans become apparent.
Hardcore or diehard fans differ from avid fans and casual fans and nonfans. What is
perhaps most interesting among these fan groups is how the interest in the sport is
demonstrated. For example, only “23% of diehard or avid NASCAR fans attended a

NASCAR event” during the 2000 season (Harris Interactive, 2001, p. l).

 

3 Interestingly, there is a movement currently to have all stock car vehicles appear the same with

the exception of sponsorship decals on the vehicle’s exterior.

37

Table 2.3

NAS CAR Fan Characteristics

 

Weekly and Touring Racing Series Fans

 

* 72% are hardcore fans

* Average attendance is four NASCAR races in the past year

* Follow races on the radio and Internet more than other fans

* Purchase intent of these fans is more than double that of other fans

* 51% say they feel that they are part of the “NASCAR family”

* 90% see NASCAR drivers as positive role models for families

* Purchase more than $1,100 in NASCAR-related products each year

* Are more likely to be overall sports fans, attending professional or college sporting
events

* Are more likely to be active in sports or outdoor recreational activities
* 55% regularly work on their cars

 

This means that marketing to fans only at racing events is perhaps only one way
to reach fans of every level. Many diehard NASCAR fans regularly watch races at home
on television or listen to them on the radio. This implies that sponsorship of events at the
events alone might not be sufﬁcient to reach these diehard fans. Additional media
purchases might also be called for.

It is also important to acknowledge that sports and racing fans are also very active
in other leisure pursuits. This gives rise to cross-entertainment marketing and, perhaps
more importantly, a new way to View sports fans. In addition to viewing the diehard fan
differently, it is perhaps advantageous to take a closer look at the casual and nonfan as
well. For example, those people not interested in motorsports but viewing races for a
variety of reasons (including time spent with signiﬁcant others and a party atmosphere)
might also be targets for advertisements about certain products and services.

The implication of casual or nonfans being signiﬁcant targets for media

communication ﬂies in the face of specialization theory, which posits that highly

38

involved fans or sports participants are the only fans with high interest levels. In fact,
looking more broadly at the concept of involvement, one might argue that some less avid
fans do indeed have high situational involvement. These fans might ﬁnd the race highly
relevant, motivating, and interesting on race day, but at very few other times, for
example.

However, taking the notion of involvement and subsequent media usage one step
back, it seems important to ask whether involvement does, in fact, have an effect on
media usage. The present study examines this question—this is a ﬁrst step in applying
specialization theory to media usage. It would seem only logical for both manufacturers
and researchers to assess whether involvement inﬂuences or predicts media usage—

especially when message placement is under consideration.

Research Questions

The following research questions, then, are put forth for consideration. The
analysis performed with the present data set attempts to answer these questions. Chapter
5 discusses the implications of the analysis and answers to these questions.

RQl: Do any of the items from the original questionnaire measure involvement?

RQ2: If so, what items from the questionnaire measure involvement?

RQ3: What kinds of items measure involvement?

RQ4: Finally, do the involvement items measure involvement as it has been
deﬁned in the academic literature?

RQ5: Does involvement contribute to the prediction Of media usage?

RQ6: What role do other concepts play in the prediction of media usage?

39

The next chapter describes the methods used for analysis of research questions 1
through 6. In addition, a discussion of the analysis done for the original study is

presented, as well as a description of the variables used for the present study.

40

CHAPTER 3

METHODOLOGY

This chapter describes the research methods used to answer the questions put
forth in chapter 2. This study relied on secondary analysis of data from another study to
which the author had access. A description of the original study, including its sample, is
provided, followed by a description of the analyses performed to investigate the present

study’s research questions.

Background

The data for this study came from a survey developed by a major Midwestern
advertising agency and its automotive client, in conjunction with an independent research
ﬁrm. A major portion of the survey itself (a grouping of 30 questions) was devoted to the
measurement of consumers’ involvement in motorsports.

This survey represented the second phase of a longitudinal look at motorsports
interest among automotive consumers. The funding corporation has an enormous
investment in sponsorship of motorsports, and annually there is concern regarding
whether such sponsorship is fruitful for the company. The underlying question is whether
these sponsorship dollars could be better spent elsewhere (e.g., in more advertising, or
toward youth or women’s initiatives).

The portion of the questionnaire that was developed to analyze respondents”
involvement had never been tested for validity or reliability. Post-hoc reliability and

validity testing of the scale are presented in this study.

41

Data were also collected on respondents’ use of speciﬁc motorsports media; for
example, the number of auto or truck newspaper articles read. These questions were used
in the present analysis to determine whether involvement variables are predictive of

media usage.

Sample

Consumers were selected for completion of a telephone interview based on
several criteria outlined for the original study conducted by an independent research
organization. To qualify for inclusion in the study, respondents must have owned an
automobile from the 1996 to 2000 model years and also have been intending to purchase
or lease another vehicle within the next 5 years. In addition, respondents had to be
involved in the purchase or lease decision for the next vehicle. Finally, respondents could
not work for an advertising agency, marketing ﬁrm or department, research company, or
automotive manufacturer.

The original list of names was supplied by R. L. Polk’s auto registration records
and consisted of 13,283 consumers in the United States. Consumers from this list were
randomly chosen. Of those called and screened, 1,380 respondents qualiﬁed and
completed the telephone questionnaire.4

Telephone interviews were conducted between April 12 and 29, 2000. The

purpose of this tinting was to allow interviewing to coincide with the seasonal lull in

 

4 It should be noted that respondents’ privacy was protected since respondents’ names were not
connected in any way with the resulting data responses. In addition, the research supplier adheres strictly to

research guidelines regarding consumer privacy.

42

motorsports activity. The thinking was that respondents would not be inundated with
motorsports-related messages (e.g., commercials with motorsports connections) and
mistakenly report having viewed motorsports events. Attempted call-backs for
respondents who were unavailable ranged anywhere from ﬁve to seven calls, depending

on urgency (i.e., how close to completion of the study) and availability of calling staff.

Original Analysis

The original intent of the survey was twofold. The primary purpose was to assess
the extent to which consumers expressed an interest in motorsports and the number of
consumers at each level of interest. For example, are there more (or fewer) respondents
who are avid fans of motorsports than who are nonfans?5 The rationale behind this
research objective was to justify the existence of motorsports sponsorship. If, for
example, the ﬁndings concluded that there were more avid motorsports fans than in
previous years, there was justiﬁcation in that the dollars spent sponsoring auto racing
were reaching more people.6 If, in fact, the sheer number of people who reported being
racing fans increased, there was felt to be justiﬁcation in spending dollars to reach more
people.

The second purpose of the study was to measure how, if at all, the funding

corporation’s name and the name of competitors were associated with motorsports.

 

5 The study ﬁndings are interesting in that they categorize respondents based on fan level ranging
from diehard fans to reluctant attendees. These categories are of great interest to a motorsports sponsor.
6 Of course, this assertion relies on direct comparison of fans’ reported level of interest in

motorsports from the previous study.

43

Again, the funding motorsports sponsor wanted to assess whether the money being spent
on such sponsorship was worthwhile. The questions asked by the sponsoring company
not only included those pertaining to association with the sport, but also the association
of competitors who also sponsored motorsports. In essence, the hope was to get a better
feel for what sponsoring company is best recalled as the top sponsor of the sport. For
example, not only were respondents asked which sponsoring brands they were aware of,
they were also asked to rate sponsors on their level of involvement in motorsports and
whether each sponsor was viewed as “a winner” in the sport.

As mentioned above, this was the second phase of a longitudinal analysis
regarding how consumers’ interest in motorsports changes over time. Beginning in 2002,

the study is to be conducted every 2 years for comparison purposes.7

Independent Variables

The original involvement variables were measured on a 5-point Likert scale
ranging from agree strongly to disagree strongly with neither agree nor disagree as the
center point. These items were created in concert with the research supplier and client
based on intuition and knowledge of the target audience. Due to this, the items have
several innate groupings. For example, it was the intent of the client to measure beliefs in

technology transfer, the social aspect of the sport (how much it is shared with friends),

 

7 The original study was conducted in 1996. The study whose data is used in the present
discussion, was conducted in 2000. The parent company of the funding corporation expects to continue

conducting the same study every 2 years beginning in 2002.

44

and interest in the sports’ participants (i.e., race car drivers), all of which were considered
indicators of involvement.

Research Questions (RQs) 1 through 4 were designed to address the issue of
whether involvement is truly measured by this industry-commissioned survey (see the
Research Questions on p. 39). RQl was designed to simply assess whether any of the 30
items in this survey actually measured involvement. RQ2 and RQ3 were designed to help
the researcher better understand—if in fact involvement was measured through any or all
of these items—what kinds of items measure involvement. Put another way, it was hoped
that RQ2 and RQ3 both would answer the question of which items measured involvement
and how involvement was expressed in these items. Finally, RQ4 was designed to assess
whether the items used in the survey that were found to measure involvement actually
reﬂected involvement as it is deﬁned in the literature. Answers to the research questions
were ascertained using factor analysis to deﬁne what types of concepts were being
measured in the original 30—item scale. RQ5 was designed to address the issue of the link
between involvement and media usage variables (described below). Finally, RQ5 was
designed to help researchers better understand if any of the other variables measured in
the original survey might assist in better understanding motorsports media usage. In
particular, demographic variables (including age, gender, employment status, occupation
and household income) were used for this analysis. These variables are, taken together, a

combination of ratio and nominal measures. Methodology is detailed below.

45

Dependent Variables

In the original survey, media usage questions were included to analyze simple
frequencies and means (for example, “What is the average number of reported newspaper
articles read in the past month by the sample?”). In addition, separate analyses were
completed on respondents who reported themselves as “NASCAR only” fans.

Media usage questions allowed respondents to report usage of traditional media
such as newspapers, television, and magazines. In addition, more contemporary media
such as cable and Internet usage (as it related to motorsports) were also included. These
questions allowed respondents to report approximate numbers of each media vehicle used
within a speciﬁc time. The time frames used (3 and 5 months) were random, based on

input from the funding client, research supplier, and advertising agency.

Current Study Variables and Methodology
The current study sought to analyze questions regarding industry questionnaire
validity and reliability, as well as the role of involvement in media usage. These
questions were addressed using factor analysis, reliability analysis, and regression

analysis. These methods and their role in the present study are described below.

Reliability Analysis

Reliability analysis was performed Via SPSS’s reliability analysis function. The
option of “scale if deleted” alpha was included. This allowed analysis of the 30-item
scale’s reliability if each item were to be deleted. This provides an accurate picture of

how reliability might change when an item is not included in the total scale.

46

Factor Analysis

Factor analysis was performed to examine whether the 30 variables in the original
questionnaire were in fact measuring a single concept—involvement. Exploratory factor
analysis was done using a portion of the sample and conﬁrmatory factor analysis was
performed using the full sample. Principal components extraction was indicated as it
extracts factors that account for less and less variance. In addition, varimax rotation was
performed in an effort to minimize items loading highly on multiple factors. Also
indicated, and reported in the next chapter, are the Kaiser-Meyer-Olkin Measure of
Sampling Adequacy (KMO) and Bartlett’s test of sphericity. KMO ranges from 0.0 to 1.0
and tests whether variables are measuring a common factor. When the KMO ﬁgure
approaches 1.0 the variables are measuring a common factor and, conversely, a KMO
ﬁgure near 0.0 indicates that the variables are not measuring a common factor. Bartlett’s
test of sphericity is a measure of whether all correlations, when tested together, are not
statistically different than 0. In other words, Bartlett’s test of sphericity tests whether the
correlation matrix differs from the identity matrix (Pedhazur & Pedhazur—Schmelkin,

1991).

Regression Analyses

With the involvement variables better deﬁned, analysis of the ﬁnal research
questions could take place. These research questions had to do with the effect of
involvement on media usage. Given the nature of the data set, these questions speciﬁcally

set out to answer whether interest in motorsports was a feasible predictor of speciﬁc

47

motorsports media usage. The statistical method used to answer this question was

regression, which allows the researcher to answer the following questions:

I Is there a statistical relationship between the dependent and independent variables
that affords predictability?

I How strong is the statistical relationship?

I Can a formula be derived that looks at each independent variable and its effect on the
dependent variable? (This would, of course, be the regression equation.)

While correlation alone might be able to shed light on the ﬁrst two points above,
regression allows the researcher to formulate the effect of the independent variables,
taken together, on the dependent variable (Hays, 1988). This is what makes regression
unique and useful for the present study.

Linear regression was performed with the “enter” method speciﬁed. Stepwise
regression was also performed, but found to not contribute much. Resulting output of
interest included R2, signiﬁcance of F from the ANOVA test, beta coefﬁcients, and
signiﬁcance of the t test for each of the variables. The R2 measurement measures the
proportion of variance in the dependent variable explained by variations in the
independent variables. The AN OVA signiﬁcance ﬁgure reported in the output allows one
to answer whether the model worked to explain deviations in the dependent variable.
Typically, the researcher is seeking a signiﬁcance ﬁgure of less than .05 for signiﬁCance
at the 95% level (Gupta, 2000). Beta coefﬁcients provide information regarding the effect
of each variable on the dependent variable (e. g., negative, positive, strong, or weak). T
test signiﬁcance allows the researcher to assess the statistical strength of the contribution

of each variable in the regression equation. These are reported and discussed in chapter 4.

48

Analysis of Variance (ANOVA)

Analysis of variance (ANOVA) allows the researcher to compare the means of
one or more groups taking into consideration one independent variable. An example
would be using the independent variable of age and the dependent variable of number of
diet sodas consumed weekly. Using this example and the assumption that respondent’s
ages were categorized in such a way that there were distinct groupings (e.g., 12 to 17, 18
to 25, and 26 to 34), ANOVA allows comparison of these groups’ means to determine
whether each age group differs signiﬁcantly from another in number of sodas consumed.

For the present study, AN OVA allows analysis of whether the demographic
independent variables have an effect on media usage (dependent) variables. In particular,
categorical analysis of responses gives the researcher a better picture of the possible
effects on the dependent media variables.

In SPSS, one-way ANOVA was speciﬁed with descriptive statistics shown.

Between groups, the F statistic and signiﬁcance are also reported.

49

CHAPTER 4

RESULTS

This chapter presents the results of the two areas of inquiry. First, the analyses
performed to assess reliability and validity of the involvement battery items are
presented. This includes ﬁndings from the SPSS reliability analysis and the various factor
analyses. These analyses are used to assess what items purported to be involvement
variables in the original research survey actually do measure involvement given the
literature on the topic. Once these analyses are performed and there is conﬁdence
regarding which variables measure involvement, the research question regarding
involvement’s role in media usage can be addressed. This is done by regressing the “true”
involvement variables on media usage variables. These variables include traditional
media such as newspaper and magazine readership in addition to Internet usage (see

Table 4.1 for these variables).

50

Table 4.1

Media Usage (Dependent) Variables

 

 

Variable Variable
number wording
Q12. About how many newspaper articles about auto or truck racing, if any,

have you read in the past 3 months?

Q13. About how many auto or truck racing magazines or newspapers, if any,
have you read in the past 3 months?

Q14. How many auto or truck races, if any, have you listened to on the radio in
the past 5 months?

Q15. How many auto or truck races, if any, have you watched on TV in the past
5 months?
Q16. How many times have you used the Internet speciﬁcally to ﬁnd

information about auto or truck racing in the past 5 months?

Q18. In the past 3 months, how many times have you watched cable TV shows
about auto or truck racing, such as SpeedWeek on ESPN?

 

Reliability Analysis
The ﬁrst part of the analysis involved measuring the reliability of the scale that
summarized the 30 items in the survey that were supposed to measure involvement. The
variables from the original questionnaire are presented in Table 4.2. This table also

includes the mean and standard deviations for each variable.

51

Table 4.2

 

 

Original Involvement Variables

How much do you agree/disagree? Mean Standard
deviation

Q30a. The only auto races that I watch are the major events like

the Indy 500. 3.59 1.55

Q30h. I like to go to local auto and truck races in my area. 4.12 1.35

Q30c. I spend a lot of time watching auto or truck races on TV. 4.03 1.38

Q30d. I like to keep up-to-date with the careers of certain drivers. 3.93 1.44

Q30e. Knowledge that a car manufacturer gains from racing helps

it build more durable street vehicles. 2.59 1.47

Q30f. My main interest in auto racing lies more with the drivers

than with the cars. 3.58 1.43

Q30g. I know much more than average about the design and

engineering of race cars. 4.20 1.24

Q30h. I follow auto or truck racing so I can keep up with the latest

automotive technologies. 4.32 1.12

Q30i. When watching racing, I watch the drivers’ skill

and technique. 3.62 1.55

Q30j. The knowledge that a car manufacturer gains helps build

vehicles with better performance and handling. 2.43 1.47

Q30k. I like auto and truck racing events for the party atmosphere. 4.22 1.18

Q301. The best thing about going to auto or truck racing events

is having a good time with my friends. 3.55 1.54

Q30m. I love the sound of roaring engines. 3.63 1.64

Q30n. The only reason I would go to an auto or truck race is if a

friend or family member wanted to go. 3.02 1.60

Q30o. The knowledge that a car manufacturer gains helps build

vehicles that are fun to drive. 2.69 1.49

52

Q30p. I often fantasize about being a race car driver.
Q30q. My friends and I often talk about auto or truck racing.
Q30r. I often videotape auto or truck races on TV.

Q3OS. One of the best things about watching auto or truck racing
is the crashes.

Q30t. Auto and truck racing are great entertainment for the
whole family.

Q30u. I often buy products or services from companies that
sponsor race teams.

Q30v. The knowledge that a car manufacturer gains helps build
vehicles that are sporty.

Q30w. I pay a lot of attention to the brands of cars and trucks
that participate in auto or truck racing.

Q30x. The brands of cars and trucks that win racing events are
also the best ones on the road.

Q30y. If a brand that I was thinking about buying won a racing
event, it would increase my interest.

Q302. The knowledge that a car manufacturer gains helps build
safer street vehicles for me.

Q30aa. Compared to others, I have a real passion about the
vehicles I drive.

Q30ab. I like to think of ways to make my vehicle higher
performance by buying performance parts or kits.

Q30ac. I View my own vehicle primarily as just a means
of transportation.

Q30ad. I enjoy doing work on my car to enhance its performance.

4.35

4.07

4.58

4.03

3.42

3.50

2.69

3.89

3.90

4.14

2.43

2.96

4.26

2.46

4.36

Table 4.2

1.23
1.38

.95

1.39

1.56

1.54

1 .48

1.46

1.31

1.26

1.43

1.57

1.23

1.53

1.19

 

Note: Valid N (listwise) = 1380

53

Reliability was assessed using the SPSS reliability method. This procedure
includes a “scale if delet ” option that allows the researcher to View the corresponding
alpha when a scale item is deleted, which provides a better picture of how each item
“ﬁts” with the other items in the scale. This analysis allows measurement of the extent to
which reported responses correlate highly with each other. This is not, however, a
measure of unidimensionality. Instead it is a measurement of the level of mean
intercorrelation of the data. This is in contrast to the standardized item alpha. The results

of the reliability analysis are discussed in the paragraphs that follow and are presented in

 

 

Table 4.3.
Table 4.3
Reliability Analysis
Item Scale mean Scale variance Corrected item Alpha if
item deleted if item deleted total correlation item
deleted
Q30a. 102.7700 593.7142 .3163 .9269
Q30h. 102.2400 581.9014 .5831 .9233
Q30c. 102.4700 576.5951 .6034 .9229
Q30d. 102.6400 574.2529 .6054 .9228
Q30e. 103.8500 582.7146 .5146 .9241
Q30i. 102.9800 580.0602 .5381 .9238
Q30g. 102.2200 584.01 17 .5762 .9234
Q30h. 102.1400 592.9903 .4729 .9247
Q30i. 102.9800 562.6865 .7088 .9212
Q30j. 103.8500 575.2803 .5801 .9232
Q30k. 102.1800 591.4218 .4899 .9245
Q301. 102.9500 572.5530 .6022 .9229
Q30m. 102.9800 562.2420 .6818 .9216
Q3011. 103.1900 609.8524 .1059 .9298
Q30o. 103.6200 567.3693 .6991 .9215
Q30p. 102.1 100 585.8363 .5102 .9241
Q30q. 102.4700 574.1 102 .6306 .9225
Q30r. 101.9100 591.4767 .5078 .9243
Q30s. 102.3100 592.2565 .3884 .9257

54

Table 4.3 (cont’d)

Q30t. 103.0500 570.8157 .6323 .9224
Q30u. 102.8700 564.4981 .6888 .9215
Q30V. 103.9200 575.7309 .5558 .9235
Q30w. 102.7600 556.1236 .7919 .9200
Q30x. 102.6500 587.9066 .5031 .9243
Q30y. 102.3100 590.1151 .4134 .9254
Q302. 103.9600 584.3620 .4858 .9245
Q30aa. 103.550 575.5429 .5425 .9237
Q30ab. 102.1300 587.9324 .4697 .9246
Q30ac. 103.8400 620.0752 -.0223 .9312
Q30ad. 102. 2100 580.7127 .5685 .9234

 

Descriptive scale statistics were speciﬁed, which allowed for the mean, standard
deviation, and alpha for each item to be shown. In addition to reporting the overall alpha,
the item deletion alpha was reported as well. This gives a better feel for what items ﬁt in
the battery of items tested. That is, when the item speciﬁed is taken out of the scale, how
is the overall reliability of the scale affected?

Cases were included to ﬁlter 100 randomly from the ﬁrst 500 cases. This partial
sample analysis allows reliability testing on a smaller subset of the sample. This allows
for analysis on part of the sample that will be used for ﬁnal analysis (in this case,
regression) with “essential variation” captured.

Before running the analysis, an alpha of .75 was chosen to be the cut-off point at
which higher than .75 would be deemed acceptable alphas (for the scale overall and for
items individually). In the social sciences, an acceptable level for the reported alpha is

.70,8 although some researchers report levels of anywhere from .60 (relatively liberal) to

 

8 That .70 is as low as one may wish to go is reﬂected in the fact that when alpha is .70, the

standard error of measurement will be over half (0.55) a standard deviation.

55

.80 (rather stringent). The a priori cut-off point for the alpha levels was determined to be
.75 for the 30 items used in this analysis. This was determined to be neither too stringent
nor too liberal for a battery of items that had not been previously tested for reliability in
any way.

The overall scale alpha was .9263, which signiﬁes that the scale of 30 items has
high internal consistency. As Schmitt (1996) noted, internal consistency is not the same
as unidimensionality. Schmitt posited that internal consistency can be thought of as
interrelatedness, while unidimensionality can be thought of as homogeneity. Obviously,
one needs internal consistency for unidimensionality, but the idea of consistency among
variables is not sufﬁcient to deﬁne unidimensional constructs (Schmitt, 1996). This
coefﬁcient of reliability will be low when items have a multidimensional structure. That
is, there are many different concepts or ideas being measured with the variety of
variables. The alpha will be high (approaching 1) when there is higher inter-item
correlation.

It is interesting to note that alpha will rise as the number of items input increases.
However, this is not a concern in the present study because it only becomes truly
problematic when the number of items nears the sample size. In this instance, the number
of items used to measure reliability was 30—nowhere near the 1,380 sample size used for
analysis.

Alphas remain fairly consistent for the scale even with items deleted. Perhaps the
most notable contributing variable (when deleted) is variable 30ac (“I View my own

vehicle primarily as just a means of transportation”). Interestingly, this item is negatively

56

correlated (although slightly) with the total scale and the overall scale alpha is highest

when this item is deleted.

Validity Analysis

In an effort to test the validity of these variables, factor analyses were performed
to validate which variables did, in fact, measure involvement. It is important to recall
here that the 30 items from the questionnaire used for this analysis were developed
intuitively to measure respondents’ involvement with motorsports. All of the items were
included based on brainstorming sessions and areas of interest to the end-user or client.
That is, no a priori testing of any of the variables was done to assess if they did, in fact,
measure involvement. Put another way, the research sponsors basically formed the ‘
variables by asking themselves what things people said, did, or felt to show they were
involved in motorsports.

Validity testing was done in a series of phases to best understand the underlying
structure of the scale items and their validity. First, an exploratory analysis was
performed using all 30 of the involvement items as they appeared on the survey (see
Table 4.2). This exploratory analysis (principal components extraction) extracted factors
and reported those with eigenvalues over 1 (SPSS default). Varimax rotation was
performed to minimize the likelihood that items would load highly on more than one

factor. This also simpliﬁed the interpretation of the factors. Cases were excluded listwise

57

so as to not include missing values in the analysis. In addition, KMO and Bartlett’s test of
sphericity9 were measured and reported.

From this analysis, the factor loadings were analyzed. Upon initial examination, it
appeared that there were three distinct factors, a fourth rather nebulous factor, and
another very poorly deﬁned factor. These results are shown in Table 4.4. For example,
the ﬁrst factor had loadings from .586 to .848, the second factor had loadings ranging
from .665 to .820, and so forth. While the cut-off point for variables’ loadings to be
attributed to a factor were set at .500, a closer inspection of the data revealed that some of
these loadings, while at ﬁrst appearing high and contributing only to one factor, were
clearly shared across factors (e. g., having a loading of .586 on Factor 1 and .389 on

Factor 3, as variable 30m does).

 

9 KMO measures the sampling adequacy via partial correlation among the variables. Bartlett’s test

of sphericity tests the appropriateness of the factor model.

58

Table 4.4

Exploratory Factor Analysis Results

 

 

Variable Factor (loadings)

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5
30a .124 .171 6.714E-03 .452 8.215E-02
30b .613 .162 .226 .207 -9.054E-02
30c .848 .1 l l .152 8.89OE-02 2.995E-03
30d .767 .165 .1 1 1 .172 4.778E-O2
30c . 196 .754 8.394E-02 6.503E-02 1.782E-02
30f .478 .135 -3.168E-02 .408 . 165
30g .594 .170 .403 3.874E-02 3.109E.02
30h .555 .179 .415 .102 7.683E-02
30i .703 .219 .185 .258 -3.253E-02
30j .148 .820 8.007E-02 .145 -3.372E-02
30k .348 .104 .230 .504 —2.605E-02
301 .370 .159 .208 .620 -.207
30m .586 .205 .389 .262 -. 182
30n .217 .163 -3.984E-02 .459 .110
300 .184 .729 .154 .215 1.691E-02
30p .491 .138 .424 3.55E-02 -4.440E02
30q .740 .154 .275 7.148E-02 7.279E-04
30r .559 8.659E-02 .244 6.901E-02 .127
30s . 165 6.670E-02 8.275E-02 .467 5.069E-02
30t .504 .308 .101 .413 -7.214E-02
30u .487 .329 .246 .260 4.804E-02
30v .1 19 .665 .143 .208 8.826E-02
30w .639 .288 .351 .167 . 184
30x .323 .362 9.982E-02 .237 .401
30y .309 .280 .343 .225 .353
302 .161 .745 6.914E-02 .1 17 7.068E-O2
30aa .270 .292 .395 .188 -5.964E-02
30ab .242 7.07OE-02 .675 .129 8.724E-02
30ac .148 -l .678E-O2 -.334 5.807E-02 .285
30ad .289 .147 .703 7 . 133E-02 2.577E-02

 

59

Table 4.5 shows further relevant results of this exploratory factor analysis, which
helped to assess just how many factors were usable in this analysis. This analysis had a
KMO measure of sampling adequacy of .95. This number can be thought of as very good
if the goal is to assess whether the variables share a common factor with other variables

(i.e., their partial correlation will be small, indicating the unique variance they share).

Table 4.5

Exploratory Factor Analysis (partial results)

 

 

Initial eigenvalues Extraction sums of squared

loadings
Factor Total % Cumulative Total % Cumulative
variance % variance %
1 11.153 37.176 37.176 10.717 35.722 35.722
2 2.703 9.011 46.188 2.278 7.594 43.316
3 1.710 5.699 51.887 1.158 3.861 47.177
4 1.341 4.470 56.358 .856 2.852 50.029
5 1.082 3.605 59.963 .545 1.816 51.845
6 .92 1 3.070 63.033

 

Table 4.5 shows eigenvalues and variance for Factors 1 through 6 only, to
illustrate how the value of the eigenvalues (and percent of variance explained) gradually
decreases and then drops below 1.0 after Factor 5. In fact, there does not appear to be
much more variance explained as one adds Factors 4 and 5. So, upon examination of this
information, there are ostensibly four really useful factors. This information needs to be
used in conjunction with the factor loadings (see Table 4.4) to assess which factors can be

useful.

Upon closer inspection of the items and their loadings in the rotated factor
analysis (see Table 4.4) the only factors that produce clear, highly loading items are
Factors 1, 2, and 4. Only two items have loadings above .500 on Factor 3, and no items
have high loadings on Factor 5. Factor loadings were again examined, and items were
included in a factor if the loading was higher than .500 and also if there were no other
relatively high loadings on other factors. For example, item 30h loads highly (.555) on
Factor 1 but also relatively highly (.415) on Factor 3. Item 30h (“I follow auto or truck
racing so 1 can keep up with the latest automotive technologies”) contributes to the
composition of these two factors.

Factor 1 has the highest number of items loading (at .500 or higher with no
“shared” relative loadings) on it. Factor 1 items clearly have to do with consumers’
involvement in auto racing events, their sponsors, or drivers. In addition, most of them
imply some action on the part of the respondent (e.g., “I like to go to local auto and truck
races in my area,” “My friends and I often talk about auto or truck racing,” and “I like to
keep up to date with the careers of certain drivers”).

These items appear to be about involvement as described in the academic
literature reviewed in chapter 2 and were named General Involvement. However, as was
noted in the literature, there are many facets to the construct of involvement, even with
the items associated with Factor 1. For example, there are aspects of investment and
socialization, as well as indirect and direct participation in the viewing process in the
above items.

Factor 2 had the most clear high loadings (very few shared loadings). In fact, the

lowest loading on the factor was .665. The items that loaded highly on Factor 2 are:

61

I The knowledge that a car manufacturer gains from auto racing helps it build more
durable vehicles.

I The knowledge that a car manufacturer gains from auto racing helps it build vehicles
with better performance and handling.

I The knowledge that a car manufacturer gains from auto racing helps it build vehicles
that are more fun to drive.

I The knowledge that a car manufacturer gains from auto racing helps it build safer
street vehicles for me.

These items do not represent the respondents’ involvement, but rather their
perceptions of the consequences of the car manufacturer’s involvement in auto racing.
Unlike the items in Factor 1, there is not direct action or participation on the part of the
respondents for these items. The respondent is more of a third-party beneﬁciary of the
actions taken in these items.

These items represent technology transfer, the notion that professional
performance of a manufacturer’s product helps make it better for the consumer.
Technology transfer is important to manufacturers. If, for example, only a minority of
consumers believed that professional auto racing helped build better cars, then there
would be little payback on the investment of sponsoring such vehicles. While getting the
manufacturer’s name out in front of the buying public is important, some in the industry
believe that technology transfer is the most important reason for allowing professional
use of a consumer goods product in motorsports.

Upon examination of the literature and deﬁnitions of involvement, these items

were thought to not capture the involvement concept. Rather than relating to how

62

consumers are interested in auto racing, these items measure how consumers feel about
auto manufacturers’ involvement. When the factor analysis was done forcing the results
to three or four factors, the technology transfer items still clearly fall under one factor.

The third clear factor from the original factor analysis includes items that revolve
around social issues related to racing (e.g., “I like auto and truck racing events for the
party atmosphere” and “The best thing about going to auto or truck racing events is
having a good time with my friends”). These items also illustrate involvement but in a
different manner than Factor 1. While Factor 1 illustrates involvement with the event or
racing itself, Factor 3 illustrates involvement with others (friends, family), with the race
being the facilitator of involvement. Thus, this factor is labeled social involvement.

The next phase of factor analysis was performed using only those items that fell
into Factor 1 and was done using the entire sample. While these items appeared to best
capture the idea of involvement, further analysis was performed to be certain that these
variables did, indeed, fall under one factor. Thirteen variables were put into this analysis.
This included the eight highly loading and “unique” (not shared loadings) variables from
the original factor analysis, "in addition to those variables that shared loadings with
another factor.

When this analysis was performed, the resultant KMO measure of sampling
adequacy was .94 and, while two factors were shown, many loadings on the rotated factor
matrix were often shared between the factors. Six variables were thought to best measure
involvement in the racing (versus, for example, involvement with working on one’s

vehicle). These results are shown in Table 4.6.

63

Table 4.6

Final Variables Used to Measure Involvement

 

 

Variable Description Factor loading_
30b I like to go to local auto and truck races in my area. ‘ .631
30c I spend a lot of time watching auto or truck races on TV. .820
30d I like to keep up-tO-date with the careers of certain drivers. .804
30i When watching auto and truck racing, 1 watch the drivers’ skill

and technique. .734
30q My friends and I often talk about auto or truck racing. .743
30w 1 pay a lot of attention to the brands of cars and trucks that participate

in auto and truck racing. .679

 

A reliability analysis was performed on these “ﬁnal” variables to be certain that,
as a scale, they represented a reliable measure of the concept. Table 4.7 provides the
details of this analysis. However, the most important output of this analysis is the overall

alpha of .9053. This indicates that the scale of six items has internal consistency.

Table 4.7

Reliability Analysis of Final Involvement Items

 

Item Scale mean Scale variance Corrected item— Alpha
if item if item total if item
deleted deleted correlation deleted

30b 19.5420 37.4000 .6487 .901 1

30c 19.6319 34.8869 .8070 .8788

30d 19.7406 34.8928 .7605 .8853

30i 20.0478 33.9832 .7480 .8876

30q 19.5971 35.6664 .7533 .8865

30w 19.7739 35.2295 .7242 .8908

 

Note: Cases = 1380.0 N of Items = 6 Alpha = .9053

64

These variables were used to create a composite variable called NEWINVOL.10
The variables that, in the exploratory analysis, measured involvement with the
atmosphere of racing (e.g., “The best thing about going to auto or truck racing events is
having a good time with my friends”) were not included. The thought in excluding these
variables was to keep NEWINVOL truly a measurement of involvement with the sport
itself and not of the surroundings.

Research Questions (RQ) 1 to 4 were posed to validate the ad hoc involvement
scale put forth in the original national survey. The ﬁndings above clearly answer these
research questions. In particular, RQl sought to determine if any of the original 30 items
from the scale did, in fact, measure involvement. Clearly, not all of the items included
measured involvement. However, a core of six items appear to measure involvement.
These items (RQ2 and RQ3) are those that imply active viewing, discussing, and
listening. In addition, the ﬁnal involvement items imply motivation and relevance in that
there is an importance about auto and truck racing that motivates respondents to watch,
talk about, and pay attention to brands raced. In addition, the admission of respondents
that they keep up to date with drivers and pay attention to the participating brands implies
relevance—these things are important to the respondents. (Ascertaining the reasons for
the relevance and motivation, however, is beyond the scope of this study.)

It would appear then, that—while less explicitly than might be desired——

involvement as it is expressed in these ﬁnal items does reﬂect involvement as it is

 

'0 This is done using the transform function in SPSS, which allows the researcher to combine

variables to create a single new variable.

65

presented in the academic literature. For example, the notion put forth by Andrews,
Durvasula, and Akhter (1990) that highly involved people will seek out inﬂuences or
communication in agreement with the object of involvement is reﬂected in the idea that
respondents spend time conversing about racing with friends. They are seeking out not
only the content but also the people who will agree with the importance of these
discussions. Perhaps Havitz and Dimanche’s (1999) discussion about how the motivation,
arousal, and interest in an activity or product affect our behavior captures the essence of
all six of the ﬁnal involvement statements. Clearly, respondents who are highly involved
are (more) motivated (titan those not involved) to watch, discuss, pay attention to, and
keep up-to-date on racing.

In addition, composite variables for social involvement (NEWSOC) and
technology transfer (NEWTECH) were created to use in predicting media usage.
Reliability and validity analyses of these two composite variables were performed and are

reported below.

NE WTECH and NE WSOC Reliability and Validity

Reliability analysis was performed on the variables that were used to compose
both NEWTECH and NEWSOC and was performed on the entire sample. The reliability
analysis alpha for the combined four-item NEWTECH variable was .8763, indicating
internal consistency. The alpha for the items used to comprise NEWSOC was .6558,
indicating lower internal consistency. While this alpha can be considered of low or
modest reliability, Pedhazur and Pedhazur-Schmelkin (1991) note that alphas of .50 or

higher are adequate for research purposes.

66

Validity was assessed using factor analysis (principal components analysis) for
each set of variables used to comprise NEWTECH and NEWSOC. For each of the sets of
variables, the factor analysis revealed one component (factor). These are shown below
with factor loadings, the KMO measure of sampling adequacy and Bartlett’s test of
sphericity. Since the KMO ﬁgure reported below approaches 1.0 (=.822) the variables are

measuring a common factor.

 

 

Table 4.8
NE WTECH Factor Analysis
Item Variable Loading_
30c. The knowledge that a car manufacturer gains from auto racing

helps it build more durable vehicles. .854
30j. The knowledge that a car manufacturer gains from auto racing

helps it build vehicles with better performance and handling. .889
300. The knowledge that a car manufacturer gains from auto racing

helps it build vehicles that are fun to drive. .840
302. The knowledge that a car manufacturer gains from auto racing

helps it build safer street vehicles for me. .833

 

Note: KMO = .822 Bartlett’s test of sphericity signiﬁcance = .000

Factor analysis was done for the variables that resulted in the composite variable
NEWSOC. The results are reported in Table 4.9. It should be noted that the loadings for
all of the variables fall within the range outlined above as acceptable (>.500) with the
exception of 30D, whose loading is .453. This variable is still included because of
rounding considerations. Therefore, the variables shown in Table 4.9 are those used for

creation of the composite variable NEWSOC.

67

Table 4.9

 

 

NE WSOC Factor Analysis
Item Variable Loading—
30a The only auto races that I watch are the major events like the Indy

500 or the Daytona 500. .601
301 The best thing about going to auto and truck racing events is having

a good time with my friends. .798
30k I like auto and truck racing events for the party atmosphere. .759
30n The only reason I would go to an auto or truck race is if a friend

or family member wanted to go. .453
303 One of the best things about watching auto or truck racing is the

crashes. .636

 

Note: KMO = .822 Bartlett’s test of sphericity signiﬁcance = .000

Regression Analyses
The research questions put forth earlier laid the foundation for better
understanding two areas of inquiry. First, does an intuition-based set of questions
measure what it is purported to measure? This was answered above in the factor analyses,
which found that some, but not all, of the variables put forth to measure involvement do
in fact measure this concept as it is generally deﬁned in the academic literature. The next
area of interest is what, if any, effect involvement has on media usage with the speciﬁc

sample of respondents interested in motorsports.ll Media usage was measured using

 

‘1 While not all respondents were avid motorsports fans, the majority of the sample reported being
moderate to avid fans. The results of analysis discussing levels of fanship were the focal point of the

original study.

68

speciﬁc self-reported, open-ended questions for speciﬁc media (refer to Table 4.1 for a
listing of these variables). Respondents were asked to ﬁll in a number in response (e.g.,
“In the past 3 months, how many times have you watched cable TV shows about auto or
truck racing, such as Speedweek on ESPN?”).

As was the case with the involvement questions, these variables were formulated
based on the opinions and interests of the researchers, not academic literature or previous
research efforts.12 The same holds true for the time periods designated within each
question (e.g., “How many auto or truck racing magazines have you read in the past 3
months?)

Regression analysis was performed in the following manner. The independent
variables used for the regression analysis were the newly formed involvement, social
involvement, and technology transfer variables called NEWINVOL, NEWSOC, AND
NEWTECH, respectively. While the research question of interest is how does
involvement predict speciﬁc media usage, it also seemed feasible that these other
variables, which measure association with motorsports in different ways, could also
affect media usage.

Initially, all regreSsions were run using both the SPSS enter method, which tests
the full regression model with all independent variables included in the analysis.
Subsequently the stepwise method was used in the hope that it would provide insights
into which independent variable, in fact, was best in predicting the particular media-
usage-dependent variable and provide a more parsimonious model. However, upon

inspection of both the enter and stepwise methods, it was evident that there was very little

 

'2 It should also be noted that these questions were modiﬁed somewhat from the earlier study.

69

to be gained in performing the stepwise analysis. Results presented below, therefore, are
those found in the regressions that used the full model.

The ﬁndings (below) are reported in the same manner in which the data were run.
Regressions were performed by individual dependent variable. As noted previously, this
allows for better analysis of, for example, the impact of social involvement on Internet
usage versus that of newspaper usage. It was felt that, while somewhat laborious
(compared to, for example, creating a composite media usage variable), this method of
analysis creates a richer picture of this data and the impact of the independent variables
on media usage. In addition, it might provide better answers to the question about the
impact of involvement on traditional media versus the Internet. Finally, it should be noted
that the scales for the independent variables are 5-point Likert scales with 1 as agree
strongly and 5 disagree strongly. This affects the interpretation Of the ﬁndings as noted
below (for example, a seemingly positive beta indicates an actual inverse relationship

between the strength of the independent and dependent variables).

Motorsports Newspaper Article Usage”

As noted above, respondents were free to specify any number of auto and truck
racing newspaper articles they had read in the past 3 months. Upon inspection of the
responses, it was noted that there were two respondents whose answers to this question

were well outside the range of other respondents (e.g., 999 newspaper articles read). Of

 

'3 It should be noted that the discussion for the ﬁrst dependent variable, newspaper article
readership, will include an explanation of the data points used. Such a discussion will not be included in

subsequent discussion of dependent variables.

70

course, these data points could be attributed to such things as misunderstanding the

question (those respondents reporting incredibly high numbers might have misread the

instructions as “How many have you read within the past year,” for example) or data

entry error. Regardless of the cause, these responses were replaced with a value of 200,

which was the upper end of the responses of the other survey respondents. Responses

then fell between 0 and 200 newspaper articles read within the past 3 months.

Table 4.10 shows the output of interest in this analysis. Again, the independent

variables were the composite variables of NEWINVOL, NEWSOC, and NEWTECH.

Table 4.10

Summary Regression Output: Newspaper Article Readership

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
Model R R2 Adjusted R7 Standard error
of the estimate
1 .370 .137 .135 14.679
ANOVA
Model Sum of squares Df Mean square F Significance
1 46900.430 (reg) 3 15633.477 72.556 .000
2964856 (res) 1376 215.469
343386.03 (total) 1379
Coefﬁcients
l Unstandardized Standardized
coefficients coefﬁcients
B Standard Beta T Signiﬁcance
error
(Constant) 22.602 1.499 15.077 .000
NEWINVOL -5.479 .484 .409 1 1.317 .000
NEWTECH .212 .370 -.016 -.572 .567
NEWSOC .701 .527 -.O49 -1.331 .183

 

 

 

 

Predictors: (Constant), NEWSOC, NEWTECH, NEWINVOL

 

 

Dependent Variable: Q12. About how many newspaper articles about auto or truck
racing, if any, have you read in past 3 months?

 

71

 

An important ﬁrst step in looking at the data above is to assess the signiﬁcance of
the model. This is shown in the signiﬁcance ﬁgure under AN OVA. With this number
being .000, we can say with conﬁdence that the model ﬁts the data or, put another way,
that there is indeed a relationship between the independent variables and the dependent
variable.14 This tells us that the model can be accepted because of this relationship.

The R2 shown under Model Summary is the next important data point to examine.
The adjusted R2 measures the proportion of the variance in the dependent variable that
was explained by the variations in the independent variables.” Put another way, the
adjusted R2 then shows that 14% (. 135) of the variation in newspaper article readership
was explained by the model using NEWINVOL, NEWSOC, and NEWTECH. This is
helpful as it allows us to now understand that the three variables—taken as a model—do
in fact have some inﬂuence on the readership of newspaper articles related to
motorsports. Involvement, technology transfer, and social involvement, taken together,
have some predictive ability when looking at readership levels of newspaper articles
speciﬁc to motorsports. Of course, we are interested in better understanding how each of

these variables contributes to the predictability of the dependent variable.

 

‘4 With the signiﬁcance ﬁgure at .000, the model is then signiﬁcant at the 99% conﬁdence level. If
signiﬁcance had been reported below .05, then the model would be signiﬁcant at the 95% conﬁdence level.

'5 The R2 is the ratio of the explained or regression sum of squares (RSS) I total sum of squares
(TSS). The TSS explains the total deviations in the dependent variable and the RSS is the amount of the
TSS that could be explained by the model. Of course, then, the residual sum of squares is the amount of

deviation that could not be explained by the model.

72

Looking next at the coefﬁcients gives us a good idea of how each of the
independent variables contributes to the model, if at all. These coefﬁcients, in particular
the beta scores, provide information on the effect of each independent variable on the
dependent variable. However, the ﬁrst step before looking at the beta coefﬁcients and
what these mean, is to look at the signiﬁcance provided for each independent variable.
This allows understanding of the signiﬁcance of the contribution of the speciﬁc
independent variable in predicting the dependent variable. Of course, one is looking for a
signiﬁcance ﬁgure of .05 or less, indicating that the independent variable is signiﬁcant in
its predictive ability at the 95% conﬁdence level.

In this particular analysis, it is clear that the only independent variable that is
signiﬁcant in its inverse contribution is NEWINVOL. It appears highly signiﬁcant (.000
signiﬁcance). The remaining independent variables are not signiﬁcant in their predictive
capabilities.

Now that it is understood that NEWINVOL is signiﬁcant in its contribution to the
level of motorsports newspaper article readership, the next question becomes how much
does it contribute to predicting this dependent variable? Independent variables can be
looked at individually to assess each variable’s contribution to the ﬁnal score of the
dependent variable and also the model in its entirety can be assessed for its ability to
inﬂuence the dependent variable. The regression equation16 is ﬁt to the data set in an
effort to both describe the data and to assess the predictive ability of the independent

variables. Put in its simplest terms, the result of this equation tells us how the dependent

 

‘6 The regression equation is Y = b0 + lei + ei where b indicates the beta coefﬁcients and x

indicates the ith variable.

73

variable will change due to the inﬂuence of the predictor variables. This gives insight into
the relationship and the strength of that relationship.

In looking at the independent variables individually and the corresponding beta
levels for each, it becomes easier to understand the relationship that exists between these
variables and motorsports newspaper article readership. If one were to assume that each
of the independent variables was signiﬁcant in the regression equation, its contribution
could be assessed as follows. NEWINVOL is seen as positively contributing to the level
of motorsports newspaper articles read, but not highly (beta = 5.479). NEWTECH does
not contribute much at all to the independent variable (beta =.212) and also contributes
inversely. The same can be true for NEWSOC, whose beta coefﬁcient is .701, indicating
that this independent variable will inversely inﬂuence the number of motorsports
newspaper articles read. In other words, as social involvement rises, respondents are less

likely to read motorsports newspaper articles.

Motorsports Cable Usage

As was the case with reported newspaper articles read, there were a few
respondents whose responses fell outside what might be considered the norm for such a
question. For example, two respondents were shown to have viewed almost 1,000 cable
television shows about motorsports in the past 3 months. This, obviously, could have
been a result of data entry errors or misunderstanding on the respondents’ part with
regard to what the question was asking. In any event, these responses were deleted and
replaced with the maximum reported response for other respondents on this variable in

the survey (100). Once this step was taken, both stepwise and enter regressions were

74

performed. However, as noted above, only the results of the enter method of regression

are reported here. The ﬁndings from this regression are reported in Table 4.11.

Table 4.11

Summary Regression Output: Cable Viewership

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
Model R R2 Adjusted R2 Standard error
of the estimate
1 .420 .177 .175 6.786
ANOVA
Model Sum of squares Df Mean square F Signiﬁcance
1 13610.787 (reg) 3 4536.929 98.511 .000
63372.01 1 (res) 1376 46.055
76982.797 (total) 1379
Coefﬁcients
l Unstandardized Standardized
coefﬁcients coefﬁcients
B Standard Beta T Signiﬁcance
error
(Constant) 1 1.807 .693 17.036 .000
NEWINVOL -3.1 15 .224 .491 1 1.317 .000
NEWTECH .111 .171 -.018 .651 .515
NEWSOC .667 .243 -.098 2.739 .006

 

 

 

 

 

 

Predictors: (Constant), NEWSOC, NEWTECH, NEWINVOL

 

 

Dependent Variable: Q18. In the past 3 months how many times have you watched cable
TV shows about auto and truck racing, such as SpeedWeek on ESPN?

 

The model, represented by the independent variables NEWINVOL, NEWTECH,
and NEWSOC, does not perform well in explaining the variance in cable motorsports
viewership. In fact, only 18% of the variance (as noted in the R2) in the dependent
variable is explained by this model. However, the model itself is signiﬁcant or, put
another way, ﬁts the data well (as seen in the ANOVA signiﬁcance).

In looking at the independent variables as they contribute to or inﬂuence cable

motorsports Viewership, the coefﬁcients are quite telling. This variable is interesting in

75

 

that not only is NEWINVOL a contributor to the inﬂuence of cable motorsports
viewership, but NEWSOC (the social involvement or social aspects of being a
motorsports fan) is, too. Interestingly, while NEWSOC is not a huge contributor to the
prediction of cable motorsports Viewership, it is notable due to its inverse (albeit small)
effect. Thought of in its simplest terms, as NEWSOC increases, the amount of cable
motorsports programs viewed decreases. (This relationship is inverse because coding for
the newly created variables went from 1 to 5, with 1 being strongly agree and 5 being
strongly disagree.) This makes intuitive sense when one thinks of the social atmosphere
of races themselves and the notion that some people might attend these races simply for
that party atmosphere, in addition to watching the race or drivers. The implication of this
intuitive idea is that such people do not want to simply watch the race (e.g., on cable
television); they want to experience the entire atmosphere of race day—the pro-race
parties and socializing as well as the race itself. However, it is good to remember that this
variable is not a large contributor (beta = .667).

NEWIN VOL is also highly signiﬁcant and, as is the case for NEWSOC, does not
contribute greatly to the inﬂuence on cable television motorsports viewership. The
inverse relationship (beta = 3.115) suggests that as involvement with motorsports
increases, viewership of cable television shows about auto or truck racing will increase as
well. This makes sense when looking at the deﬁnition of involvement used for the present
study, which includes the desire and investment put into that which someone is
involved—perhaps those more highly involved are more willing to expend time on

television viewing of a race (perhaps even in addition to attending select races).

76

Motorsports Magazine and Newspaper Usage

Just as interest in motorsports has been increasing, so have the media choices
available with which a fan can follow the sport. For example, current magazine titles on
the topic include Auto Racing Digest, NASCAR RFT (Racing for Teens), Motor Trend,
and Stock Car Racing.

When the three predictor variables are placed in a regression model equation, the
model accounts for only 15% of the variance in racing magazine or newspaper
readership. This and all summary numbers are shown in Table 4.12. While the model
itself is signiﬁcant, it does not account for much of the variance in readership of racing
magazines and newspapers.

Upon looking at the independent variables individually, only NEWINVOL
appears signiﬁcant. While the beta coefﬁcient for NEWTECH taken by itself appears to
be a huge contributor to the inﬂuence of number of auto and truck racing magazines read,
this number cannot be deemed reliable due to its .498 signiﬁcance value. The implication
of this seemingly contradictory pair of numbers is that the beta coefﬁcient is not reliable
due to too much variance.

As with other dependent variables analyzed, NEWSOC contributes to magazine
and newspaper readership but is not a reliable predictor (signiﬁcance = .224). However,
NEWINVOL is reliable but contributes to the prediction of number of motorsports
newspaper and magazines read. This contribution is not large, though, and should not be

overstated (beta = 1.574).

77

Table 4.12

Summary Regression Output: Magazine and Newspaper Readership

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
Model R R2 Adjusted R2 Standard error
of the estimate
1 .390 .152 .150 4.159
ANOVA
Model Sum of squares Df Mean square F Sigﬂ'ﬁcance
1 4259.547(reg) 3 1419.849 82.101 .000
23796.580 (res.) 1376 17.294
28056.128 (total) 1379
Coefficients
1 Unstandardized Standardized
coefﬁcients coefﬁcients
B Standard Beta T Signiﬁcance
error
(Constant) 6.815 .425 16.047 .000
NEWINVOL -1.574 .137 .411 11.477 .000
NEWTECH -7.1 1 lE-O2 .105 .019 .678 .498
NEWSOC .181 .149 - 044 1.216 .224

 

 

 

Predictors: (Constant), NEWSOC, NEWTECH, NEWINVOL

 

 

Dependent Variable: Q13. About how many auto or truck racing magazines or
newspapers, if any, have you read in the past 3 months?

 

Motorsports Television Usage

This analysis proved interesting in a couple of ways. First, the amount of variance
explained by this model is roughly one-quarter (R2 = .278). This is higher than any of the
variance explained for the other media usage variables. The other interesting occurrence
to note is the near signiﬁcance (signiﬁcance = .067) of NEWTECH in the regression
equation for television viewership of auto or truck races (see Table 4.13). The implication
of this is that with some reliability, NEWTECH might be thought to contribute inversely
to the number of auto and truck races a fan watches on television. That is, if signiﬁcant,

as interest in technology transfer increases, motorsports television usage decreases. While

78

 

this is interesting and different from the contribution of NEWTECH to other media usage

variables, it should be noted that the inﬂuence is very small (beta = .309).

Table 4.13

Summary Regression Output: Television Viewership

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
Model R R2 Adjusted R2 Standard error
of the estimate
1 .527 .278 .276 6.689
ANOVA
Model Sum of squares Df Mean square F Signiﬁcance
1 23641.395 (reg) 3 7883.798 176.187 .000
61571446 (res.) 1376 44.747
85222.84] (total) 1379
Coefﬁcients
1 Unstandardized Standardized
coefﬁcients coefﬁcients
B Standard Beta T Signiﬁcance
error
(Constant) 17.226 .683 25.216 .000
NEWINVOL -3.472 .221 .520 15.736 .000
NEWTECH .309 .169 -.048 -1.833 .067
NEWSOC -.261 .240 .036 1.087 .277

 

 

 

 

 

 

Predictors: (Constant), NEWSOC, NEWTECH, NEWINVOL

 

 

Dependent Variable: Q15. How many auto or truck races, if any, have you watched on
TV in the past 5 months?

 

Aside from the implications of NEWTECH’S contribution, again it is notable that
NEWINVOL contributes in the regression equation for television viewership of auto and
truck races. As with previous analyses, this contribution is rather small, but noteworthy,

especially in a relative sense.

79

 

Motorsports Internet Usage

Internet usage was assessed by having respondents report the number of times
they accessed the Internet speciﬁcally to ﬁnd information about racing. One speciﬁc
research question was posed with the thought that the Internet is easy to access and
available to many, not to mention new and interesting to use. This being the case, it was
felt that even those who were not avid fans of motorsports might log on to the Internet to
get information. Indeed, logging on to the Internet to get race results, information about
drivers, or NASCAR products, for example, might not be something that is contingent on
someone’s involvement level with the sport.

Table 4.14 summarizes the ﬁndings of the regression of NEWINVOL,
N EWTECH, and NEWSOC on the Internet usage dependent variable. As was the case
with television viewership as the dependent variable, this analysis yielded some

interesting results.

80

Table 4.14

Summary Regression Output: Internet Usage

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
Model R R2 Adjusted R2 Standard error
of the estimate
1 .291 .085 .083 8.081
ANOVA
Model Sum of squares Df Mean square F Signiﬁcance
1 8316.211 (reg) 3 2772.070 42.451 .000
89854354 (res.) 1376 65.301
98170.565 (total) 1379
Coefﬁcients
1 Unstandardized Standardized
coefﬁcients coefﬁcients
B Standard Beta T Signiﬁcance
error
(Constant) 9.007 .825 10.915 .000
NEWINVOL -2.381 .267 .332 8.932 .000
NEWTECH -1.304E-04 .204 .000 .006 .995
NEWSOC .487 .290 —.O63 -1.681 .093

 

 

 

Predictors: (Constant), NEWSOC, NEWTECH, NEWINVOL

 

 

Dependent Variable: Q16. How many times have you used the Internet speciﬁcally to
ﬁnd information about auto or truck racing in the past 5 months?

 

What is interesting is the small percentage of variance that this model accounts
for when looking at Internet usage for auto or truck racing information. Obviously, [there
are other factors not tested here that contribute more to the variance of this dependent
variable.

However, it is interesting to note that the model itself is signiﬁcant and that
NEWSOC, while contributing very little to the regression equation by way of beta (beta =
-.487), has a signiﬁcance which, while not proving reliability of this variable, hints that
the social aspect could, perhaps be reliable and predictive of involvement. If this

composite variable were made up of different social involvement variables (what those

81

 

might be are beyond the scope of this section), perhaps this signiﬁcance would be below
.05. Of course, this is speculation and food for future thought.

Once again, NEWINVOL was shown to be a signiﬁcant contributor in the
regression equation. While its contribution is not large (beta = -2.381), it appears to be
the only variable analyzed that contributes signiﬁcantly to the number of times a person
uses the Internet to ﬁnd auto or truck racing information. As was the case with the
previously reported media usage variables, it appears that as involvement increases,
media usage also increases. Taken in the context of the questionnaire, as respondents
reported less agreement with involvement variables (e.g., “My friends and I talk a lot

about auto or truck racking”), Internet usage increased.

Motorsports Radio Usage

As shown in Table 4.15, the model that includes the predictor variables
NEWINVOL, NEWTECH, and NEWSOC explains only 12% of the variance in the
dependent variable radio usage. However, the model itself does work to explain
deviations in the dependent variable (as seen in the ANOVA signiﬁcance of .000). As
noted for other dependent variable analyses, the model does ﬁt the data given this
signiﬁcance ﬁgure.

When one looks at the speciﬁc independent variables used for this analysis, it
becomes apparent that NEWINVOL is signiﬁcant in its contribution to the explanation of
radio listenership of motorsports. However, NEWSOC is also a signiﬁcant inverse
contributing variable in the regression equation. As has been the case in previous media

usage variable analyses, this contribution is quite small. However, it is interesting that the

82

social involvement (NEWSOC) variable has a negative (or inverse) impact on radio

listenership. Perhaps, as might be the case with cable television, motorsports radio

listening is not seen as an ingredient in social events centered around motorsports.

The notion of technology transfer captured in NEWTECH is not signiﬁcant in its

contribution to the prediction of motorsports radio listenership. This is evidenced by the

.564 signiﬁcance ﬁgure.

Table 4.15

Summary Regression Output: Auto and Truck Racing Radio Listenership

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model Summary
Model R R2 Adjusted Rz Standard error
of the estimate
1 .348 .121 .119 3.14
ANOVA
Model Sum of squares Df Mean square F Signiﬁcance
1 1869.901 (reg) 3 623.300 63.126 .000
13586.473 (res.) 1376 9.874
15456.374 (total) 1379
Coefﬁcients
1 Unstandardized Standardized
coefﬁcients coefﬁcients
B Standard Beta T Signiﬁcance
error
(Constant) 4.393 .321 13.690 .000
NEWINVO -1.124 .104 .396 10.848 .000
L -4.569E-04 .079 .017 .577 .564
NEWTECH .259 .1 13 -.084 -2.294 .022
NEWSOC

 

 

 

 

 

 

 

Dependent Variable: Q14. How auto or truck races, if any, have you listened to on the
radio in the past 5 months?

 

83

 

Demographic Dependent Variables

Involvement, technology transfer and social involvement do not appear to be
Signiﬁcantly large contributors to media usage. This will be discussed at length, along
with the implications of the ﬁndings, in chapter 5.

Given that these variables do not explain the variance in media usage that was
perhaps hoped for, demographic variables were analyzed to assess what impact these had.
In particular, demographic variables were used in an effort to see if these better explain
the variance in media usage variables. However, the demographic variable of state of
residence was not included due to recoding done by the original research supplier, which
cannot be changed.17 Speciﬁcally, then, the demographic variables used for analysis were
age, employment status, household income, occupation, and gender.

On initial analysis of the responses to the demographic variables, it was noted that
respondents completed most of these demographic questions. This is not always the case
in questionnaire research—especially with an instrument roughly 15 pages lOng
(approximately 20 to 25 minutes over the telephone). Respondents get fatigued from
having answered previous questions, and might not feel comfortable giving this
somewhat personal information (especially income, for example, for fear that it might

somehow be linked to the person’s name). So, of course, it was to be expected that, while

 

‘7 In addition, the responses to “In which state do you live?” were recoded to account for regions
of the country in which respondents reside. This recoding only took into account East, Central, and West

regions, and left out the Southern region—a region of the United States where motorsports is especially

popular.

84

respondents did supply answers most of the time, often there would be refusals.18 These
refusals were changed in the coding schemata to reﬂect missing data, so as not to confuse
the ﬁndings of the subsequent analysis. The ﬁndings of the ANOVA are presented and

discussed below by independent variable.

Occupation and Media Usage

Occupation appears to have a signiﬁcant effect on certain motorsports media
usage, as shown in the summary Table 4.16. There are signiﬁcant differences between
occupational groups about how many specialty (auto or truck racing) magazines they read
and how many races they watched on television. Speciﬁcally, upon examination of the
means, those whose occupation is machine operator are more likely to read a higher
number of specialty racing magazines and newspapers, and public safety personnel
appear more likely to watch signiﬁcantly more motorsports television than other
occupations listed. For both media, those who report their occupation as farming partake
in the least of these media. Interestingly, those whose occupation is farming appear to
consume the least of any motorsports media with the exception of newspaper articles, the

medium for which technical support personnel report the lowest consumption.

 

‘8 For demographic variables, refusals occurred for occupation (371 cases) and household income

(288 cases).

85

Table 4.16

 

 

Occupation and Media Usage
Mean Between group statistics
high, low & total
Mean F Sig.
square
Q12. Newspaper articles Executive (mean=6.7) 293.205 .960 .472
Tech support (mean=2)
Total (mean=4.59)
Q13. Specialty Machine operator (mean=3.9) 92.123 3.471 .000
magazines Farming (mean=.13)
& newspapers Total (mean=1.25)
Q14. Radio Machine operator (mean=1.7) 14.224 1.030 .414

Farming (mean=0)
Total (mean=.89)

Q15. TV Public safety (mean=10.22) 207.625 3.058 .001
Farming (mean=.25)
Total (mean=3.51)

Q16. Internet Executive (mean=2.82) 62.452 1.358 .203
Farming & public safety
(both mean=0)
Total (mean=1.38)

Q18. Cable TV Machine operator (mean=3.33) 77.797 1.562 .122
Farming (mean=.13)
Total (mean=2.23)

 

Gender and Media Usage

With the exception of radio, there appear to be signiﬁcant differences (see Table
4.17) between men and women in their consumption of motorsports media as measured
in the present study. In each of these media types (newspaper articles, specialty
magazines and newspapers, television, Internet, and cable television), men report a higher

average of times using these media to gather motorsports information. The highest

86

average mean for males was reported for newspaper articles (6.65), whereas women had
a mean of 1.95 for newspaper article readership. Consumption of radio for motorsports
usage was quite low for both groups, with both groups reporting an average of less than

one incidence of listening to motorsports radio programming in the past 5 months.

 

 

Table 4.17
Gender and Media Usage
Mean Between group statistics
high, low & total
Mean F Sig.
square
Q12. Newspaper articles Male (mean=6.65) 7586.834 31.134 .000
Female (mean=1.95)
Total (mean=4. 12)
Q13. Specialty Male (mean=1.75) 516.243 25.831 .000
magazines Female (mean=.53)
& newspapers Total (mean=1.09)
Q14. Radio Male (mean=.91) 15.269 1.363 .243
Female (mean=.70)
Total (mean=.80)
Q15. TV Male (mean=4.07) 597.785 9.734 .002
Female (mean=.2.75)
Total (mean=3.36)
Q16. Internet Male (mean=2.69) 1943.317 27.829 .000
Female (mean=.31)
Total (mean=1 .41)
Q18. Cable TV Male (mean=3.76) 2625.028 48.647 .000

Female (mean=.99)
Total (mean=2.27)

 

87

Income and Media Usage

Table 4.18 shows the summary AN OVA results of inﬂuence of income on media

usage. The analysis of media usage when income is considered is an interesting one.

While there appear, on the surface, to be great differences between income groups with

regard to media consumed (e.g., the average specialty magazines consumed in the past 3

months for those whose income is under $20,000 is 0, while the average for those whose

income is between $40,000 and $59,999 is roughly 7) the only differences in

consumption of television and cable television are signiﬁcantly different. Another

interesting observation is the mean of 0 for those earning under $20,000 for all

motorsports media measured.

Table 4.18

Income and Media Usage

 

 

Mean Between group statistics
high, low & total
Mean F Sig.
square
Q12. Newspaper Under $20K (mean=0)
articles $60,000 to $79,999 (mean=5.48)
Total (mean=4.55) 258.934 .852 .530
Q13. Specialty Under $20K (mean=0)
magazines $40,000 to $59,999 (mean=6.93)
& newspapers Total (mean=4.89) 31.215 1.307 .251
Q14. Radio Under $20K (mean=0)
$60,000 to $79,999 (mean=5.48)
Total (mean=4.55) 21.904 1.759 .104
Q15. TV Under $20K (mean=.84)

$100,000 to $150,000 (mean=11.73)

Total (mean=8.3 1)

88

150.499

2.196

.041

Table 4.18 (cont’d)

Q16. Internet Under $20K (mean=0)
$60,000 to $79,999 (mean=14.79)
Total (mean=9.18) 148.778 1.774 .101

Q18. Cable TV Under $20K (mean=0)
$80,000 to $99,999 (mean=11.30)
Total (mean=7.37) 123.630 2.289 .033

 

Employment Status and Media Usage

The examination of employment status and motorsports media usage, shown in
Table 4.19, illustrates some of the intense media usage among respondents. In particular,
Internet and cable television usage for motorsports events appear to be quite high for
those unemployed (Internet average = 29 Internet “hits” over the past 5 months and
slightly more than 19 cable television motorsports shows viewed in the past 3 months).
However, aside from the sheer volume of some motorsports media being consumed, there
appear to be signiﬁcant differences between the employment groups with regard to
motorsports media consumed. With the exception of newspaper articles and radio
programming, there are signiﬁcant differences based on employment status when looking
at motorsports media consumption. It appears that students tend to consume the least

amount of motorsports media and the unemployed, in most instances, consume the most.

89

Table 4.19

Employment Status and Media Usage

 

 

Mean Between group statistics
high, low & total
Mean F Sig.
square
Q12. Newspaper Full time (mean=4.96) 425.134 1.712 .129
articles Student (mean=.25)

Total (mean=4.12)

Q13.Specialty Not employed (mean=2.33) 54.451 2.693 .020
magazines Student (mean=.00)
& newspapers Total (mean=1.09)
Q14. Radio Full time (mean=.94) 15.226 1.360 .237

Student (mean=.00)
Total (mean=.80)

Q15. TV Not employed (mean=10.67) 187.439 3.056 .010
Student (mean=2.25)
Total (mean=3.36)

Q16. Internet Not employed (mean=29.00) 1949.772 30.298 .000
Student (mean=.13)
Total (mean=1.41)

Q18. Cable TV Not employed (mean=19.17) 754.223 14.155 .000
Part time (mean=.95)
Total (mean=2.27)

 

Age and Media Usage

There appear to be signiﬁcant differences between age groupings in the
consumption of motorsports media, as shown in Table 4.20, with the exception of
television. Also, with the exception of television, the age group 35 to 44 appears to be the

highest consumer of motorsports media—especially newspaper articles (mean = 7.78).

90

Table 4.20

 

 

Age and Media Usage
Mean Between group statistics
high, low & total
Mean F Sig.
square
Q12. Newspaper 35 to 44 (mean=7.78) 1103.787 4.500 .000
articles 75+ (mean=1.37)
Total (mean=4.59)
Q13.Specialty 35 to 44 (mean=2.06) 89.448 4.463 .000
magazines 75+ (mean=.11)
& newspapers Total (mean=1.09)
Q14. Radio 35 to 44 (mean=1.33) 28.772 2.585 .017
55 to 64 (mean=.33)
Total (mean=.80)
Q15. TV 45 to 54 (mean=3.82) 94.025 1.525 .166
75+ (mean=.74)
Total (mean=3.36)
Q16. Internet 35 to 44 (mean=2.91) 213.138 3.020 ‘ .006
75+ and 65 to 74 (mean=.00)
Total (mean=1.09)
Q18. Cable TV 25 to 34 (mean=10.51) 108.247 1.947 .070

75+ (mean=.1 1)
Total (mean=2.27)

 

Chapter 5 presents a further discussion of these ﬁndings, along with implications
and potential shortcomings of the study. In addition, suggestions for future research are

presented.

91

CHAPTER 5

DISCUSSION AND CONCLUSIONS

Overview of Findings and Implications

This Study was conducted to investigate two areas of interest. The ﬁrst area
explored the reliability and validity Of an involvement scale developed by an advertising
agency. Essentially, the question was whether such an intuitive based grouping of
questions measures what it is purported to measure.

The second area of inquiry was to better understand how, if at all, involvement
and demographic variables affect media usage. Using a survey developed to speciﬁcally
measure motorsports interest and motorsports-related media usage, the ﬁndings are
applicable to this sport in this instance.

The preceding chapter outlined the analyses put forth to better understand the
implications of both areas of inquiry. Through reliability and factor analysis, a better
understanding was gained of how the original scale reﬂected a variety of concepts (e. g.,
technology transfer and social involvement). These analyses aided in answering the ﬁrst
four research questions and revealed that the scale did, in fact, contain some items that
measured involvement. In particular, the ﬁnal involvement items related to the ideas of
motivation and investment put forth in the deﬁnition used for this study. Clearly, the
involvement scale could be improved on. This is discussed below.

The remaining research questions sought to better understand how involvement
(once better deﬁned) and demographics affected media usage. In particular, speciﬁc

media usage variables were examined separately to better understand how involvement

92

affected each versus how involvement and demographics affected overall media usage.
This allowed for explication of any differences that might appear, for example, between

printed media and visual media or traditional media and the Internet.

Did the “Involvement Scale” Measure Involvement?

Reliability analysis illustrated that the 30-item scale does, in fact, have some
shared meaning. Again, this does not mean that all of the items were measuring one
single concept (i.e., involvement) but that the concepts being measured were
intercorrelated. To further explore what speciﬁc concepts the 30-item scale was
measuring, factor analyses were performed. These analyses found that the major or
clearest areas being explored in this scale were technology transfer, social involvement,
and involvement. As noted in earlier chapters, it is important to reiterate that social
involvement and involvement were treated as separate concepts since social involvement
was shown to be involvement with the atmosphere or friends during a race event, while
involvement was shown as involvement with the sport itself. Thought of a different way,
social involvement was not necessarily dependent on there being a NASCAR event, but
more dependent on a gathering of friends or a party atmosphere.

In sum, then, the 30-item scale does not measure involvement per se. However,
there are some very useful items within the scale that measure involvement. In particular,
both the social involvement and involvement would interest NASCAR sponsors. Of
course, the implications of these types of involvement might be different—those more
socially involved might be less receptive tO the sponsors’ messages, for example. This

issue is beyond the scope of this study, but certainly Of interest to sports sponsors.

93

Recalling the deﬁnition of involvement used for this study, it is apparent that the
ﬁnal items used to measure involvement certainly reﬂect the deﬁnition, although perhaps
less explicitly than they could have had the deﬁnition been used to guide questionnaire
development. In particular, consider the deﬁnition and the ﬁnal involvement items.

Involvement—A product, activity, or event that is highly relevant and

meaningful and often motivates one to invest time, money, and effort in

pursuing the activity, event, or product. In particular, high involvement

creates a desire to invest more money, time, and effort in the pursuit of

keeping current with the product, activity, or event.

The ﬁnal involvement items are:
I I like to go to local auto and truck races in my area.
I I spend a lot of time watching auto or truck races on TV.
I I like to keep up-to-date with the careers of certain drivers.
I When watching auto and truck racing, I watch the drivers’ skill and technique.
I My friends and I often talk about auto or truck racing.
I I pay a lot of attention to the brands of cars and trucks that participate in auto and

truck racing.

As noted above, these items do get at the notions Of investment and motivation to
keep current with the sport, as well as hint at the relevance of the sport (e. g., through
friendships). However, to better measure involvement based on the deﬁnition put forth,

additional questions might be included to more speciﬁcally get at the idea of monetary

94

investment (e.g., “I’ll pay whatever the going price for admission to a racing event” and
“I believe that NASCAR events are worth the cost of ticket prices”) and relevance (e. g.,

“NASCAR events reﬂect part of who I am”).

Do Involvement, Social Involvement, Technology Transfer, and
Demographics Affect Media Usage

Chapter 4 outlines speciﬁcally the analyses done in an effort to better understand
the implications of certain variables on media usage. Given the speciﬁcs Of the original
study, the real question became one of whether involvement with motorsports translates
into a higher tendency to use motorsports-related media (e.g., race-related articles in
newspapers or magazines).

As noted above, concepts other than involvement were included in the original
30-item involvement scale. Given that this was the case and that there was measurement
of some demographic variables in the original study, it was decided to see if any of these
concepts had an impact on media usage.

The ﬁrst set of regressions analyzed the impact of technology transfer, social
involvement, and involvement on media usage. The results of these analyses were
presented previously in chapter 4 and are discussed below.

Involvement (with racing, not social involvement) had a small and positive
inﬂuence on media usage. Thought of another way, those who reported themselves as
highly involved were more likely to use motorsports-speciﬁc media. This was slightly
truer for newspaper, cable television, television, and Internet, perhaps because these

media (versus specialty newspapers and magazines or motorsports radio) are more

95

commonly used by consumers in general. Another reason might be that newspaper,
television (cable and broadcast), and the Internet might provide more detail than other
media, especially radio.

Interestingly, social involvement has a slight and negative impact on radio and
cable television usage. Social involvement was shown as an insigniﬁcant contributor to
media usage for the remaining media types. This makes intuitive sense because those
who desire social interaction from racing would most probably want to be at the event
itself, rather than partake of it through media.

The notion of technology transfer did not have a signiﬁcant impact on
respondents’ media usage. Perhaps, as might be the case for those seeking social
involvement, respondents who believe in the notion of technology transfer either like to
be at races in person or don’t have an interest in the sport but believe that the transfer of
racing technology to street vehicles exists—even if they do not ﬁnd the sport particularly
interesting.

Demographically, there were many ﬁndings of interest. For example, only age
inﬂuenced motorsports radio listenership. All other demographic variables did not appear
to inﬂuence this medium. The age group with the highest radio listenership for
motorsports events was 35- to 44-year-olds. Perhaps this group was more likely to grow
up with radio as entertainment or perhaps they are more likely to listen as they drive (e.g.,
to and from work).

The ﬁndings regarding gender’s inﬂuence on motorsports media were also
interesting, although perhaps not surprising to some who feel that motorsports is

appealing mostly to men. In particular, men were found to be more likely to use

96

newspaper, cable television, specialty magazines and newspapers, and the Internet to ﬁnd
information about motorsports. Certainly this could not be due to males’ greater
adeptrress at using these media or their availability to men. Perhaps the disparity is simply
one of overall interest in the sport—that is, women are less likely to use these media for
motorsports-related news because they have less interest overall in the sport. This is an
interesting supposition and worthy of tracking because motorsports are becoming more
and more appealing to women, according to some sources (Hagstrom, 1998). When
looking at speciﬁc media, it is also of interest to note that in many instances men had an
average usage of double the motorsports media usage of women. For example, the
number of motorsports’ newspaper articles read in the past 3 months was roughly seven
for men and two for women. The disparity exists for Internet usage in the past 5 months
as well, where men report an average of roughly three and women report an average far
less than one (.31).

Age’s inﬂuence on Motorsports-related media usage was seen signiﬁcantly only
for cable television and specialty magazines and newspapers. That is, as age increases,
respondents are less likely to look to these media for motorsports information. The
explanations for this could be varied. In fact, it would be interesting to see if overall
interest in the sport increases or decreases with age, which would, Of course, have
implications for media use as well. Perhaps Older motorsports enthusiasts are less likely
to spend their income on cable or specialty magazines, relying on things such as the local
daily newspapers and broadcast television for motorsports information. However, the
present analysis seems to point to an overall decrease in motorsports media use at the

higher age ranges. The Older age group (75 or older) has the lowest average use of all

97

motorsports media except radio (where the age group 55 to 64 has the lowest average
use). The explanation for this might not be due to cost at all, but to interest in the sport.
This age group might have the least amount Of interest in the sport.

A ﬁnding of great interest to motorsports marketers is the inﬂuence of income on
motorsports media. In many instances, higher income categories consume a higher
average of motorsports media. In particular, people who fell into higher income
categories (e. g., more than $60,000) had the highest average media usage for all
motorsports media. The implications of this for marketers could prove far-reaching, as
advertising for high-end products (e.g., jewelry, designer clothing, or sports cars) and
services (e. g., interior decorating or real estate services) could be placed in these
motorsports media. This data certainly counters stereotypes of motorsports fans as having
low incomes. For example, the average number of races Viewed on television by those
whose income is between $100,000 and $150,000 was 12 over the past 5 months. This is
signiﬁcantly higher than those whose income is under $20,000 (mean = .84).

The income data is interesting in light of the employment status ﬁndings, which
run counter (intuitively) to those of income. In all Of the media categories for which
employment status was a signiﬁcant differentiator, those respondents who fell into the
category of not employed had the highest average use. In fact, the average Internet use
rate for those unemployed was 29 Internet hits over the past 5 months. Cable television
for this same group averaged 19 programs viewed in 3 months. For these signiﬁcantly
different media categories, those employed full time had fewer than half of the media
viewing averages compared with the unemployed respondents. This higher rate of media

use might simply be due to time availability—that is, unemployed respondents might

98

 

have more time to spend with these media in general. In fact, these respondents might not

be more avid fans but might have more time to consume such media.

What Next? Study Implications

This study proved interesting on several fronts, but perhaps none moreso than the
implications it has for future research. Unfortunately, the variables used in the original
study did not all measure involvement as intended. If involvement were the issue of
importance, the scale used to measure the concept could have been expanded to include
explicit measures of motivation, relevance, and investment, for example. Such items
could be borrowed from scales such as Zaichowsky’s involvement scale and modiﬁed
speciﬁcally for motorsports. Of course, these would need to be tested for reliability and
validity.

Perhaps of most importance to advertising practitioners and sponsors of events
such as motorsports is the attention to and inﬂuences on media use. Further explication of
involvement and how it affects media use is obviously called for. With better
involvement measures, researchers can better understand how highly involved fans use
(or do not use) media.

However, the present study does have some ﬁndings of interest on which to base
future research. For example, if those respondents who are highly socially involved with
motorsports do not use radio or cable television sources for motorsports events or
information, are there ways to make these media appealing to them? Are there contests or
events that could be done on radio or on cable television that would somehow replicate

the social atmosphere these people seek? In other words, can the party atmosphere of the

99

race be replicated for fans who can not attend the actual race? Or, can these people be
“turned into” motorsports enthusiasts, thereby getting them to use diverse media?

The idea of technology transfer also appears to have little impact on media use.
This is an important fact for motorsports advertisers (especially auto manufacturers who
sponsor the sport and tout the idea that racing results in better street vehicles).
Technology transfer, it appears, is not a reason for people to tune into motorsports events.
Motorsports sponsors need to assess where technology transfer falls in the scheme of
marketing. That is, is it important for people to be heavy motorsports media consumers to
believe in the notion of technology transfer or is it enough for the general population to
understand and believe it?

Demographically, a better picture of who uses motorsports media is needed.
Traditional stereotypes would say young, lower income males are the primary consumers
Of any motorsports media, but with the motorsports fan base increasing to include more
afﬂuent consumers and more females, advertisers need a measurement of how people
differ demographically in their media usage. Future research should track how the
motorsports enthusiast market changes (if at all). Questions of interest might include
whether the market is including more females and more or fewer higher income people,
and what members of what professions tend to follow motorsports, for example. More
speciﬁcally, NASCAR claims its current fan base to be 41% female (NASCAR New
York Corporate Marketing Ofﬁce, 2001). Understanding whether the female fan base
increases or decreases will be important for marketers as they create messages integrating

motorsports. The same is true for the other demographics measures in the study, such as

100

income level and age. Understanding trends in fan demographics will help sponsors of
the sport better reach audiences.

When looking at age, motorsports sponsors must understand how to appeal to an
aging population. As baby boomers age, for example, motorsports sponsors will certainly
want to appeal to them. Assessing the appeal of motorsports can be done through surveys
and qualitatively in ways such as ethnographic interviews. Combining methodologies
will give marketers a greater understanding of what motorsports means to this audience.

Further explication of the not employed category might interest motorsports
marketers. Are these people searching for employment? Independently wealthy? Too
young to be employed? Are they a market to which one might advertise—or not? These
questions are important because, given the data in the present study, this group consumes
an inordinately high amount of certain motorsports-related media (i.e., Internet and
broadcast and cable television).

Finally, additional demographic variables not analyzed in the present study would
be critical to the study of the motorsports market. Knowledge of the market’s race and
education would help complete the picture of who the motorsports fan really is. For
example, current NASCAR data shows that roughly two thirds Of NASCAR fans are
white, 11% are Hispanic and 12% are African-American (NASCAR New York Corporate
Marketing Ofﬁce, 2001). Future research can be done to verify these ﬁgures, as well as to
track the changes in minority fan levels and understand the appeal of racing to members

of minorities such as Asians and Hispanics.

101

Study Limitations

This study, while using a large, national sample, did indeed have limitations.
Perhaps most notable is the inability to measure involvement accurately given the items
presented in the original scale. Obviously, a better picture of how involvement inﬂuences
media usage might have been obtained with additional and more speciﬁc involvement
items. Of course, this limitation is based on using a preexisting questionnaire. Perhaps if
the ﬁnal involvement variables were developed based on previous research (qualitative or
quantitative), they would be better measures of involvement than the initial 30 items.

Also, demographic variables for the present analysis should have included race

and education level to better understand the motorsports audience.

102

APPENDIX

103

APPENDIX

SURVEY BOOKLET: MOTORSPORTS INTEREST STUDY

Screener:

A.

B.

C.

Are you over 16 years of age?
Yes (CONTINUE)

NO (ASK TO SPEAK TO ADULT IN HOUSEHOLD, OTHERWISE
TERMINATE)

Have you purchased or leased a NEW, NOT previously owned vehicle in the past 5
years?

Yes (CONTINUE)

NO (TERMINATE)

DO you intend to purchase or lease a NEW, NOT previously owned vehicle in the
next 5 years?

Yes (CONTINUE)

NO (TERMINATE)

 

When I refer to street vehicles in this survey, I mean vehicles that you would ﬁnd at your

1.

local dealer.

First, thinking of the next NEW, NOT previously owned, vehicle buying decision in
your household, who will be involved in the decision as to what make and model to
buy or lease?

Self only
Self and other

Other only (ASK TO SPEAK TO OTHER PERSON, IF NOT HOME,
ARRANGE FOR A CALL BACK)

104

. Now thinking of different makes and models of cars, trucks, and vans, which make

and model comes to mind ﬁrst? Which comes to mind next? Which others? (LIMIT
TO 5)

. If you were to buy or lease a new vehicle in the next 5 years, what make and model
would you be most likely to consider? What others? (LIMIT TO 5)

. You may have already mentioned this, but which are would you be most likely to
buy or lease in the next few months? (IF TYPE OF "CAR" IS NOT MENTIONED,
SKIP TO Q.6) .

. If you were to buy or lease (CAR MAKE/MODEL MENTIONED ABOVE), would it

 

be a 2-door or 4-door?
2—door C11

4-door C12

 

CHECK QUOTAS, IF RESPONDENT QUALIFIES FOR MORE THAN
ONE OPEN CELL, ASK ABOUT VEHICLE IN CELL THAT IS MOST
IN NEED.

 

 

. When do you think you will buy or lease your next new, not used, vehicle? (READ

LIST)
In the next 12 months
In 1—2 years

In 3—5 years

Now I'd like to know how familiar you are with these various vehicles. For each make
and model I mention, can you tell me if you. . ..(read scale)

Have owned one

Have driven one but never owned one

Have heard a lot of things about it but never driven one
Have heard a few things about it, or have

Just heard the name before.

105

7. Earlier, you said that you would consider a (QUALIFYING VEHICLE) for your next
vehicle purchase. Please tell me which of these statements most applies to your
familiarity with (make/model).

 

Have Have Have Have Just
owned one driven one heard a lot heard a heard
but never about it few things the
owned one but never about it name
driven one before
First choice
make/model

Chevrolet
make/model

Ford make/model

Pontiac
make/model

Dodge
make/model

BMW
make/model

Toyota
make/model

 

8. Overall, how would you rate the 2000 (VEHICLE MODEL)? Would you rate it as
Excellent Very good Good Fair Poor DK

First choice model
Chevrolet (model)
Ford (model)
Pontiac (model)
Dodge (model)
BMW (model)

Toyota (model)

106

9. Now I’d like to know how likely you are to buy these various vehicles. For each make
and model I mention, can you tell me if. . ..

It's the only model you would consider

It's one of only two or three models you would consider
It's one of several

It's one that might get on your list Of possibilities, or you
Would not consider at all

10. Thinking about the next time you buy a vehicle, how likely are you to buy a
(make/model)?

 

It’s the only It’s one of It’s one of It’s one Would not
model you only 2 or 3 several that might consider at
would models you get on your all
consider would list of
consider possibilities

 

First choice
make/model
Chevrolet
make/model
Ford
make/model
Pontiac
make/model
Dodge
make/model
BMW
make/model
Toyota
make/model

My next series of questions are speciﬁcally about auto and truck racing. By that I mean
any 919; of auto or truck racing.

107

11. Please tell me your overall level of interest in each of the following types of auto and
truck races. How interested are you in (READ EVENT)? Are you . . .(READ
SCALE)?

(ITEMS WILL BE RANDOMIZED)

Extremely Very Somewhat Not very Not at all
interested interested interested interested interested
Indy Car racing 1 2 3 4 5
NASCAR Winston Cup
Stock car racing 1 2 3 4 5
NASCAR Busch
Grand National
stock car racing 1 2 3 4 5
NASCAR Craftsman
truck racing 1 2 3 4 5
NHRA drag racing 1 2 3 4 5
Desert or
off-road racing 1 2 3 4 5
Sports car racing 1 2 3 4 5
American Speed
Association 1 2 3 4 5
LeMans 1 2 3 4 5
Formula One 1 2 3 4 5
World Road Rally 1 2 3 4 5
Other 1 2 3 4 5

 

 

11a. (If “NOT AT ALL INTERESTED” to all, Skip to Q12) Which type of auto or truck
racing are you most interested in? (READ LIST. RECORD ONE.)

Indy Car racing

108

NASCAR Winston Cup stock car racing
NASCAR Busch Grand National stock car racing
NASCAR Craftsman truck racing

NHRA Drag racing

Desert or Off-road racing

Sports car racing

American Speed Association

LeMans __
Formula One

World Road Rally

Other type of racing (Specify)

12. About how many newspapgr articles about auto or truck racing, if any, have you read
in the past 3 months?

 

13. About how many auto or truck racing magazines or newspapers, if any, have you read
in the past 3 months? Please include racing magazines that you may subscribe to, as
well as those you may have bought at newsstands.

 

14. How many auto or truck races, if any, have you listened to on the radio in the past 5
months? '

 

15. How many auto or truck races, if any, have you watched on TV in the past 5 months?

 

16. How many times, if any, have you used the Internet speciﬁcally to ﬁnd information
about auto or truck racing in the past 5 months? (IF 0, SKIP TO Q. 18)

 

17. How many Web sites have you visited in the past 5 months related to auto or truck
racing?

 

b. What Web sites did you visit?

 

109

18. In the past 3 months, about how many times, if ever, have you watched _a__ny weekly
cable TV shows about auto or truck racing such as Speedka on ESPN, or Racetalk
on TNN, or SpeedVision?

 

19. Have you personally attended any auto or truck races in the past 12 months?
Yes :1, _, (CONTINUE)
NO 02 -> (SKIP TO Q. 22)
20. How many of those attended in the past 12 months were national auto or truck races?

By national races, I mean national events like Winston Cup Series races, Indy Car
races and other major racing events.

 

 

21. How many, within the past 12 months, were local auto or truck races?

 

22. How many Sppcialty auto exhibits or shows, if any, have you personally attended i_11
the past 12 month_s_? By specialty shows, I mean those exhibiting hot rods, classic
cars, etc. and p91 the annual new car auto shows.

 

23. Are you currently a member of any auto racing car club or race car driver fan club?
Yes C11 + (CONTINUE)
NO 02 -> (SKIP TO Q. 25)

24. What auto racing clubs or driver fan clubs are you a member of? (DO NOT READ.
RECORD ALL THAT APPLY)

National Hot Rod Association (NI-IRA) .................................. C11
Sports Car Club of America (SCCA) ...................................... C12
Any race car driver fan club .................................................... as
National Street Rod Association ............................................. 04
Other (Specify) ........................................................................ D5

27. Based on races you’ve recently seen or things you’ve recently read, how would you
rate each Of the following manufacturers in terms of “its level of involvement in auto

110

and truck racing”? Please use a scale from 1 to 10 where a 10 means extremely
involved and a 1 means not at all involved. How would you rate
(MANUFACTURER)? (NOTE: DO NOT PUSH FOR AN ANSWER. IF “DON’T
KNOW”, RECORD AS SUCH.)

 

 

Extremely Not at all
involved involved

Chevrolet 10 9 8 7 6 5 4 3 2 1
Dodge 10 9 8 7 6 5 4 3 2 1
Ford 10 9 8 7 6 5 4 3 2 1
Honda 10 9 8 7 6 5 4 3 2 1
Mazda 10 9 8 7 6 5 4 3 2 1
Mercedes Benz 10 9 8 7 6 5 4 3 2 1
Nissan 10 9 8 7 6 5 4 3 2 l
Oldsmobile 10 9 8 7 6 5 4 3 2 1
Pontiac 10 9 8 7 6 5 4 3 2 1
Toyota 10 9 8 7 6 5 4 3 2 1

28. Based on anything you might have seen or heard, how would you rate each of the
following manufacturers in terms of “being a winner in auto and truck racing”?
Please use a scale from 1 to 10 where a 10 means describes perfectly and a 1 means
does not describe at all. How would you rate (MANUFACTURER)? (NOTE: DO
NOT PUSH FOR AN ANSWER. IF “DON’T KNOW”, RECORD AS SUCH.)

 

 

Describes Does not

perfectly describe at all
Chevrolet 10 9 8 7 6 5 4 3 2 1
Dodge 10 9 8 7 6 5 4 3 2 1
Ford 10 9 8 7 6 5 4 3 2 1
Honda 10 9 8 7 6 5 4 3 2 1
Mazda 10 9 8 7 6 5 4 3 2 1
Mercedes Benz 10 9 8 7 6 5 4 3 2 1
Nissan 10 9 8 7 6 5 4 3 2 1
Oldsmobile 10 9 8 7 6 5 4 3 2 1
Pontiac 10 9 8 7 6 5 4 3 2 1
Toyota 10 9 8 7 6 5 4 3 2 1

29. (Ask only if answered extremely, very, or somewhat interested in NASCAR car or
truck racing in Q. 10) Which NASCAR vehicles are you aware of? (DO NOT READ
LIST)
Grand Prix
Intrepid
Monte Carlo
Taurus
Thunderbird

 

 

 

 

 

111

Indy Car racing 1 2 3 4 5

30. Please tell me how much you agree or disagree with each of the following statements.
(READ ROTATED STATEMENT, READ SCALE.)

 

Agree Agree Neither Disagree Disagree
strongly somewhat agree nor somewhat strongly
disagree

 

The only auto races that I watch are
the major events like the Indy 500 or
the Daytona 500. 1 2 3 4 5

I like to go to local auto and truck
races in my area. 1 2 3 4 5

I spend a lot of time watching auto
or truck races on TV. 1 2 3 4 5

I like to keep up-tO-date with the
careers of certain drivers. 1 2 3 4 5

The knowledge that a car manufacturer
gains from auto racing helps it build
more durable street vehicles for me. 1 2 3 4 5

My main interest in auto racing lies
more with the drivers than with the
cars or trucks. 1 2 3 4 5

I know much more than average about
the design and engineering Of race cars
and trucks. 1 2 3 4 5

I follow auto and truck racing so I can
keep up with the latest automotive
technologies. 1 2 3 4 5

When watching auto and truck racing,
1 watch the driver’s Skill and
technique. 1 2 3 4 5

The knowledge that a car manufacturer

gains from auto racing helps it build

street vehicles with better performance

and handling. 1 2 3 4 5

112

I like auto and truck racing events
for the party atmosphere. 1

The best thing about going to auto or
truck racing events is having a good
time with my friends. 1

I love the sound Of roaring engines. 1

The only reason I would go to an auto
or truck race is if a friend or family
member wanted to go. 1

The knowledge that a car manufacturer
gains from auto racing helps it build
street vehicles that are fun to drive. 1

I often fantasize about being a race
car driver.

My friends and I often talk about auto
or truck racing. 1

I often videotape auto or truck races
on TV. 1

One of the best things about watching
auto or truck racing is the crashes. 1

Auto and truck racing are great
entertainment for the whole family. 1

I often buy products or services from
companies that sponsor race teams. 1

The knowledge that a car manufacturer
gains from auto racing helps it build
street vehicles that are sporty. 1

I pay a lot of attention to the brands

of cars and trucks that participate
in auto or truck racing. 1

113

45

The brands Of cars and trucks that
win racing events are also the best
ones on the road. 1 2 3 4 5

If a brand of car or truck that I was

thinking about buying won a racing

event, it would increase my interest

in buying it. 1 2 3 4 5

The knowledge that a car manufacturer
gains from auto racing helps it build
safer street vehicles for me. 1 2 3 4 5

Compared to Others, I have a real
passion about the vehicles I drive. 1 2 3 4 5

I like to continually think of ways

to make my vehicle higher performance

by buying performance parts, kits,

or upgrade packages 1 2 3 4 5

I View my own vehicle primarily
as just a means Of transportation. 1 2 3 4 5

I enjoy doing work on my car to
enhance its performance, such as its
speed and handling. 1 2 3 4 5

31. What impact do you think a car manufacturer’s involvement and success in auto or
truck racing has on your overall interest in the vehicles it makes? Does it increase
your interest . . .

A great deal

Somewhat

Not very much

Not at all

32. How important is it to you that technology used on race cars is also used on street
vehicles?
Very important
Somewhat important
Neither
Somewhat unimportant
Very unimportant
Don't know

114

33. Based on anything you have seen, heard, or experienced, how would you rate each of
the following manufacturers in their ability to transfer successful race track
technology from race cars to street vehicles?

Excellent Very good Good Fair Poor DK

Chevrolet)
Ford
Pontiac
Dodge
BMW
Toyota
34. Other than routine maintenance or repairs, during the past 3 years, have you done any
work on your vehicle (either yourself or paid someone else) speciﬁcally to achieve
higher performance or to enhance its appearance?

Yes
No If no, skip to Q. 36

 

35a. Which of the following did you purchase?
Wheels, tires, suspension
Internal engine, drive train, ignition, exhaust, or fuel systems
Accessories and appearance (i.e., interior trim, specialty waxes, sunroof, spoiler)

Other

 

35b. Did you install some or all of these parts yourself?
Yes - all parts myself
Some parts myself
None of the parts myself

 

35c. Approximately how much in total did you spend on these upgrades parts or packages
whether they were installed by you or someone else? $

 

My next few questions are for classiﬁcation purposes only.

36. What is the make and model of the vehicle that you, yourself, own and drive most
often?

 

 

Make: Model:
What other vehicle makes and models, if any, are in your household?
Make: Model:

 

 

115

 

Make: Model:

 

37. Which of the following categories best describe your age?
16—24
25-34
35—44
45—54
55—64
65-74
75 or Older

38. Which of the following categories best describes your employment status?
Employed full time
Employed part time
Retired
Homemaker
Student
Not employed

39. Which one of the following categories b_es_t describes your occupation? (READ LIST)

Professional specialty

Executive, managerial, or business owner
Administrative support

Sales

Mechanics, repair, construction

Machine Operator, transportation, warehouseman
Technical support for health, engineering or sciences
Public safety such as police, ﬁre, etc.

Farming, forestry or ﬁshing services

Or some other occupation

40. Which of the following categories includes your total household income last year
before taxes?
Under $20,000
$20 to $39,000
$40 to $59,000
$60 to $79,000
$80 to $99,000
$100,000 to $ 150,000
More than $150,000

41. In which state do you live?

 

42. Record sex:
Male
Female

116

REFERENCES

117

REFERENCES

American Association of Advertising Agencies/Association of National Advertisers.
(2002). 2001 television commercial monitoring report. New York: Authors.

Andrews, J. C., Durvasula, S., & Akhter, S. (1990). A framework for conceptualizing and

measuring the involvement construct in advertising research. Journal of Advertising,
19(4), 27—40.

Antil, J. H. (1984). Conceptualization and operationalization of involvement. Advances in
Consumer Research, 11, 203—209.

Arora, R. (1982). Validation of the S-O-R model for situation, enduring, and response
components of involvement. Journal of Marketing Research, 19, 505-516.

Arora, R. (1985). Consumer involvement in retail store positioning. Journal of the
Academy of Marketing Science, 10(2), 109—124.

Backman, S., & Crompton, J. (1991). Differentiating between high, spurious, latent, and
low loyalty participants in two leisure activities. Journal of Park and Recreation
Administration, 9(2), 1—17.

Beatty, 8., Homer, P., & Kahle, L. (1988). The involvement-commitment model: Theory
and implications. Journal of Business Research, 16(2), 149—167.

Beatty, S., & Smith, S. (1987). External search effort: An investigation across several
product categories. Journal of Consumer Research, 14, 83—94.

Bloch, P. (1981). An exploration into the scaling Of consumers’ involvement with a
product class. Advances in Consumer Research, 8, 61—65.

Bloch, P. H., & Bruce, G. D. (1984). The leisure experience and consumer products: An
investigation of underlying satisfactions. Journal of Leisure Research 16(1), 74—88.

Bryan, H. (1977). Leisure value systems and recreation specialization: The case of trout
ﬁsherman. Journal of Leisure Research, 9, 184—187.

Bryan, H. (1979). Conﬂicts in the great outdoors: Toward understanding and managing
for diverse sportsmen ’s preferences. University, AL: Bureau Of Public
Administration, University of Alabama.

118

Bryan, H. (2000). Recreation specialization revisited. Journal of Leisure Research, 32(1),
18—21.

Buchanan, T. (1985). Commitment and leisure behavior. Leisure Sciences, 7, 401-420.

Celsi, R. L., & Olson, J. C. (1988). The role of involvement in attention and
comprehension processes. Journal of Consumer Research, 15, 210—224.

Cotter, T. (2001, May). NASCAR’S 2001 race Odyssey. Road & Track, 111-117.
Crane, D. (1972). Invisible colleges. Chicago: University of Chicago Press.

Day, E., Stafford, M., & Camacho, A. (1995). Opportunities for involvement research: A
scale-development approach. Journal of Advertising, 24, 69—75.

Dimanche, F., Havitz, M., & Howard, D. (1991). Testing the involvement proﬁle (IP)
scale in the context Of selected recreational and tourist activities. Journal of Leisure
Research, 23(1), 51—66.

Ditton, R., Loomis, D., & Choi, S. (1992). Recreation specialization: Re-
conceptualization from a social worlds perspective. Journal of Leisure Research,
24(1), 33—5 1.

Donnelly, M. P., Vaske, J. J ., & Graefe, A. R. (1986). Degree and range of recreation
specialization: Toward a typology of boating related activities. Journal of Leisure
Research, 18(2), 81-95.

Ellis, G. D., & Witt, P. A. (1994). Perceived freedom in leisure and satisfaction:
Exploring the factor structure of the perceived freedom components of the leisure
diagnostic battery. Leisure Sciences, 16(4), 259-270.

Ferrand, A., & Pages, M. (1996). Football supporter involvement: Explaining football
match loyalty. European Journal for Sport Management, 3(1), 7—20.

Goode, W. (1957). Community within a community: The professions. American
Sociological Review, 22, 194—200.

Greenwald, H. (1965). The involvement controversy in persuasion research. [unpublished
manuscript], Columbia University.

Greenwald, H., & Leavitt, C. (1984). Audience involvement in advertising: Four levels.
Journal of Consumer Research, I I, 581-592.

119

Gunter, B. G., & Gunter, N. C. (1980). Leisure styles: A conceptual framework for
modern leisure. The Sociological Quarterly, 21, 361—374.

Gupta, V. (2000). Regression explained in simple terms. In SPSS for beginners.
Washington, DC: VJBOOks.

Hagstrorrr, R. G. (1998). The NASCAR way: The business that drives the sport. New
York: Wiley.

Harris Interactive. (2001, May), NASCAR fan not typical anymore, and sponsors wake
up to catch new opportunities. Sportainment News, 1(2). [online newsletter; no longer
available].

Havitz, M. E., & Dimanche, F. (1997). Leisure involvement revisited: Conceptual

conundrums and measurement advances. Journal of Leisure Research, 29(3) 245—
278.

Havitz, M. E., & Dimanche, F. (1990). Propositions for testing the involvement construct
in recreational and tourism contexts. Leisure Sciences, 12(2), 179-195.

Havitz, M. E., & Dimanche, F. (1999). Leisure involvement revisited: Drive properties
and paradoxes. Journal of Leisure Research, 31(2), 122-149.

Havitz, M. E., & Howard, D. R. (1995). How enduring is enduring involvement? A
seasonal examination of three recreational activities. Journal of Consumer
Psychology, 4, 255—276.

Hays, W. (1988). Statistics. Chicago: Holt, Rhinehart and Winston.

Higie, R. A., & Feick, L. F. (1989). Enduring involvement: Conceptual and
methodological issues. Advances in Consumer Research, 16, 690—696).

Houston, M., & Rothschild, M. (197 8). Conceptual and methodological perspectives on
involvement. In S. Jain (Ed.), Educators’ Proceedings (pp. 84—87). Chicago:
American Marketing Association.

Hupfer, N. T., & Gardner, D. M. (1971). Differential involvement with products and
issues: An exploratory study. In D. Gardner (Ed.), Proceedings of the Association for
Consumer Research (pp. 262—270).

Iwasaki, Y., & Havitz, M. E. (1998) A path analytical model of the relationships between
involvement, psychological commitment, and loyalty. Journal of Leisure Research,
30(2), 256—280.

120

Katz, F. (1958). Occupational contact networks. Social Forces, 37, 52—55.

Krugman, HE. (1965). The impact of television advertising: Learning without
involvement. Public Opinion Quarterly, 29, 349—356.

Lastovicka, J. L., & Gardner, D. M. (1979). Components Of involvement. In J. C.
Maloney & B. Silverrnan, Eds.), Attitude Research Plays for High Stakes (pp. 53-73).
Chicago: American Marketing Association.

Laurent, G., & Kapferer, J. (1985). Measuring consumer involvement proﬁles. Journal of
Marketing Research, 22, 41—53.

Laverie, D. A., & Amett, D. B. (2000). Factors affecting fan attendance: The inﬂuence of
identity salience and satisfaction. Journal of Leisure Research, 32(2), 225—246.

Mannell, R. C. (1980). Social psychological techniques and strategies for studying leisure
experiences. In S. E. Iso-Ahola (Ed.), Social psychological perspectives on leisure
and recreation (pp. 62—88). Springﬁeld, IL: C. C. Thomas.

McAdams, D. (2001). For women only. Broadcasting and Cable, 130(3), 100.

McFarlane, B. L. (1994). Specialization and motivations of birdwatchers. Wildlife Society
Bulletin, 22, 361—370.

McIntyre, N. (1989). The personal meaning of participation: Enduring involvement.
Journal of Leisure Research, 21(2), 167—179.

McIntyre, N., & Pigram, J. J. (1992). Recreation specialization reexamined: The case of
vehicle-based campers. Leisure Sciences, 14(1), 3—15.

McQuarrie, E. F., & Munson, J. M. (1987). The Zaichowsky personal involvement
inventory: Modiﬁcation and extension. Advances in Consumer Research, 14. 36—40.

Mullins, N. (1973). Theory and theory groups in sociology. New York: Harper and Row.

NASCAR New York Corporate Marketing Ofﬁce (2001). NASCAR demographics. New
York: NASCAR Information Fax Service.

Nayyar, S. (2001). A matter of interpretation. American Demographics 23(3), 6.

Oliver, R., & Bearden, W. (1983). The role Of involvement in satisfaction processes.
Advances in Consumer Research, 10, 250—255.

121

Ostrom, T., & Brock, T. (1968). A cognitive model of attitudinal involvement. In R. P.
Ableson, et al. (Eds.), Theories of cognitive consistency. New York: Rand McN ally.

Parkinson, T. L., & Schenk, C. T. (1980). An empirical investigation of the S-O-R
paradigm of consumer involvement. Advances in Consumer Research, 7, 696—699.

Parnell, R. (1999, October 31). Green ﬂags are waving for NASCAR. In The Dismal
Scientist (http://www.dismal.com), West Chester, PA. Retrieved from
http://www.economy.com/dismal/default.asp; subscription required.

Pastore, M. (1999, November 30). Internet may lead to power shift in sports industry.
Retrieved August 3, 2001, from http://cyberatlas.internet.com/markets/
professional/article/O, 1323,5971_25 1761 ,00.html

Pedhazur, E., & Pedhazur-Schmelkin, L. (1991). Measurement, design, and analysis: An
integrated approach. I-Iillsdale, NJ: Lawrence Erlbaum Associates.

Petty, R. E., & Cacioppo, J. T. (1981). Issue involvement as a moderator of the effects on
attitude of advertising content and context. Advances in Consumer Research, 8, 20—
24.

Petty, R., Cacioppo, J ., & Schumann, D. (1983). Central and peripheral routes to
advertising effectiveness: The moderating role of involvement. Journal of Consumer
Research, 10, 135—146.

Picard, R. (1999, October 21—23). Audience fragmentation and structural limits on media
innovation and diversity. Paper presented at the Second Expert Meeting on Media in
Open Societies, University of Amsterdam, Amsterdam.

Redden, J ., & Steiner, C. J. (2000). Fanatical consumers: Towards a framework for
research. Journal of Consumer Marketing, 1 7(4), 322-337.

Richins, M. L., & Bloch, P. H. (1986). After the new wears off: The temporal context of
product involvement. Journal of Consumer Research, 13, 280—285.

Richins, M. L., & Bloch, P. H. (1991). Post-purchase product satisfaction: Incorporating
the effects of involvement and time. Journal of Business Research, 23, 145—158.

Rothschild, M. (1979). Marketing communications in nonbusiness situations, or why it’s
so hard to sell brotherhood like soap. Journal of Marketing, (43), 11-20.

Schnritt, N. (1996). Uses and abuses Of coefﬁcient alpha. Psychological Assessment, 8(4),
350—353.

122

Schreyer, R., & Beaulieu, J. T. (1986). Attribute preference for wildland recreation
settings. Journal of Leisure Research, 18(40), 231-247.

Scott, D., & Godbey, G. (1994). Recreation specialization in the social world of contract
bridge. Journal of Leisure Research, 26(3), 275—295.

Selin, S. W., & Howard, D. R. (1988). Ego involvement and leisure behavior: The
concept of situational self-image. Journal of Leisure Research, 20(3), 237—244.

Shamir, B., & Ruskin, H., (1984). Sport participation vs. sport spectatorship: Two modes
Of leisure behavior. Journal of Leisure Research, 16(1), 9—21.

Sherif, C. W. M., & Nebergall, R. E. (1965). Attitude and attitude change: The social
judgment involvement approach. New Haven, CT: Yale University Press.

Sherif, M., & Cantril, H. (1947). The psychology of ego-involvements, social attitudes,
and identiﬁcations. New York: Wiley and Chapman and Hall.

Shibutani, T. (1961). Society and personality. Englewood Cliffs, NJ; Prentice-Hall.

Smith & Street. (2001). Top 10 cable sports telecasts. Charlotte, NC: Smith & Street ’s
SportsBusiness Journal.

Snyder, E. E., & Spreitzer, E. A. (1974). Family inﬂuence and involvement in sports. The
Research Quarterly, 44(3), 249—255.

Stossel, J. (2001, December 19). Advertising in the bathroom? It ’s getting out of hand.
ABC News Internet Ventures. Retrieved from
http://abcnews.go.com/sectionS/2020/DailyNews/GMAB_ads_01 1219.html

Williams, D. R. (1984). A development model of recreation choice behavior, In G. H.
Stankey and S. F. McCOOl (Eds) Proceedings: Symposium in Recreation Choice
Behavior (pp. 31—37). Washington, DC: US. Department of Agriculture.

Wright, P. W. (1974). Analyzing media effects on advertising responses. Public Opinion
Quarterly, 38, 192—205.

Zaichowsky, J. L. (1985). Measuring the involvement construct. Journal of Consumer
Research, 12(3), 341-352.

123

 

 

   

ljljjjjjjjji(gift