MSU

 

 

 

RETURNING MATERIALS:
ace 1n oo rop to
remove this checkout from

 

w your record. FINES wHI

‘ be charged if book is
returned after the date
stamped beIow.

ww "a

 

 

 

CONSUMER CHARACTERISTICS AND POST-PURCHASE SATISFACTION:
A COMPARATIVE STUDY OF ACCOUNTING AND

HOTEL AND RESTAURANT MANAGEMENT MAJORS

BY

James Carroll Cooper

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

DOCTOR OF PHILOSOPHY

Department of Marketing and Transportation Administration

1981

ABSTRACT

CONSUMER CHARACTERISTICS AND POST-PURCHASE SATISFACTION:
A COMPARATIVE STUDY OF ACCOUNTING AND
HOTEL AND RESTAURANT MANAGEMENT MAJORS

BY

James Carroll Cooper

This study examined the relationship between con-
sumer characteristics and post-purchase satisfaction. The
question at issue was whether the relationships differed
between purchasers of competing brands. The expectation
that differences exist between brands receives implicit
support in market segmentation theory. The marketing lit-
erature, however, provides no empirical evidence on this
question although evidence does exist that consumer char-
acteristics are associated with satisfaction.

The brand choice upon which the study focused was
the selection of academic major by university students.
The research subjects were Michigan State University seniors
enrolled in accounting or hotel and restaurant management.
The research design was a cross-sectional, ex post facto
survey. The data were collected using a questionnaire
personally distributed to classes in which subjects were
concentrated.

The personal characteristics data collected included

measures of demographics, self-image, values, and activ-
ities, interests, and opinions. All data, including the
satisfaction measures, were subjected to cluster analysis
reduction procedures. For each group of students, the re-
lationship between consumer characteristics and satisfaction
was evaluated using both bivariate (product-moment correla-
tion) and multivariate (multiple regression) analysis. At
both levels attention was directed to a comparison between
groups of the strength and direction of associations ex-
hibited between given variables and satisfaction.

The tentative results of the bivariate analysis
were reinforced by the results of the more rigorous mul—
tivariate analysis. The multiple regression analysis yielded
clearly identifiable differences between groups in the asso-
ciation of given variables with satisfaction. Satisfaction
was related to choice of major (brand choice). Specifically,
an individual's satisfaction with accounting was greater if
that individual reflected the characteristic labeled social
desirability but was low in the characteristic labeled in-
dividualism. In contrast subjects high in individualism re-
ported greater satisfaction with hotel and restaurant man-
agement. Althought the emphasis of the study was not placed
on maximizing the prediction of satisfaction, it should be
noted that the majority of variance in satisfaction was not
explained by consumer characteristics.

The essence of the analysis is that the belief that

consumer satisfaction is related to brand choice is supported.

It is, therefore, consistent with the implications of market—
ing theory and segmentation theory in particular. An in-
creased understanding of the processes involved in the gen-
eration of consumer satisfaction requires, therefore, that
the analysis not be restricted to consideration of sellers'
actions. Instead more attention should be paid to varia-
tions among consumers and the market mechanisms through

which buyer-seller interaction takes place.

This dissertation is dedicated
to my wife Sandra
and to my sons Lee, Reid, and Kelly

ii

ACKNOWLE DGMENTS

The author would like to acknowledge the invaluable
support, both personal and professional, of his major pro-
fessor Dr. William Lazer. Without it this project would not
have been completed. Additionally, acknowledgments are due
to Dr. Donald Taylor and Dr. Gilbert Harrell for their con-
tributions as dissertation committee members and advisors.

Appreciation is also noted to James Anderson who
provided technical assistance in analysis of data and to the
faculty and students in accounting and hotel and restaurant
management whose cooperation enabled the collection of the
data base.

iii

TABLE OF CONTENTS

Page
LIST OF TABLES. . . . . . . . . . . . . . . . . . . . Vi
Chapter
I. INTRODUCTION. . . . . . . . . . . . . . . . . 1
Purpose of Study. . . . . . . . . . . . . . 1
Theoretical Background. . . . . . . . . . . 2
Problem Statement . . . . . . . . . . . . . l7
Methodological Summary. . . . . . . . . . . 19
Limitations . . . . . . . . . . . . . . . . 22
II. LITERATURE REVIEW . . . . . . . . . . . . . . 27
Scope . . . . . . . . . . . . . . . . . . . 27
Post-purchase Processes . . . . . . . . . . 28
Complexity Theory . . . . . . . . . . . . . 31
Attitude Change . . . . . . . . . . . . . . 35
Reward and Satisfaction . . . . . . . . . . 35
Consumerism . . . . . . . . . . . . . . . . 40
Complaining Behavior. . . . . . . . . . . . 41
Satisfaction. . . . . . . . . . . . . . . . 49
Segmentation. . . . . . . . . . . . . . . . 54
Summary . . . . . . . . . . . . . . . . . . 59
III. RESEARCH METHODOLOGY. . . . . . . . . . . . . 63
Introduction. . . . . . . . . . . . . . . . 63
Research Approach . . . . . . . . . . . . . 66
Research instrument . . . . . . . . . . . 67
Dependent variable. . . . . . . . . . . 67
Independent variables . . . . . . . . . 74
Analytical procedures . . . . . . . . . . 82
Data gathering procedures . . . . . . . . 85
IV. RESEARCH FINDINGS . . . . . . . . . . . . . . 89
Bivariate Analysis. . . . . . . . . . . . . 89
Multivariate Analysis . . . . . . . . . . . 93
Brand Perceptions . . . . . . . . . . . . . 101

iv

V.

SUMMARY AND CONCLUSIONS . . . . . .1

Objectives. . . . . . . . . . . .
Background. . . . . . . . . . . .
Methods . . . . . . . . . . . . .
Summary of Findings . . . . . . .
Implications for Marketing Theory

Practice. . . . . . . . . . . .
Limitations and Future Research .
Concluding Comments . . . . . . .

APPENDICES

A

B

RESEARCH INSTRUMENT . . . . . . . .

PREDICTOR VARIABLE COMPONENTS . . .

BIBLIOGMPHY O O C C O O O O O I O O O O O

Page
111
111
112
116
118
119
121
123
125
132

135

Table

10.

11.

12.

13.

LIST OF TABLES

Predictor Variables . . . . . . . . . . . . .
Correlations between Clusters . . . . . . . .

Bivariate Correlations:
Satisfaction with Predictors. . . . . . . .

Transformations:
Correlation Coefficients to Z Values. . . .

Multiple Regression Results: ACC . . . . . .
Multiple Regression Results: HRI . . . . . .
Multiple Regression Results: Total Sample. .
Significant Predictors and R2 by Group. . . .
Perceptions of ACC as a Major . . . . . . . .
Perceptions of HRI as a Major . . . . . . . .

ACC Students:
Paired Comparisons of ACC and HRI as Majors

HRI Students:
Paired Comparisons of ACC and HRI as Majors

PREDICTOR VARIABLE COMPONENTS . . . . . . . .

vi

Page
80
81

90

92
95
96
98
98
103

105

106

107

132

CHAPTER I

INTRODUCTION

Purpose of Study

 

The purpose of this study is to investigate selected
aspects of marketing theory dealing with market segmentation
and post-purchase evaluative phenomena. Segmentation and
post-purchase issues are important components of both
marketing theory and marketing practice. Of the two areas
post-purchase questions have the shorter tradition of em-
pirical research. The tradition is even shorter for the
area within the range of post-purchase concerns that is
coming to be known as satisfaction research. This study
focuses on this area.

Pioneer studies in the area of consumer satisfaction
frequently show the influence of prior segmentation studies.
For the most part, however, these studies have drawn on
segmentation theory rather than attempted to test or extend
.it. Consequently, satisfaction research has developed with
a set of focal concerns largely independent of segmentation
theory. The intent of this study is to attempt to establish
stronger links between the two research domains. The purpose
is to investigate whether the logic of segmentation theory
is substantiated by empirical evidence when extended beyond

purchase to include evaluative judgments.

2

'Theoretical Background

 

Webster defines satisfaction as "1. satisfying or
being satisfied 2. something that satisfies; specif.,
a) anything that brings gratification, pleasure, or content-
ment."1 Given this definition it would seem that the sub-
ject of satisfaction would inspire neither controversy nor
differences of opinion. For the most part this has been
true. Satisfaction is highly valued and pursued. It is
likely that most people would agree with Confucius that:
"The superior man is satisfied and composed; the mean man

is always full of distress."2
There are indications, however, that in certain
circumstances, times, or places other goals or values
are preeminent. John Stuart Mill illustrated this very
aptly with the statement: "It is better to be Socrates
dissatisfied than a pig satisfied."3 Charles Steinmetz
elaborated on the theme that the world belongs to the dis-
satisfied when he said:
For me the one underlying principle of all human pro-
gress is that "divine discontent" which makes men
strive for better conditions and improved methods. . . .
Through my years of experience I have come to see that
one can never rest in this game of life, because it is
a race that is never-ending. Those who stay in front
of the pack have learned this truth early in their
careers, and that is why they are in front. The men
in the rear are the Easily satisfied, the ones who are
content with little.
Steinmetz clearly valued the dynamic notions of achievement
and proqress as more worthy than satisfaction which he saw

as a static state.

More recently it has been observed that satisfaction

3

in the context of consumption may be a mixed blessing.
Olander (1977) argued that satisfaction measures are in-
adequate as indicators of consumer welfare for purposes
of public policy formation. His argument is premised

on a view of satisfaction as a function of the level of
aspiration. Since aspirations change, this is equivalent
to measuring with a rubber yardstick. Likewise, if
frustration leads to lowered aspirations, what appears to
be satisfaction may be little more than resignation.
Olander concluded that more time should be spent on
objective indicators of welfare rather than using subjective
measures such as satisfaction.

Engledow (1977) addressed the value of satisfaction
as a managerial performance indicator and concluded that it
doesn't measure up. He saw the concept as ill-defined and
lacking operational measurement devices, pointing out that,
"a thorough reading of any current marketing text reveals
that the discussion of satisfaction generally ends with the
ringing endorsement of the marketing concept in chapter
one."5 Generally speaking, his view is more benign than
that of Olander who is less certain that satisfaction should
be a major objective of the firm or is admirable in society.
Olander's reluctance is a reminder that guidance of an
economic system by overt preferences via the principle of
consumer sovereignty is not ethically neutral. Thorstein
Veblen, best known to marketing students as the originator

of the phrase "conspicuous consumption" which redirected

attention toward consumer motivation and social motives

in particular, expressed serious reservations about ground-
ing analysis of economic efficiency upon consumers. He
viewed the tendency toward conspicuous consumption as be-
ing very common and reflective of competitive urges.

He concluded that:

In order to meet with unqualified approval, any
economic fact must approve itself under the test of
impersonal usefulness — usefulness as seen from the
point of view of the generically human. Relative

or competitive advantages of one individual in
comparison with another does not satisfy the economic
conscience and therefore competitive expenditure has
not the approval of this conscience.

A reviewer of Veblen's work reemphasized this point:

Thus for Veblen the orthodox analysis of the allocative
mechanism had little meaning. The definition of
efficiency vis—a-vis the ability of the system to

meet the demands of the individual consumer was in-
consistent with Veblen's definition of efficiency
vis-a-vis society. It further overlooked the

important fact that the individual was manipulated

by business through salesmanship to suit the ends of
business. The consumer could scarcely be taken as the
benchmark for the analysis.

The question raised by Olander and Engledow cannot
be ignored by satisfaction researchers. It should be
noted, however, that they are minority voices. Most students
of satisfaction issues appear to be in sympathy with Pfaff's
answer to the question of whether satisfaction is good.
I say yes, indeed. I prefer satisfaction to dissat-
isfaction. More seriously, we have to consider several
goals simultaneously, of which consumer satisfaction
is one. Other goals being attained equally, I prefer
more satisfaction to less satisfaction. Beyond a
certain level of consumer satisfaction, other goals
are not equal any more and this is where the tradeoff
question comes in.

In similar fashion there is general support for the point

of view expressed by Day: "I see no insurmountable
difficulties in measuring an individual's satisfaction or
dissatisfaction with the overall outcome of a specific
recent personal consumption experience."9 The growing
body of research dealing with conSumer satisfaction provides
evidence of agreement with Day.

An understanding of the reasons for the growing
interest in satisfaction research requires a brief review
of the treatment of satisfaction in economic and marketing
theory. The economists of the nineteenth and early
twentieth centuries believed that the amount of satis-
faction derived from the consumption of a particular market
basket of goods was measureable. It was called utility and
the unit of measurement (with a cardinal scale) was the util.
This was a dubious assumption, and modern utility theory
has progressively trimmed away strong assumptions about
utility and human psychology as they were demonstrated to
be unneeded for efficient theory.

Modern utility theory employs ordinal utilities,
and in modern theory "the only thing that can be said on
the basis of these utilities is that market baskets with
higher utilities are preferred by the consumer to market
baskets with lower utilities, and that market baskets with
equal utilities are equivalent in the eyes of consumers."10
The question of which market baskets maximize the consumer's
utility is answered by reference to the consumer's indif-

ference map, or set of indifference curves, which represent

tastes or preferences. The consumer's indifference map

is deduced from actual behavior according to the theory of
revealed preference. Since a rational consumer is postu-
lated and the consumer possesses complete market information,
satisfaction is presumably maximized within the budget
constraint by the choice process.

An elaboration of conventional demand theory has
been offered by Lancaster (1971). In his view, "traditional
demand theory, like the traditional theory of the firm, has
its roots in the economics of an earlier, and simpler,
society, when there were fewer products, each more or less
standard, and a simpler technology."11 Consequently, it
possesses the "coarse structure" necessary to demonstrate
broad substitution effects among goods which are typically
agreggates such as "automobiles," "clothing," and "food"
rather than individual items. While it serves this purpose
well, it is inadequate to deal with important current issues
in demand:

The omission in the traditional analysis of any
provision for using information concerning the tech-
nical characteristics of goods renders it completely
incapable of handling the most important aspects of
demand in an advanced society -- the effects of product
variations and differentiates, modeizchanges, new goods,
and new variants of ex1st1ng goods.

In an earlier work on demand theory, Hicks (1956)
had discussed the possibility that economists should view
consumer behavior as similar in structure to that of entre—

preneurs. That is, objectives are established according to

preferences, and then an evaluation of alternative means of

7

reaching the objectives is required. The commodities
purchased would more accurately be viewed as means to the
attainment of objectives rather then as the objectives
themselves. While Hicks concluded that technical
difficulities precluded developing this view, Lancaster
set about to overcome the problems.

Lancaster's analysis is premised on the belief
that "it is the characteristics of goods, not the goods

themselves, in which people are interested."13

The working
hypothesis of his book is that "individuals differ in their

reactions to different characteristics, rather than in their

 

assessment of the characteristics content of various goods
collections."14

The differences between Lancaster's analysis and
traditional theory are differences of degree rather than
kind. He views traditional theory as a special case of
his own more general theory. To the extent this is true,
his work brings us closer to an understanding of consumer
preferences and purchase behavior than the work of his
fellow economists. Because of his reluctance to
"psychologize", however, we merely substitute unanswered
questions about the relationships among preferences and
goods characteristics for the older questions about pre-
ferences and goods. We are left, then, with the same almost-
axiomatic identity between consumption goals and the ex-
pression of preferences through purchases which character-

izes demand theory.

8

The empirical study of consumer behavior by
economists has undoubtedly been impeded by this basic
component of demand theory. Much of the effort that has
gone toward improving consumer welfare has been devoted
to the supply side of the market equation, i.e. to in-
dustrial organization and structure and the market be-
havior of individual firms.

Within the marketing discipline the situation has
been somewhat different. The roots of the field in
economics are evidenced by the key role played by the
marketing variant of the theory of revealed preference:
that consumers are the best judges of what is in their own
best interests. The professional, managerial, decision
emphasis which characterizes marketing has encouraged
more extensive empirical investigations of consumer
behavior in recent years, particularly in the highly
developed western economies. Indeed, recognition of the
need to know more about consumer behavior evolved as
marketing became a more important function within the firm.

Yet, the quality of the need for information kept
the scope of this investigation within relatively narrow
confines. In economics the welfare orientation required a
level of abstraction sufficiently removed from the realities
of the marketplace that many serious questions relating to
consumption were not scientifically studied. In marketing,
by way of contrast, the needs of the firm required that

attention be focused on the choice process; the determinants

of repurchase behavior. Evaluative judgments were simply
seen as less useful than behavioral measures of brand switch-
ing and brand loyalty. The prevailing View in the early
sixties was expressed in Hollander's statement that "the
devices for identifying and measuring consumer satisfaction
in any general sense are limited to votes in the marketplace,
which is probably the most significant argument for a free

15 This statement was accurate not because

marketplace."
efforts to develop alternative measures had failed but
rather because of the absence of any sustained effort in
that direction.

Even prior to that time, Alderson (1958) had
anticipated the likelihood that "some day marketing may
need to look beyond the act of purchasing to a study of

16

consumption proper." He saw the need arising as a con-

sequence of choice problems in conditions of abundance and
involving marketers' attempts to encourage acceptance of
product proliferation and other innovations. He raised
the questions:

What is the use of a plethora of goods unless the

buyer derives genuine satisfaction from them? What

is the justification of surfeit if the acquisition

of goods serves as a distraction from activities

which are essential to the preservation of our

culture and of the integrity of our personalities?17

He suggested the name hedonomics for a field of study in-

18 This

volving the "management of the capacity to enjoy."
hypothetical discipline would clearly differ from marketing
in which consumer motivation and choice processes are im-

portant because of their impact on the firm. It is also

10

unlike consumer economics or home economics where in-
dividual or household purchase behavior is studied within
a framework of economic optimization.

Alderson's proposed new discipline is also
dissimilar to the concerns with social indicators and the
quality of life although these areas of research are of
recent origin and share an interest in human welfare. Each
of these fields of study contains its own stimuli for a
renewed interest in the consequences of exchange processes.

Two relatively recent phenomena have forced
recognition within marketing of the limited understanding
that exists about the post-purchase phase of consumption
behavior and, more specifically, the determinants and
predictors of satisfaction and dissatisfaction. The first
of these was the diffusion of the marketing concept through
the business world and, subsequently, through organizations
not characterized by profit-seeking. The second was the
consumer movement.

The marketing concept shifted the focus of organ-
izational influence and decision-making away from an in-
ternal, production orientation and toward an external
market orientation. The welfare consequences of any
system of voluntary exchange began to surface for more
explicit treatment as marketers postulated consumer sat-
isfaction as a commercial objective. A fairly typical
statement of the role of satisfaction in the marketing

concept is provided by DeLozier and WOodside:

11

"The marketing concept can be defined as the philosophy

 

of focusing all the activities of the organization on

satisfying customer needs through the use of integrated

 

 

marketing to achieve maximum profits through customer
satisfaction."19

Explicit inclusion of customer satisfaction in
discussions of the firm's objective function heightens
both its sophistication and its complexity. The decision-
making process is complicated by the expansion of inputs.
Likewise, attention is called to the paradoxical combina-
tion of both mutuality and conflict of interest between
the parties involved in exchange.

Obvious problems associated with maximizing customer
satisfaction (e.g., distributing products free) have forced
recognition of the need for balance and compromise in
structuring organizational actions toward achieving mul-
tiple objectives. As a result, customer satisfaction may
be compromised or subordinated in the objective function
such that the amount of satisfaction the firm attempts to
deliver is related to its competitive objectives and the
competitive environment.

Although the problems associated with implementing
the marketing concept have been recognized (Stampfl, 1978),
it has been argued that it is not inclusive enough.

DeLozier and Woodside cite two common criticisms of the
'marketing concept: "(1) The marketing concept overlooks

the conflict between individual want satisfaction and long—run

12

public interest. (2) The goal of the marketing concept
represents only interest -- profits for the organization."20
One response to these complaints is reduced concern with
stockholders and expanded interest in "stakeholders",
pe0ple who are influenced in a variety of ways by the
organization's decisions. The societal marketing concept
developed by Kotler reflects this approach: "The societal
marketing concept is a management orientation aimed at
generating customer satisfaction and long-run consumer and
public welfare as the key to satisfying organizational
goals and responsibilities."21

To the extent the marketing concept is difficult
to operationalize, the societal marketing concept is even
more so. Neither concept incorporates the reality of
consumer dissatisfaction with existing purchase patterns
which is the intentional result of aggressive marketing
efforts designed to build demand for a new product or
service. Notwithstanding conceptual and operational
difficulties and an occasional minority Opinion, the concept
of consumer satisfaction has assumed an important position
in marketing theory.

It is not surprising, however, given the prevailing
pragmatism of modern managers, that operationalized
versions of dissatisfaction have received more managerial
attention than satisfaction per se. Brand or institutional

loyalty has been interpreted as an acceptable indicator

of satisfaction. Attempts to measure satisfaction directly

13

have been viewed as difficult, expensive, and of little
value if indeed it was even possible.

Dissatisfaction, however, was another thing. It
need not be sought in order to be measured. The firm can
tolerate or perhaps even ignore small amounts of dis-
satisfaction, treating it as a noncritical concern. By
the same token, low intensity dissatisfaction may be ex-
pressed through brand switching or other non—threatening
means. However, the firm's ability to compete and its
social sanction are threatened by expression of high
levels or large amounts of dissatisfaction.

It is for this reason that the consumer movement
has also been important in the expansion of interest in the
evaluative component of the post-purchase phase. The
movement encompasses a broad range of issues touching in
some fashion virtually all aspects of modern marketing
programs. Among them are selling and advertising practices,
safety concerns, and resource waste. These and related
issues are important but incidental to the purpose of this
study. The facets of the consumer movement of greater
relevance to this study are those providing specific
evidence of consumer dissatisfaction with goods and services.
It is this dissatisfaction which is so antithetical to
normative marketing theory and the consumer satisfaction
which it embraces.

The most tangible evidence of dissatisfaction is

the large volume of consumer complaints filed regularly

14

with the Better Business Bureaus, Attorneys General
offices of the various states, the Office of Consumer
Affairs at the federal level, assorted trade associations,
and numerous regulatory agencies. Unlike other features
of the consumer movement which critics discount because
evidence of general support is lacking, complaint volumes
do provide documentation of widespread dissatisfaction.

At the same time, there is growing research support for
the belief that complaints are merely the tip of the
dissatisfaction iceberg.

The consumer movement clearly constitutes a basic
challenge to current marketing thought and practice. It
challenges students of the field to address the meth-
odological and measurement issues related to consumer sat-
isfaction that have escaped emghfical examination and have
been discussed more from a philosophical point of view. A
better understanding is needed of the interactive process
involving consumer characteristics, market stimuli, decision—
making, and post—purchase evaluation of the level of satis-
faction attained. As Engledow pointed out: "Empirical
evidence on how satisfactions link together and interact
and conceptual frameworks for analyzing satisfaction in-
formation are lacking."22

There has been a dramatic increase in recent years
in the number of research studies focusing on the post—

purchase phase of consumer decision processes and satis-

faction issues in particular. Most of the material cited

15

subsequently in the literature review chapter was published
in the 1970's. As the number of studies has increased,

the subject area has become increasingly fractured with
growing diversity in research topics and methodologies
employed.

The researcher faces a number of Options in select-
ing productive research areas and specific topics within the
domain of satisfaction research. It is an ironic parallel
with the problems consumers face in choosing among the
proliferation of products and services available in a
modern economy.

Traditionally, economists and marketers have agreed
that a wide range of choices is valuable for consumers and
an asset for an economy. The reasons for favoring extensive
choice differ in the two disciplines. In virtually all
schools of economic thought, choice is seen as an indicator
of healthy competition. Healthy competition is believed to
be beneficial to the public because market power resides
more with consumers than producers, and there is reduced
likelihood that sellers can charge high prices and command
monopoly profits. It shOuld be emphasized that these
conditions hold when the goods in question are undifferen—
tiated commodities, and the choice is therefore restricted
to source of supply. It presumably is made solely on the
basis of price. A differentiated product provides the
seller with a degree of monopoly power and therefore works

against market efficiency. Mansfield points out that

16

efficiency is the dominant interest of most economists:

Some people favor competition simply because it
prevents the undue concentration of power and the
exploitation of consumers. But to the economic
theorist, the basic argument for a perfectly com-
petitive economy is the fact that such an economy
satisfies these conditions [for welfare‘maximization].23

Alternatively, marketers promote the advantages
of choice as flowing from the natural heterogeneity of
demand. Their belief is that consumers' preferences are
frequently characterized by dissimilarity rather than
similarity, and the availability of many choices, par-
ticularly in the non-price attributes of products, insures
that demand will be well met. This theory was most fully
developed in Alderson's functionalist theory.24 One of
the fundamental conceptions of this theory was the
heterogeneous market which was "one in which there was
a precise match between differentiated units of supply

25

and differentiated segments of demand." Heilbroner

concedes that point somewhat grudgingly although his
preference for efficiency is not concealed:

And where product differentiation results in varia-
tions in the product itself -- and not just in its
"image" -- one must ask whether the aim of an
affluent society is to produce the largest possible
quantity of a standardized product at the cheapest
possible cost or to offer an array of differing
products that please our palates, admitedly at
somewhat higher costs. Few consumers in a rich
society would prefer an inexpensive uniform to more
expensive but highly individualized clothes. From
this point of view, even the wastefgl parade of

car styles has a certain rationale. 6

It is unarguable that consumers respond positively to a

wide range of choices in the market place. The managerial

17

strategy of market segmentation is a pragmatic response

to this situation.

It is also true that such phenomena as product
proliferation, expansion of shopping opportunities,
scrambled merchandising, and a mass of advertising
messages create special problems for consumers. Limita-
tions faced by consumers in time, energy, and informa-
tion processing capabilities place barriers in the way
of utility maximization as postulated by economists
and optimal need satisfaction as seen by marketers. If
choice is generally viewed as good, it is also true that
some observers see the modern consumer as suffering from
too much of a good thing. Birmingham states flatly:

"The increasing complexity of the modern market place has
rendered the unaided individual almost defenseless against

modern merchandising techniques."27

"Problem Statement

 

The benefits of product—market competition to the
consumer are clearly not obtained without cost. Con-
spicuously absent has been any attempt to assess whether
the benefits in terms of customer satisfaction are worth
the cost. Marketers typically accept a positive accounting
of the issue a priori or ignore it completely.

This study addresses a question which is a subset
of the larger problem. No attempt is made to build in

or evaluate costs. The investigation does deal, however,

18

with one aspect of the benefit side of the equation. The

research problem is to determine whether it is possible
to identify through measures of satisfaction whether
consumers gain by virtue of the availability of purchase
options. Stated more narrowly the question is this:
Is customer satisfaction independent of brand choice?
The term satisfaction is used here in a manner
consistent with Hunt's conclusions after reviewing several
different conceptualizations:
Satisfaction is a kind of stepping away from an
experience and evaluating it. You may have a
pleasurable experience and then be satisfied as you
evaluate that experience. If it were not a stepping
away, we'd call it happiness or pleasure or a warm
feeling or something like that. Satisfaction is not
the pleasurableness of the experience, it is the
evaluation rendered that the experience was at least
as good as it was supposed to be. One could have a
pleasurable experience that caused dissatisfaction
because even though pleasurable, it wasn't as pleas-
urable as it was supposed or expected to be. So
satisfaction/dissatisfaction isn't an emotion, it's
the evaluation of an emotion, and as such it be-

comes a quasi-cognitive construct and we woulg expect
the laws of judgment to affect satisfaction.

The prediction based on marketing theory is fairly
straightforward. The relevant theory is segmentation
theory which is premised on a conception of the market
similar to Alderson's, that is, a market in which both
supply and demand are heterogeneous. Such a market accord-
ing to Alderson "can be said to be cleared by information.
The consumer must know that the appropriate segment of
supply is precisely what he specified."29 The implication

is that because of unique personal characteristics or

l9

motivation structure the consumer will be dissatisfied
with any product other than the differentiated product
specified. Therefore, satisfaction would be associated
with rather than independent of brand choice.

Wendell Smith recognized the limitations of this
view in his early work on market segmentation. He saw
that "some divergence in demand is the result of shopping
errors in the market. Not all consumers have the desire or
the ability to shOp in a sufficiently efficient or rational
manner as to bring about selection of the most needed or

30 What is not known is how

most wanted goods or services."
frequently this occurs or how best to detect it when it
does happen. It would be cause for concern with the per-
formance of the economy if an increasing trend in "shopping
errors" were detected.

If, however, there were no discernible differences
in levels of satisfaction with different brands, one might
argue that the differences among brands were trivial, and
the justification for multiple brands must rest on other
grounds. To be more precise, this interpretation might

arise if the predictors of satisfaction do not differ among

brands.

Methodological Summary

 

A brief summary is presented here of the material
given a detailed treatment in Chapter III. The core of

the analysis is a comparison of satisfaction prediction

20

functions for two selected brands within a product

class. The product class selected for study was academic
majors within the College of Business of Michigan State
University. The two brands chosen for analysis were majors
in accounting (ACC) and hotel and restaurant management
(HRI) .

The population of interest was defined as individ-
uals who had attained senior status in either ACC or HRI
at the time of data collection, which was during the Fall
1977 and Winter 1978 terms. A complete census of the
population was attempted. Questionnaires were distributed
with the c00peration of instructors in classes known to
contain high concentrations of seniors in the two majors.
Completed questionnaires totaled 306 of which 180 were
from ACC majors and 126 were from HRI majors.

The instrument itself was designed to gather two
principal kinds of information. The first was data to be
employed as the dependent or criterion variable. Twelve
Likert-type scale items were used to tap the satisfaction-
with—choice-of-major construct. The twelve items were
reduced to a six item scale through the use of cluster
analysis and reliability analysis.

The second main category of information collected
was intended for use as independent or predictor variables.
Three kinds of personal characteristics data composed this
category. They were demographic, self-image descriptors,

and activities, interests, and opinion (AIO) statements.

21

The analysis centered on the self-image and A10 data
whfljmtotaled sixty four items. These items were sub-
jected to cluster and reliability analysis which reduced
the number of items to forty-four and assigned the forty—
four items to eight clusters. These eight clusters

then became the predictor variables of central interest
in the subsequent analysis. Both criterion and predictor
variables were developed using intermingled data from all
306 cases.

The predictor variables had their origins in a
principal components factor analysis employing oblique
rotation. Multicollinearity was then a possibility.
Pearson product-moment coefficients were computed for all
combinations of the eight clusters in order to check for
multicollinearity.

The first phase of the analysis itself was the
computation of correlation coefficients for satisfaction
and each predictor variable for both ACC and HRI students.
In order to test whether the coefficients for satisfaction
and a given predictor variable differed significantly
between ACC and HRI students, it was necessary to use the
Fisher r to Z transformation.

Stepwise multiple regression procedures were
employed to develop predictive equations for both ACC and
HRI students. Regression weights were checked to see which
predictor variables were significantly different from zero.

In the critical test, regression weights for variables

22

significant in at least one of the equations were tested
for significant differences.

A third regression equation was computed with the
combined responses of ACC and HRI students. The purpose
was to determine whether the conclusions regarding pre-
dictive relationships were altered by intermingling data
from the two groups.

A working hypothesis underlying this study, was
that majors in ACC and HRI were perceived as dissimilar
by the subjects. This assumption was tested with t-tests

of the differences between means.

Limitations

 

The principal limitations of this study lie in the
research design employed and the particular purchase
situation studied. The design had the capability of
providing partial data at best in assessing the net gain
or loss resulting from product differentiation. Assuming
that the research hypothesis is supported, cost inputs are
required to make a full accounting of affects.

A longitudinal design would have advantages over
the cross-sectional design used. With a cross-sectional
design it is impossible to say anything about purchase
*motives, their stability, and their relationship to evalua-
tive criteria. It is also likely that more light can be
shed on the importance of personal characteristics in

evaluative criteria if the design were ambitious enough

23

to allow comparisons with other possible factors of im-
portance, e.g., product performance.

The choice of academic major as the purchase
decision to study has advantages discussed elsewhere.
The disadvantage is that the great bulk of consumer pur-
chases are very different in terms of expense, time re—
quirements, personal commitment, and so on. Consequently,

the results cannot be generalized.

FOOTNOTES: CHAPTER I

_ 1Webster's New World Dictionary, ed. David B.
Guralink (2d CoIl, ed,; Cleveland: World Publishing Co.,
Inc., 1974), p. 1265.

 

2Confucius, The Confucian Analects, bk. 7:36, quoted
in John Bartlett, Bartlett's Familiar Quotations, ed.
Emily Morison Beck (14th ed.; Boston: Little, Brown and
Company, 1968), p. 71b.

 

 

3John Stuart Mill, Utilitarianism (1863), quoted in
John Bartlett, Bartlett's, p. 619b.

 

 

4Charles P. Steinmetz, "The world Belongs to the

Dissatisfied," The American Magazine 85 (May, 1918), p. 39.

 

5Jack L. Engledow, "Was Satisfaction a Pig in a

Poke?" Business Horizons 20 (April, 1977), p. 87.

 

6Thorstein Veblen, The Theory of the Leisure Class
(New ed,; New York: Modern Library, 1918), p. 98.

 

7Lawrence Nabers, "Veblen's Critique of the
Orthodox Economic Tradition," Thorstein Veblen: A Critical
Reappraisal, ed. Douglas P. Dowd TIthaca, N.Y.: Cornell
University Press,l958), p. 111.

 

 

8 Martin Pfaff, quoted in H. Keith Hunt, "CS/D -
Overview and Future Research Directions," Conceptualization
and Measurement of Consumer Satisfaction and Dissatiisfaction,
ed. H. Keith Hunt (Cambridge, Massachusetts: Marketing
Science Institute, 1977), pp. 485-486.

 

 

9Ralph L. Day, "Toward a Process Model of Consumer

Satisfaction" Conceptualization and Measurement, p. 177.

 

10Edwin Mansfield, Microeconomics (New York:
W. W. Norton & Co., 1970), pp. 30-31.

 

lKelvin Lancaster, Consumer Demand (New York:
Columbia University Press, 1971), p. 8.

 

24

25

lzIbid.

13Ibid., p. 7.

14Ibid.

 

15Stanley C. Hollander, "Measuring the Cost and
Value of Marketing," MSU Business Topics (Summer, 1961),
p. 22.

 

16Wroe Alderson, "The Analytical Framework for

Marketing," Proceedings: Conference of Marketing Teachers
from Far Western States, ed. Delbert Duncan (Berkeley:
University of California, 1958), p. 27.

 

 

17Ibid.

 

lBIbid., p. 28.

19M. Wayne Delozier and Arch G. WOodside, Marketing
Management (Columbus, Ohio: Charles E. Merrill Publishing
Company, 1978), p. 6.

 

 

ZOIbid.

21Phillip Kotler, Marketing Management: Analysis,
Planning, and Control (3rd ed.; Englewood Cliffs, N.J.:
Prentice-Hall, Inc., 1976), p. 18.

 

 

22Engledow, p. 90.

23Mansfield, p. 422.

24Wroe Alderson, Dynamic Marketing Behavior (Homewood,

Illinois: Richard D. Irwin, Inc., 1965), pp. 25-32.

 

251bid., p. 26.

 

26Robert L. Heilbroner, The Economic Problem (3rd.
ed.; Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1972),
p. 562.

 

26

27 . . .
Robert L. Birmingham, "The Consumer as King: The

Economics of Precarious Sovereignty," Consumerism: Search
for the Consumer Interest, ed. David A. Aaker and George
8. Day (2nd. ed.; New York: The Free Press, 1974), p. 183.

 

 

28H. Keith Hunt, "CS/D - Overview and Future
Research Directions," Conceptualization and Measurement,
pp. 459-460.

 

29Alderson, Dynamic Marketing Behavior, p. 30.

 

30Wendell R. Smith, "Product Differentiation and
Market Segmentation as Alternative Marketing Strategies,"
Journal of Marketing 21 (July, 1956), p. 4.

 

CHAPTER II

LITERATURE REVIEW

eggs—2.

A researcher studying the post-purchase phase
of consumer behavior may find that several existing lines
of research relate to or offer insight into the problem
at hand. Among the areas of research most likely to have
a direct impact are those related to brand loyalty, brand
switching, cognitive dissonance, satisfaction and dissat-
isfaction, and complaining. More general studies
dealing with pre-purchase aspects of the choice process,
learning theory and market segmentation may also be rel-
evant, but the relationship would typically be more in-
direct. Indeed, the link simply grows more tenuous as
other marketing research areas are scanned. It is im-
possible, however, to divorce post-purchase phenomena
totally from the pre-purchase activities of buyers and
sellers and from the act of purchase itself. Consequently,
the researcher is obliged to become increasingly more
selective in reviewing the literature as the search pro-

ceeds further from the area of most direct concern.

27

28

Post-purchase Processes

 

This study is most closely associated with the
investigation of consumer satisfaction and dissatisfaction.
Empirical studies in this area have a very recent history
in the marketing literature although proprietary studies
have undoubtedly been numerous over the years. The
absence of studies of satisfaction as well as the direction
and sc0pe of post—purchase research prior to the seventies
is suggested by Engel, Kollat, and Blackwells' (1973) re-
view of the evidence.

Their chapter titled "Postpurchase Processes" is
the shortest chapter in the book, a graphic demonstration
of the relative lack of interest in this part of the buy-
ing process. The content of the chapter also illustrates
the dominant perspectives employed in early post-purchase
research. The general model used specifies two outcomes:
(1) further behavior and (2) postpurchase evaluation.

The discussion of behavior focuses on finances, installa—
tion, use, and complementary goods and services. More
germane for this study is their discussion of postpurchase
evaluation which is divided into three categories: (1) re-
ward and satisfaction, (2) attitude change, and (3) post-
decision dissonance. The subject attracting the most
attention was dissonance.

Festinger's (1957) theory of cognitive dissonance
sparked substantial interest in marketing. Underlying

coqnitive dissonance theory is the broader psychological

29

perspective of consistency or balance theory which is

premised on "the notion that the person tends to behave
in ways that minimize the internal inconsistency among
his interpersonal relations, among his interpersonal
cognitions or among his beliefs, feelings, and actions."1
Consistency theory is compatible with both the simple,
rational man of economics and his more sophisticated,
complicated brother residing in marketing, and it is not
surprising that Festinger's theory stimulated wide
interest in its marketing implications.

The theory postulates that the psychological tension
created by dissonant cognitions is a motivating force for
the restoration of consonance or consistency. Since
decision making requires the selection among alternatives,
including the rejection of attractive ones, it almost
always leads to dissonance. Purchase behavior, of course,
falls under the heading of decision making. Much of the
body of dissonance theory research has focused on the
extent to which dissonance is a pervasive phenomenon.

Summaries and reviews of the research findings
have been published by, among others, Aronson(l968),
Chapanis and Chapanis (1964), and Margulis and Songer
(1969). Engel, Kollat, and Blackwell (1973) extracted a
number of antecedent conditions for the occurrence of
dissonance from this extensive body of literature. It

was most likely to occur at a significant level when:

30

(l) a minimum level of disanance which is
tolerable is surpassed. 3

(2) the action is irrevocable.

(3) unchosen alternatives have desirable
features. 5

(4) several desirable alternatives exist.

(5) the decision engenders a high level of
committment because of its psychological
importance.

(6) each of the available altegnatives has some
desirable unique features.

(7) the decision process i3 voluntary and free
from outside pressure.

Presumably, dissonance would be greater if several of
these conditions exist simultaneously. It is understandable
that Engel, Kollat, and Blackwell would conclude that
"postdecision dissonance is largely confined to extended
problem-solving situations."9 The extended problem-
solving situation obviously contains more of the ante-
cedent conditions than occur with limited problem solving
or routinized response behavior.

Interest in cognitive dissonance within marketing
has waned during the seventies in comparison with the
volume of research published during the sixties. A
number of reasons may be advanced for the decline. Among
them are:

(l) diminishing marginal returns from research
effort

(2) frequent inability to rule out alternative
explanations

(3) dissonance reduction may occur in several ways,
'many of which are not threatening to sellers

(4) dissonance is essentially a short-range
phenomenon and therefore not a major concern

(5) the antecedent conditions normally do not exist
in the packaged goods area where behavioral re-
search has a longer history and greater financial
support

(6) interest has shifted to more topical areas of
concern.

31

The possibilites are not mutually exclusive.

Although a decline in interest in cognitive
dissonance in marketing situations is evident, the idea
cannot be ignored. It frequently appears, as it should,
as a possible or alternative explanation for aspects of
other processes being investigated. Despite the attention
it has received, cognitive dissonance theory has not been
integrated into the more general theoretical work in
consumer behavior. It is not indexed in Howard and
Sheth's (1969) early theoretical work which emphasized
information processing in a learning theory context and
has since become a major reference work in the consumer
behavior literature. Neither is it included in Howard's

(1977) more recent applications-oriented book.

Complexity Theory

 

Consistency theory itself is increasingly being
viewed as too narrow to provide general explanation for
consumer motivation although its applicability to studies
of the post-purchase phase has not been diminished. The
inadequacies of consistency theory have been most evident
to those seeking a better understanding of consumer
reactions to new products, other marketing innovations, and
novelty seeking in general. They have turned to what have
been called "complexity theories" for additional insight
into these processes (McGire, 1972; Venkatesan, 1973).

The most ambitions advocate of a central role for

32

complexity theory is Scitovsky (1976). The extent of

his committment is indicated by his statement that "variety
is not the spice of life, it is the very stuff of it."10
His book is of particular interest because it deals not
just with the psychology of motivation but with evaluative
judgments as well. This is indicated by the subtitle:

An Inquiry into Human Satisfaction and Customer Dissatis—
faction.

The sc0pe of his essay clearly extends beyond the
short-range effects of cognitive dissonance and, indeed,
beyond the market—related events of principal interest to
marketers. His argument might be interpreted as a spec-
ulative extrapolation of selected psychological findings.
Consequently it is not proposed as a general theory of
consumer behavior but is offered as an attempt to integrate
empirical results of studies in psychology with the largely
non-empirical consumer behavior propositions in economics.
Scitovsky's background in economics required him to study
an entirely different literature in preparing his analysis,
and some omissions were to be expected. Among the
ommissions, however, were Katona's work and the entire
marketing literature of consumer behavior. Despite these
limitations, Scitovsky's essay is of interest because it
deals at a very general level with a subject of emerging
interest in marketing.

Scitovsky reports the results of studies in which

citizens of various countries were asked to rate their

33

own happiness. His conclusion was that one's chances of
happiness seems to improve with a rise in rank on the
national income scale. If, however, everybody's income
is rising, no change in happiness is reported. He pro-
posed several possible explanations involving the satis-
faction of status, work satisfaction, enjoyment of novelty,
and addiction. He also emphasized that market exchanges
and economic conditions are only part of the happiness
equation.

The theme that evolves from Scitovsky's book is
that income level and purchasing power by themselves are
no guarantee of happiness. In fact, the mass production
and specialization which characterize the advanced
economies and are necessary for a high standard of living
‘may interfere with the processes which create human satis-
faction. The search for satisfaction in work is consistent
with the Puritan heritage but consumption activities have
no such legitimating foundation in American history.
Consumption skills, therefore, are underdeveloped just as
culture is undervalued. In the principal consuming choice
which we face, the choice between comforts and pleasures,
we lack skill in ordering and skill in consuming. Scitovsky
argues that "when we behave rationally, we seek the best
available choice or the best compromise between mutually
incompatible considerations; one can go very wrong, however,
in identifying the available choices and the relevant con-

siderations."11 Additionally "we gradually dismantled

34

the Laws of God and came to believe in man as final

arbiter of what is best for himself."12

In doing so not
only was research discouraged and delayed but the belief
itself is arguably indefensible.

The merits of the arguments notwithstanding,
economists and marketers alike have operated upon the
assumption that man is the best judge of his own interests
or, more moderately, that the distance between buyer and
seller prevents the seller from being a good judge. The
economic incentives inherent in trade prevent the seller
from being an unbiased judge. Also, it has been argued
that marketers responsibility should stop with the
development and offering of a product/service mix. The
emergence of an institutionalized consumer protection
movement has made this largely an academic issue.

A related issue involves the extent to which the
assumed rationality of the consumer yields decisions not
only consistent with his welfare but meet a higher standard
of satisfaction per se. Scitovsky argues that for a
variety of reasons rationality provides no guaranty of
satisfaction. Implicit in his conclusions is that in-
dividuals, particularly in the United States, exhibit
tendencies to misunderstand their needs and to misinterpret
the possible means of satisfaction.

The temptation exists to view the possible
marriage of consistency theory and complexity theory as a

major step toward a general theory. The post-purchase

35

studies of the seventies have not, however, employed any
framework general enough to include both types of phenomena.
Brand loyalty studies might be viewed as investigations

of an aspect of consistency, but its most common expres-
sion, cognitive dissonance theory, rarely appears in

recent studies of posterior evaluations. The orientation
of the general model building attempts in consumer behavior
toward learning theory and information processing makes
them ill-suited as a mechanism to integrate consistency

and complexity theories.

Attitude Change

 

A second category of postpurchase evaluation in
the Engel, Kollat, and Blackwell book (1973) attitude
change, is discussed briefly and adds very little to
knowledge of post-purchase processes. The issue is dis-
cussed in the context of encouraging customers to engage
in attitude discrepant behavior by means of cents-off
coupons or free samples. Consequently, it is narrower
and of less value than the third category of their dis-

cussion of postpurchase evaluation, reward and satisfaction.

Reward and Satisfaction

 

 

This section is illuminating both for what it says
and for what it doesn't say. The authors report that
"thousands of studies in the learning of literature [gig]
affirm that a response is reinforced to the extent that

it serves to reduce the motivation drive state."13

36

Therefore, positive reinforcement of an act of purchase

and use will increase the probability of a similar pur—
chase. The mathematically oriented models of brand loyalty
are built upon this proposition.

Interesting questions arise, however, when atten-
tion is focused on what is meant by reinforcement or re-
ward when the situation is one involving consumer judgment.
The questions involve not only consideration of the
standards against which comparisons are made but the nature
of the decision itself, the decision environment, and
the timing of the evaluation.

Review of the satisfaction related studies
published in the marketing literature and the extent to
which they address these issues will be deferred momentarily.
The empirical nature of many of these studies requires a
somewhat narrow focus. One general, almost global, in—
quiry into various aspects of satisfaction will be con-
sidered first. This is the work of William Leiss (1976).
Leiss' essay is a provocative piece in which marketers will
find some conventional, unconventional, and radical ideas.

Although Leiss approaches his subject matter from
a slightly different direction than Scitovsky, he arrives
at the similar conclusion that "individuals are led to
misinterpret the nature of their needs and to misunderstand
the relationship between their needs and the ways they may

14

be satisfied." The motivations that underlie the two

books are, however, clearly at odds. Scitovsky is

37

vulnerable to being labeled as an epicure. He respects
the yearning for new things as a source of progress,
civilization, and satisfaction. Leiss, on the other
hand, recommends that the "present rate of innovation
and productive applications should be reduced consider-

15

ably" and that we need "a reorientation of our social

and economic behavior away from dependence on a high-

consumption life-style"16
Leiss argues that an understanding of the possibil-

ities for satisfaction that exist in highly developed

economies or in his terminology, "high-intensity market

settings," are obscured by debates about wants versus

needs, need hierarchies, authentic versus artifical needs,

biological versus cultural, or simple versus complex

needs. To him the main point is that "all human needs

are routed indirectly toward sources of satisfaction by

the agency of culture."17 He views modern society as

the first large-scale attempt to strive for stability and

authority through the achievements of economic production

and the satisfaction of needs rather than inherited

privilege or traditional associations. Necessary

conditions included production and exchange with cap-

italism providing catalytic and legitimating functions.

Inevitably, there was an emergence of "the necessity of

stimulating and managing the growth of the sphere of

18

consumption [marketingJP

A central position in the Leiss essay is held by

38

his statement that "high—intensity market society
encourages its citizens to orient their search for the
satisfaction of their needs more and more exclusively
toward consumption activities in part by neglecting other
possibilities for individual self—fulfillment."19 He
extended this into a version of Greshams Law applied to
consumption: "Wants for ever greater numbers of
commodities tend to depreciate all types of desires that
are not dependent upon the consumption of things"20
Leiss admits that it is difficult to develop an under-
standing of what satisfaction of needs really means, and
he contends, additionally, that the "meaning of satisfac-
tion of needs will become increasingly ambiguous as the
consumer pursues the will-o'-the-wisp of gratification
through a jungle of commodities."21
Given the sheer number of items available to
consumers, Leiss argues that even the most unbiased
package of information about goods is of little assistance.
As a consequence choice is little more than volunteering
for a grand experiment to determine which claims are most
valid. Leiss appears indifferent to the direct implica-
tions of this belief. Indeed, he fails to develop the
effects upon satisfaction in situations where purchase
is infrequent and the only claim that can realistically
be evaluated is for the product that is purchased. He

is more concerned with the implications of his statement

that "the realm of needs becomes identical with the range

39

of possible objects."22 Product proliferation is followed
by need proliferation and, inevitably, by increased
likelihood of dissatisfaction. This argument is con-
sistent with that developed by anthropologist Marshall
Sahlins (1972) in a somewhat polemical essay on economic
anthropology. Sahlins views scarcity as a relation be-
tween means and ends, that is, as a situation in which
the only reliable guide is felt or perceived scarcity.
Thus, it is more a psychological phenomenon than an
economic one. It follows that the volume of goods and
advertising messsages in the United States combine to
produce scarcity unapproached in degree anywhere else.
There is a kinship between this position and the ex-
pression of similar thoughts by Daniel Bell, (1975) an
observer of the social scene who occupies a position more
in the mainstream.

Sahlins provides an interpretation of anthro-
pological evidence, albeit a minority viewpoint, that
primitive hunter-gatherer societies maintained an adequate
level of satisfaction with a working day less extensive
than our own rather than occupying all their waking hours
seeking subsistence as is commonly believed. Leiss'
contention that material abundance provides no sense of
contentment, that progress is partially illusory, is
softened by a paradoxical statement about a possible
solution. A concern underlying his essay is his view

that the advanced economies are ransacking the natural

40

environment in a search for resources that is ultimately
futile. By the same token the inevitable shortages of
materials and energy are not catastrophic because the
high-intensity market setting provides a mechanism through
the education of desires to substitute images for things.
Those persuaded by his analysis can breathe easier about
human survival. It would seem, however, that in deference
to his arguments about the relation between goods and
satisfaction one must conclude that human society will not
find satisfaction in images either.

Despite the growing literature in marketing under
the rubric of macro-marketing, there is no work directly
comparable to that of Scitovsky or Leiss, although some
evolution in that direction is discernible. The marketing
literature over the past ten to twelve years has reflected
a gradual transition in this area from concern with narrow,

short-range outcomes to more general, long-range evaluations.

Consumerism

 

The social phenomenon variously described as
consumerism or the consumer movement provided the impetus
behind much of the current literature relating to consumer
satisfaction. Consumerism has been described by one
observer as "a social movement seeking to augment the
rights and powers of buyers in relation to sellers."23

In this view it had to do with the reactions of consumers

to institutional business practices such as advertising,

41

pricing, warranties, etc. and economic structures and
motivations in general rather than displeasure with
dealings involving specific firms and specific situations.
It was to be expected, then, that a general interest in
diagnosis and prognosis would arise. A number of works
appeared that addressed questions related to the origins
or causes of the movement, its durability, its direction
and its impact on business and government (Aaker and Day
1974; Broffman, 1971, Cravens and Hills, 1970; Gaideke and
Etcheson, 1972, Greyser and Diamond, 1974; Hermann, 1970;
Kotler, 1972; Monsen, 1973; Murray, 1973).

The general, historical orientation which char—
acterized these works suggested a variety of areas of
inquiry amenable to empirical research. One such line of
research focused on the extent of sympathy for the consumer
movement, identification of issues of greatest concern,
and profiling those most attracted to the movement
(Barksdale and Darden, 1972; Diamond, et a1, 1976; Granzin
and Grikscheit, 1974; Grikscheit and Granzin, 1975;

Hustad and Pessemier, 1973).

Complaining Behavior

 

A second line of research developed around the
complaining behavior of consumers. Research in this area
has the advantage of dealing with a visible, tangible
manifestation of dissatisfaction. While general con-

sumerism concerns are of interest to public agencies in

42

the formation and administration of public policy,
complaining behavior is arguably a more important man-
agerial issue because of the need to interpret and respond
to individual complaints.

One segment of these studies has a strong man-
agerial flavor. Kraft's (1977) work supports the ex-
pectation that attitude change is negative and repatronage
intentions diminish markedly when the retailer is blamed by
the consumer for resulting dissatisfaction. Other studies
have developed descriptive data about the rate and nature
of corporate responses to consumer complaints (Resnik, et a1,
1977; Boschung, 1976; Kendall and Russ, 1975; Pearson, 1976:
Thomas and Shruptine, 1975). The conclusion generally
drawn is that complaints, either unsolicited or associated
with warranties, are an important source of feedback about
the operation of marketing programs. Firms can and should
use this data in a positive fashion to improve competitive
performance, but too few take advantage of the opportunity.

Another branch of the literature on complaints
emphasizes better understanding of the process that
generates complaints. The focus is more basic than applied.
Understanding and prediction are central elements and
control is peripheral. The descriptive approach is typified
by the Best and Andreasen (1976) study. The authors
perceived the complaint process as composed of problem
perception, voicing of complaints, and achievement of

results. They examined the relationship between each part

43

of the process and problem type, purchase type (34
categories), and hOusehold type. Best and Andreasen found
that problem recognition and complaining were lower in
low socio-economic groupings than in higher groups. A
more noteworthy finding, however, was that purchase
shortcomings lead to complaints in only about one third
of all instances where dissatisfaction was recalled.
Additionaly, only a very small fraction of consumer
problems move into the domain where third party institutions
such as Better Business Bureaus or courts are involved.

This is an important point in light of the in—
creasing attention being paid to complaints as an indicator
of the general level of satisfaction among consumers.
Critics of this practice have been able to point to both
increasing numbers of official outlets for complaints and
more forthright policies of complaint solicitation as
tentative explanations of the growing volume of documented
complaints. The argument is that there is no evidence of
any change in the level of dissatisfaction, only that the
number of official expressions has increased. The absence
of good historical benchmark data renders this argument in-
conclusive. It is true, however, that only a very small
percentage of transactions ever result in complaints.

A stated objective of the Best and Andreasen study
was "to see how well the complaints households make to
businesses and complaint handling institutions represent,

in subject matter and number, the problems households

44

perceive about goods and services."24 Their findings

with respect to the number of complaints lead to the
conclusion that complaints to businesses are not represent-
ative and those to complaint handling institutions are
even less so. Other investigators whose research dealt
with this issue as a secondary interest have generated
similar results (Day and Landon, 1976; Mason and Himes,
1973; Thomas and Shruptine, 1975; Wall, et a1, 1977;
Warland, et a1, 1975). With respect to the subject matter
of perceived problems, Best and Andreasen's findings
support the interpretation that complaints are only partly
representative. The evidence is that costly problems, not
surprisingly, are more likely to elicit complaints.

The need to interpret and respond to complaints
has also focused attention on complainers, again to deter-
mine if they are representative of the general population.
Implicit in this interest was the existence of a hypo-
thetical chronic complainer. The identification of such
an individual would probably lead to minimization of the
importance of complaints. To this point no such chronic
complainer has been isolated.

Investigations of complainer characteristics have
run the full range of individual attributes commonly
associated with segmentation studies. Consistent with
purchaser segmentation studies, demographic data have been
associated with complaining behavior in irregular fashion.

The closest thing to a pattern in the repeated attempts

45

to link complaining behavior with demographic variables
is a frequent finding that complainers are younger, have
more income, and have more education that non-complainers
(Landon and Emery, 1974; Liefeld, et a1, 1975; Warland,
et al, 1975; Thorelli and Puri, 1977). Wall, Dickey and
Talarzyk (1977) found a significant relationship for in-
come only. In contrast Kraft (1977) found associations
between complaining behavior and occupational, racial,
marital, and age characteristics but none for education
and income. Granbois, Summers, and Frazier (1977) found
virtually no association at all between the tendency to
complain and demographic variables across the five problem
areas which they investigated.

Skeptics of the use of demographics in purchase
segmentation studies have long argued that the relatively
easy access to and measurement of demographic data is in-
sufficient reason to expect them to be associated with
purchase behavior. To the extent this is true for pur—
chase behavior, it is also true for complaining behavior.
Although demographics are routinely included, a number of
studies have tested hypothetical mechanisms which lead to
complaining behavior. Almost all of the studies deal with
psychological processes in some respect.

While the cognitive dissonance processes discussed
earlier suggest a theoretical explanation of complaining
behavior, there are no empirical studies to assist in

evaluating the theory. Most studies, by way of contrast,

46

deal in some degree with the role of expectations.
Miller (1977) drew heavily on level of aspiration theory
(Lewin, et al, 1944) in postulating fimnrtypes of expecta-
tions which he labeled ideal, expected, minimum tolerable,
and deserved. These reflect respectively what an in—
dividual feels performance "can be", "will be", "must be",
and "should be". Similar concepts of predictive expecta-
tions (straightforward estimates) and normative expectations
(performance that ought to be delivered) were tested for
their relationship with complaining behavior by Granbois,
Summers, and Frazier (1977). They found, much to their
surprise, that neither had much of an association with
complaining. Furthermore, their evidence suggested that
those holding less idealistic norms for problem frequencies
had a tendency to complain more. The most striking finding
in this study was that the variable with the strongest
association with complaining behavior was the individual's
perception of the stores willingness to provide a remedy.
The implication is that consumers are more likely to complain
when the effort is likely to be fruitful. This study dealt
only with the relationship between customers and retail
stores. Conflict avoidance, therefore, might also be a
factor in the absence of complaining.

Kraft's (1977) study, in contrast, focused on the
distinguishing features of individuals who had gone
beyond the buyer-seller interaction and had filed complaints

with a District Attorney's office. He found strong support

47

for his hypothesis that complainers are more likely to
report a perception that purposeful deception or intent to
defraud on the part of the seller was the cause of the
problem. An earlier attempt to determine whether
differences between complainers and noncomplainers was
related to whether locus of control was internal or ex-
ternal (Collins, 1974) found only small differences.
Attribution theory does offer, however, a largely unex-
plored potential to improve insight into these processes.

In a study which employed a variety of predictor
variables, Wall, Dickey, and Talarzyk (1977) found that
internal influences and personal characteristics were the
best predictors of the tendency to complain. Product
(clothing) performance problems, and satisfaction levels
were not good predictors. In another recent study
(Zaichkowsky and Liefeld, (1977), however, researchers
were unable to distinguish complaint letter writers from
a general sample of the population on the basis of per-
sonality type.

Part of the difficulty might exist in the likeli-
hood that both complaining and non—complaining groupings
are relatively heterogeneous. Warland (1977) provides
support for this view in a study which grouped consumers
using clustering techniques. He found two types of
complainers which he labeled concerned and unconcerned and
three types of noncomplainers which he labeled unconcerned,

concerned, and uniformed. It seems likely that information

48

is lost when subcategorizations such as these are ignored.
The confusion arising from diverse and often con—
tradictory findings about complaining behavior may also be
related to the absence of an integrating framework.
Landon (1977) has proposed a phenomenological model with
the intent of organizing and synthesizing existing know—
ledge as well as suggesting important conceptual linkages.
In Landon's model complaint behavior is a function of dis—
satisfaction, the importance of the dissatisfaction, the
benefit from complaining, and personality. With the
exception of personality, each of these predictor variables
is also treated as a criterion variable. Dissatisfaction,
for example, is viewed as a function of the discrepancy
between expectations and perceived performance. Importance
is viewed as a function of product cost, search time,
physical harm, and ego involvement. The third variable,
benefit from complaining, is perceived as a function of
the expected value of the payoff from complaining less the
cost of complaining. Both of these variables are again
treated as criterion variables. The expected payoff from
complaining is a function of the importance of dissatisfac—
tion and the nature of the defect while the expected cost
of complaining is a function of the firm's image, complain-
ing experience, and the nature of the defect. Not too
surprisingly, demographic variables are not proposed as
important considerations, although further elaboration of

the model would probably lead to their inclusion.

49

Many of the studies cited previously yielded results
consistent with Landon's model, but the nature of the
variable linkages remains uncertain because there have
been no studies of sufficient scale and complexity to
control for even a few of the possible influences. Al—
though numerous studies of complaining behavior deal with
satisfaction only implicity, there is an increasing
tendency to investigate the two subjects jointly in a
manner consistent with Landon's model. Satisfaction is
clearly a factor in complaining, but the subject has a
research tradition, albeit brief, which is independent

of any concern with complaints.

‘ Satisfaction

 

The first study reported in the marketing litera-
ture to purportedly address the measurement of satisfaction
and its relationship with key variables was Cardozo's
(1965) experiment involving ballpoint pens. His article
was mistitled since no measurement of satisfaction itself
was reported. Rather, product performance ratings were
presumed to be the equivalent of satisfaction measurement.
Although his study was later criticized on methodological

grounds,25

his conclusion that effort expended in purchas-
ing positively influenced product evaluation led him to the

provocative conclusion that "customer satisfaction, then,

may be more a global concept than simply product evaluation".

Cardozo's study inspired Olshavsky and Miller (1972)

26

50

to undertake an experiment with a broader set of hypoth-
eses regarding the effects of disconfirmed expectancies.
The outcomes they were interested in were product per-
formance ratings. and perceptions of product quality
rather than satisfaction per se. They concluded, inter-
estingly, that engaging in puffery or overstatement of the
quality of a complex product (a reel tape recorder) led

to more favorable evaluation (and presumably satisfaction)
than did understatement but that ethics and consumerism
direct sellers away from such practices.

Anderson (1973) continued this line of research
with another ballpoint pen experiment which dealt with
disconfirmed expectancies in light of the outcomes pre-
dicted by four psychological theories: (1) cognitive
dissonance (2) contrast (3) generalized negativity, and
(4) assimilation-contrast. Satisfaction was again treated
vaguely since perception of product performance was the
dependent variable. His findings supported assimilation-
contrast theory in which the size of the disparity between
expectations and product performance is critical. Large
disparities, i.e., those which fall within the consumers
latitude of rejection, are subject to contrast effects,
but small disparities, i.e., those within the latitude of
acceptance,are subject to assimilation effects. Anderson
also produced evidence that was supportive of Cardozo's
suggestion that effort is related to positive evaluation.

Anderson obtained a positive relationship between evaluation

51

and the amount of product information provided. He ex-
plained it in terms of committment to the product, but it
should be noted that information search was not incorp-
orated in the design.

The most recent study in this line of research
(Oliver, 1977) tended to reinforce the conclusions of
the earlier research. Swan (1977) investigated the im-
portance of expectations in the setting of a field ex-
periment involving the opening of a new department store.
He concluded, unsurprisingly, that satisfaction is re-
lated to fulfillment of expectations, but he also suggested
that confidence may act as a moderator variable in the
process of evaluation.

Miller (1977) argued very persuasively, however,
that simplified measures of expectations may act to screen
or mask fundamental evaluative processes. He contended
that it quxssense to think of four "types" of expecta-
tions: ideal, expected, minimum tolerable, and deserved.
Unless the type of expectation is dealt with explicitly,
there may be no way of determining how the types interact
or which may be most relevant for the determination of the
expectations-performance gap that is frequently viewed as
critical in the evaluative process. With the exception of
the Granbois, Summers and Frazier (1977) study cited pre-
viously, this expanded notion of expectations has not been
explored empirically. Their study involved a limited

adaptation of Miller's framework but employed it in a

52

study of complaining behavior rather than satisfaction
per se.

A number of studies referenced in this review
were published as part of proceedings from a conference
which marked an expansion of interest in the study of
satisfaction and indicated the multiple directions of
current research.27 One obvious change is a direct con-
frontation with issues previously dealt with only im-
plicitly or at best by assumption. The operational
simplifications employed in viewing product ratings as
measures of satisfaction and complaints as dissatisfaction
are recognized as impeding enhancement of understanding
of the processes involved.

The 1976 conference emphasized conceptual issues
and measurement problems. In a follow-up conference in
1977 complaining behavior was added as a focal topic, and
the emphasis shifted to empirical studies.28 A number of
these studies referenced with respect to findings on
complaining behavior also produced results of direct
interest on satisfaction. Wall, Dickey, and Talarzyk
(1977) examined both satisfaction and complaining activity
related to clothing performance. They concluded from a
multivariate analysis that "knowledge of consumers'
activities, interests, and opinions, demographic character-
istics and textile product experience and knowledge can be
used to effectively 1) predict consumers' clothing per—

formance satisfaction, 2) discriminate between consumers

53

who are satisfied and consumers who are not satisfied with
clothing performance."29 Most marketers would likely

find reassurance, however, in their findings that the
first variable to enter the regression to predict sat-
isfaction was a factor score for performance problems

with clothing. Performance problems were less important
than personal characteristics and internal influences in
the determination of whether or not action would be taken
if an item were judged to be unsatisfactory.

One characteristic of the more recent explorations
of satisfaction related phenomena is more finely developed
theory. Recent studies are less likely to appear as fish-
ing expeditions or trial and error efforts where the
objective is the identification of useful satisfaction
correlates. A parallel may be seen in the early segmenta—
tion studies where the search for purchase correlates was
associated more with ease of variable measurement and in-
clusiveness of variables considered than with explanatory
theory. Despite this parallel there was no recognition in
the satisfaction literature of the overlap of the two areas
until Miller (1977) addressed satisfaction measurement in
explicit segmentation terms.

Miller attempted to extablish the existence of
segments based on demographic and psychographic character—
istics before testing for differences in satisfaction with
food stores where the subjects regularly shopped. This was

an alteration of the more common practice of first

54

classifying consumers as satisfied or dissatisfied. His
objective was to identify the "types of characteristics
typical of highly satisfied or dissatisfied consumers."30
This is consistent with many other satisfaction studies
and with purchase segmentation studies with allowance for
the concern with brand choice rather than satisfaction.

Miller's purpose was primarily methodological.

He did conclude, however, that his approach yielded
results similar to those of other studies.

It should be noted that not all research on the
subject of satisfaction was reviewed here. Specifically,
the substantial literature dealing with job satisfaction
and the more modest literature on life satisfaction were
excluded. Also omitted was work in psychology which
emphasized the importance of reinforcement. The rationale
for these exclusions was that they were too tangential to

the purpose of this study.

'Segmentation

 

Positing the prediction of satisfaction in segmenta-
tion terms may have a substantial effect in satisfaction
research. It was argued previously that much of the sat-
isfaction research fit established segmentation research
structures. If, however, the two areas overlap it seems
prudent to evaluate the state of the art in segmentation
studies in order to build upon it rather than simply
replicate it.

Weisenberger's (1977) study is useful in this regard

55

because of its thorough review of the bases and methods
of investigation that have been employed in segmentation
research. In attempting to identify generalized, non-
product-specific market segments, Weisenberger faced a
decision on the appropriate bases(s) to employ, and pre-
paration for this decision required an evaluation of all
possibilities. His review led to the conclusion that
demographic and socioeconomic bases Offer little promise,
partly because of contradictory or ambiguous evidence
generated in numerous studies and partly because they can-
not contribute much to an understanding of why an in-
dividual buys a product. The early satisfaction research
is much like the early segmentation research in focusing
first on demographic and socioeconomic variables.

Weisenberger concluded that the best opportunity
to identify generalized market segments lay in the use of
life style concepts as expressed by an individual's
activities, interests, and Opinions. Ths conclusion is
apparently shared by other researchers since most recent
segmentation research dealing with issues related to bases
have employed life style notions (Richards and Sturman,
1977; Goldberg, 1976). For certain kinds of segmentation
studies, demographic variables may prove to be more
valuable, however, than they are at present. The best
evidence was provided by Blattberg,et a1 (1978) in a
study of deal-prone segments. The distingushing

feature of his research was a model which incorporated

56

specific hypothesis relating the demographic predictor
variables and the criterion variable.
Wind (1978A) introduced a speCial market segmenta-

tion research section in the Journal of Marketing Research

 

by drawing four conclusions about current efforts. The
first was that the preponderance of attention is still
devoted to searching for bases of segmentation. In one
of the special section articles, Wind (1978B) used the
term "basis" as synonymous with dependent variable.
Consequently,there is some confusion about his intended
meaning. The fact that a given variable may be treated
as either a predictor or criterion variable depending
on the researchers purposes, does not help clarify matters.

While treating satisfaction as a criterion variable
has the appearance of being no different that using deal-
proneness, brand preference, or product usage, there is a
notable difference. Segmentation has assumed a position
of importance in marketing because of its presumed value
as an aid to managerial decision making. Homogeneous
clusters of customers are of value because of similarity
of elasticities, response function coefficients, or
marginal responses. With respect to satisfaction segments
(assuming they exist and can be identified), it is not
readily apparent what the marketing program implications
are.

Interest in overlapping satisfaction and segmenta-

tion research may be sustained by development Of its

57

utility to management, but it seems more likely that in
the immediate future macromarketing concerns and general
theory development will provide the motivation. Miller's
(1977) study provided much of the stimulus for this study,
but Westbrook and Newman (1978) helped establish its
relevance by demonstrating that the process of purchasing
may be a source of dissatisfaction and segmentation con-
cepts help to understand why.

The present research is premised on the hypothesis
that a consumer may judge a product or service as unsatis—
factory for reasons unrelated to the objective performance
criteria normally considered most important. The reasons
postulated are largely unexplored and involve consumers
purchasing in an imperfect market environment. Also, it
is not necessary to disavow rationality in purchase be-
havior to Observe that "perfection" in identifying and
selecting the best purchase Option is difficult to achieve.
Dissatisfaction with purchase processes, such as that
identified by Westbrook and Newman (1978), is consistent
with the possibility that an unknown part of dissatisfaction
with specific goods and services results from imperfections
in buying behavior and in the market environment.

If this scenario is plausible, one aspect of a
competitive market system that deserves attention for its
effect on customer satisfaction is the brand choice
decision. In the idealized model, informed consumer

choice among the numerous product configurations available

58

should lead to high levels of satisfaction. The applied
orientation of the marketing researcher, however, requires
that the idealized model be tested against reality. The
consumer choice commonly promoted as a major benefit of a
market economy is clearly one of its distinguishing
characteristics and begs evaluation for its effect on
satisfaction.

To this point satisfaction research, even satis-
faction segment research, has not been designed with this
issue in mind. A variety of issues exist for exploration.
Miller (1977) suggested that accounting for differences
in performance levels might be accomplished by holding
the brand constant while identifying satisfaction segments.
Again, the interest lies in performance levels rather than
the process and effects of brand choice. Miller's study
of satisfaction segments addressed satisfaction with the
foodstore usually patronzied as the criterion variable.

The subjects patronzied numerous foodstores and satisfac-
tion levels varied across foodstores, but the only recogni-
tion of possible interaction among consumer characteristics,
store characteristics, and patronage motives was a satis-
faction index involving a questionable multiplication of
importance weights and store descriptor ratings.

In a study which is closer to this research
methodologically, Maddox (1977) explored supermarket
satisfaction factor analytically. Maddox concentrated on

store attributes rather than customer characteristics as

59

predictors and found that, consistent with segmentation
theory, customers weighting certain factors highly tend
to patronize stores posseSsing those factors.

Store patronage is somewhat similar to brand pre-
ference, but no studies have yet been published which
address satisfaction as it relates to brand choice
rather than product classes and focus on the most promis-

ing customer characteristics, life style variables.

Summary

Each of the areas reviewed in this chapter is of
potential value to marketing managers in developing and
implementing successful marketing programs. They are
of value, also, in public policy formation and evaluation.
The principal limitation, however, is the frequent absence
of recognition of interdependence among these concerns.
This study explores one such area, the relationship be-

tween market segmentation and post-purchase evaluation.

FOOTNOTES: CHAPTER I I

1William J. McGuire, "The Current Status of
Cognitive Consistency Theories," Behavioral Science
Foundations of Consumer Behavior, ed. Joel B. Cohen (New
York: The Free Press, 1972), p. 253.

 

 

2M. T. O'Keefe, "The Anti-Smoking Commercials:
A Study of Television's Impact on Behavior," Public
Opinion Quarterly 35 (1971), pp. 242-248.

 

3H. B. Gerard, "Basic Features of Committment,"
Theories of Cognitive Consistency: A Sourcebook, ed.
R. P. Abelson et aI (Chicago: RandiMcNally,’l968),
pp. 456-463.

 

4H. J. Greenwald, "Dissonance and Relative vs.
Absolute Attractiveness of Decision Alternatives,"
Journal of Personality and Social Psychology 11 (1969),
pp. 328-333.

 

5J. W. Brehm and A. R. Cohen, "Re-evaluation Of
Choice Alternatives as a Function of Their Number and

Qualitative Similarity," Journal of Abnormal and Social
Psychology 58 (1959), pp. 373-378.

 

 

6C. A. Kiesler, 9Committment," Theories of COgnitive
Consistency, pp. 448-455; Gerard; J. W. Brehm and A. R.
'Cohen, Explorations in Cognitive Dissonance (New York:
Wiley, 1962), p. 300.

 

 

 

7Brehm and Cohen, "Re-evaluation of Choice."

8A. R. Cohen, H. I. Terry, and C. B. Jones,

"Attitudinal Effects of Choice in Exposure to Counter- Prop-
aganda," Journal of Abnormal and Social Psychology 58 (1959),
pp. 388-391; L. Festinger and J. M. Carlsmith, ITCOgnitive
Consequences of Forced Compliance,” Journal Of Abnormal and
Social Psychology 58 (1959), pp. 203-210; A. R. Cohen, J. W.
Brehm, and W. H. Fleming, "Attitude Change and Justification
for Compliance," Journal of Abnormal and Social Psychology
56 (1958), pp. 272-278; T. C. Brock, "Cognitive Restructur-
ing and Attitude Change," Journal of Abnormal and Social
Psychology 64 (1962): PP. 264-271.

 

 

 

 

 

60

61

9James F. Engel, David T. Kollat, and Roger D.
Blackwell, Consumer Behavior (2d. ed.; New York: Holt,
Rinehart, and Winston, Inc., 1973), p. 537.

 

loTibor Scitovsky, The Joyless Economy (New York:
Oxford University Press, 1976), p. 23.

 

11Ibid., p. viii.

lzIbid., p. 4.

l3Engel, Kollat, and Blackwell, p. 532.

14William Leiss, The Limits to Satisfaction: an
Essay on the Problems of Needs and Commodities (Toronto:
University ofiToronto Press, 1976), p. x.

 

 

lsIbid., p. 91.

16Ibid., p. x.

l7Ibid., p. 54.

18Ibia., p. 87.

19Ibid., p. 28.

201bid., p. 20.

211bid., p. 6.

221bid., p. 93.

23Philip Kotler, "What Consumerism Means for

Marketers," Harvard Business Review 50 (May - June, 1972),

 

p. 49.

24Arthor Best and Alan R. Andreasen, Talking Back
to Business: Voiced and Unvoiced Consumer Complaints
(Center for Study of Responsive Law, 1976), p. 100.

 

 

62

25Richard W. Olshavsky and John A. Miller,
"Consumer Expectations, Product Performance, and Per-
ceived Product Quality," Journal of Marketing Research
9 (February, 1972), pp. 19-21.

 

26Richard N. Cardozo, "An Experimental Study
of Customer Effort, Expectations,and Satisfaction,"
Journal of Marketing Research 2 (August, 1965), p. 249.

 

27H. Keith Hunt (ed.), Conceptualization and
Measurement of Consumer Satisfaction and Dissatisfaction
(Cambridge, Massachusetts: Marketing Science InsEitute,
1977).

 

 

28Ralph L. Day (ed.), Consumer Satisfaction, Dis-
satisfaction, and Complaining Behavior (Bloomington,
Indiana: Indiana University, 1977).

 

 

29Marjorie Wall, Lois E. Dickey, and W. Wayne
Talarzyk, "Predicting and Profiling Consumer Satisfaction
and Propensity to Complain," Consumer Satisfaction, p. 97.

 

30John A. Miller, "Data Reduction Techniques and
the Exploration of Satisfaction Segments," Consumer
Satisfaction, p. 102.

 

. CHAPTER III

RESEARCH METHODOLOGY

Introduction

 

This research was designed to address the question
of whether or not consumer satisfaction is related to
brand choice. As discussed previously, the expectation
created by the natural heterogeneity of demand and diversity
of choice in the market is that peOple are satisfied by
different things. If so, satisfaction may be related to
brand choice.

Alternative research designs are available to study
this question, and the design choice depends partly on the
researcher's motivation. Experimental and longitudinal
research designs are most desirable from the point of view
of illuminating causal relationships, but they have been
excluded here for lack of sufficient time and resources.
Also, research efficiency is typically improved by a
sequential approach which postpones causal analysis until
associational relationships are well established.

This study is more properly described as a cross-
sectional, ex post facto survey. While not so rudimentary
as to be described as exploratory, it is perhaps best
described as preliminary. It is preliminary in the sense

that longitudinal and experimental studies require data

63

64

such as is sought here in order to be designed effectively.

The research hypothesis at the heart of this re-
search is: there are no differences in the variables pre-
dicting (associated with) customer satisfaction between
purchasers of two brands within a given product class.

This hypothesis is best tested in the context of a
purchase decision likely to be characterized by extensive
problem solving. According to Howard, this occurs "when
consumers are confronted with a brand from a product class

1 In a situation such

they have never before encountered."
as this consumers encounter greater complexity and ex-
perience higher levels of uncertainty than in situations
where familiarity leads to habitual and limited problem
solving behavior. Extended problem solving is characterized
by the processing of great amounts of information and a
long decision time. These characteristics would be most
pronounced when they exist jointly with a purchase decision
involving substantial committments of time, expense, and
ego-involvement. This obviously would be a major decision
with important consequences. Any effects of brand choice
on satisfaction are more likely to surface in such an en-
vironment than in one involving frequently purchased, low
valued, mature products for which the differences among
brands are small.

The purchase decision which.meets the stated

criteria and was chosen for investigation in this study

was the selection of academic majors by university

65

undergraduate students. It is a decision characterized by
substantial financial outlays, ego-involvement, and long-
range consequences. It is also a decision for which there
is little opportunity to draw on prior purchase and con-
sumption experience.

The selection of a specific undergraduate population,
product class, and brands was made on the basis of access-
ibility for data collection purposes and size of the
various purchase groups. For reasons of convenience and
accessibility, the subjects chosen were undergraduate
students at Michigan State University. The product class
selected from the set of possibilities offered by this in-
stitution was academic majors within the College of
Business, which had the largest number of undergraduates
enrolled of any college within the University. Similarly,
within this product class attention was directed toward
two specific brands (academic majors): accounting (ACC)
and hotel and restaurant management (HRI). They were
selected because they were the two largest majors within
the College of Business. Consequently, they had the
potential to produce the largest number Of survey respon-
dents. This was an important consideration because the
anticipated use of multivariate statistical analysis
rendered the question of sample size indeterminate.

The principal limitation of selecting for investiga-
tion the choice of academic majors is inseparable from its

advantages. As a purchase situation offering a probability

66

Of positive research findings in excess of the typical
purchase situation, this situation is clearly not repre-
sentative of the broad range of purchase decisions. The
value of this research lies more in the attempted illum-
ination of relationships among variables than in its
generalizability. The common criticism of the use of
student subjects for study of purchase decisions of
general interest is not relevant here. The decision sit-
uation providing the context for this study clearly re-
quires a student subject group.

Subsequent sections of this chapter will provide
an overview of the research approach and discuss in some
detail the rationale for and development Of the research
instrument, the analytical procedures to be employed, and

data gathering procedures.

Research Approach

 

The approach taken in this study involves the
measurement of satisfaction with choice of major. This
process is reviewed in the following section titled
"Dependent Variable". It also involved measurements of
several different kinds which may be jointly described by
the phrase "personal characteristics Of respondents." The
issues involved and techniques employed are discussed in
the subsequent section titled "Independent Variables".

The section titled "Analytical Procedures" reviews the

statistical procedures used to search for the existence of

67

associations between dependent and independent variables
for the two groups of students and to test for the sig-
nificance of any differences that occur. The last section
Of this chapter reviews the procedures used to acquire

the data.

Research Instrument

Dependent Variable

 

The dependent variable in this study is identified
as satisfaction with choice of undergraduate major. The
central issue, however, is not satisfaction per se. Rather
it is the question of whether or not the predictors of
satisfaction vary between those students majoring in ACC
and those majoring in HRI.

A thorough analysis of satisfaction measures was
conducted by Aiello, Czepiel, and Rosenberg (1977). They
reviewed the diverse ways of operationalizing satisfaction,
the variations in scope of satisfaction interests from
product/service to system, the variation in conceptual
approaches, and the different scale formats available.

The practice employed in this research follows
that used by Czepiel, Rosenberg, and Akerele (1975).

They stated that: "To avoid confusion of the terms satis-
faction and dissatisfaction, the word satisfaction will
refer to the global concept. Its subsets would then

consist Of positive satisfaction and negative satisfaction

68

(dissatisfaction)."2 There has been a difference of
Opinion in the literature, however, as to whether satis-
faction and dissatisfaction are different points on a
single scale or are in fact parallel concepts. In re-
viewing the limited research on this subject Aiello,
Czepiel, and Rosenberg concluded that "the question of
whether the concepts of satisfaction and dissatisfaction
are different ends of the same scale or parallel scales,
or even a mixed scale remains unresolved."3
An objective of their research was to evaluate
alternative measures of the satisfaction construct and the
relationship between measures taken at the system, enter-
prise, and product/service levels. The scales used in-
cluded a simple satisfaction scale, an expectational scale,
a mixed scale, and an affect scale. With the exception of
the affect scale, these scales are purely historical. That
is, they involved a present judgment or evaluation of a
prior experience. This is a common practice and is con-
sistent with their observation that "to the extent that
there is any agreement among consumer satisfaction re-
searchers, there is agreement that satisfaction can be ex-
perienced only during or after consumption."4 It appears
that this conclusion follows from a priori agreement
among researchers rather than from fruitless attempts to
identify other forms of satisfaction. There is no evidence
in the literature, for example, of the possibility of

"anticipatory satisfaction." This hypothetical state

69

might exist when expectations, which are largely cognitive
in nature, are combined with evaluative judgments assoc-
ciated with satisfaction. The result would be satisfac-
tion with an upcoming experience or event as distinguished
from a present or past occurence.

For many products and services, the determination
of when purchase takes place or consumption is completed is
not a major difficulty. For the selection of academic major,
however, these determinations are more problematic. The
acquisition of the "product" takes place fragmentarily
over an extended period of time, sometimes several years,
depending on the requirements imposed by the university
and the inclinations of the student. In a sense, con-
sumption begins with the first course in the discipline.
What the student receives in the transaction with the
university is not immutable. The educational product
clearly is related to the effort and abilities of students
as well as those of the university. The duration Of con-
sumption is also variable. This is particularly true of
occupationally related or professional education as
Opposed to a traditional liberal education. For reasons
having to do with opportunity and personal preference, an
individual who has completed a major program may use the
acquired knowledge and skills over a lifetime, intermittently,
or not at all.

The indeterminate period of consumption is similar

to that of a durable good with a long expected life. In.

70

both cases questions are raised as to the appropriate time
or times for the measurement of satisfaction. Because of
changes taking place over time, multiple measures are
perhaps desirable. Also, it seems likely that the
measure(s) of satisfaction employed may need to be altered
because of the changes in experience and expectation that
occur over time. If so, the point during the consumption
process at which the measurement is to be taken should be
selected before the measurement device.

In this study, the decision was to measure satis-
faction among individuals with declared majors in ACC and
HRI who had attained senior status. This is clearly a
relatively early point in the consumption process, but
one that is not without its advantages. At this level
students have taken a substantial part of the coursework
required. As a result they have also had opportunities
to test their expectations regarding the substantive con-
tent of the coursework, the type of instruction, the
difficulty of the work, and so on. They have also Oppor-
tunities to become members of a group of people with
similar occupational interests and have access to the in-
formation flowing through the group's information network.
Information available in this network frequently involves
data, rumor, or speculation about consequences of major
selection that remain in the future such as changes in
job availability, salaries, working conditions, etc. It

can therefore play a role in creating or changing

71

expectations, and more importantly for this study, have

an influence on satisfaction with the selection of major
decision. Measurement of satisfaction among seniors should
therefore reflect the possibility of the joint influence

of experience and expectations.

Measurement Of satisfaction at a later point in
time, after occupational experience has been acquired and
the influence of expectations diminishes, has some merit
but also has disadvantages. This particular area has
been studied extensively under the title of job satisfac-
tion by organizational behavior researchers. The emphasis
in this research has quite logically been on the conditions
and requirements of employment. As the length of time in-
creases between the decision to pursue a major and the
measurement of satisfaction with that decision, the likeli-
hood increases that changes in the subjects will interfere
with accurate measurement.

A distinction must be drawn between two types of
evaluations in selecting and measuring the dependent
variable. The difference between satisfaction with
specific employment in a field and satisfaction with the
decision to prepare and train for employment in a selected
field must be noted. While the two evaluations are similar,
they are not identical.

In this study attention was focused on the latter
evaluation, the evaluation of the decision to pursue a

specific career. This led to a somewhat unique resolution

72

of the concern with specification and measurement of
satisfaction.

At the center of this issue is the fact that there
is a lack of consensus in the marketing literature as to
what satisfaction is and, as a result, how to Operation-
alize its measurement. As Aiello, Czepiel, and Rosenberg
pointed out, "consumer satisfaction has been defined in
many ways, most of which recognize satisfaction as a
measure of the 'gap' between the actual performance and
some standard."5 There is substantial disagreement,
however, about what to use as a standard and whether such
a standard can be obtained indirectly from measures of
dissatisfaction. The lack of accepted general measures
combined with the possibility of an expectational component
in this research led to the conclusion that a measure of
satisfaction designed specifically for major selection
was appropriate.

The construction of the satisfaction measure used
in this study began with small focus group interviews con-
ducted with convenience samples of seniors in ACC and HRI.
The group discussions focused on the identification of
specific goals, objectives, and expectations associated
with the selection of an academic major and upon features
of the students' experiences that led to evaluative,
satisfaction-type judgments. As a result of these inter-
views, twelve questionnaire items were constructed to tap

various aspects of the satisfaction construct. These

73

twelve items, which were included in the questionnaire
in the same AIO statement format as items subsequently
analyzed as predictors, are listed below:

1. Major courses are pretty much what I thought
they would be like.
2. If I were starting college all over again, I
would pick the same major.
3. I expect to get the kind of job that I want.
4. At this point, I would not consider changing
majors.
5. I have no regrets about picking the major I am
now enrolled in.
6. I think my selection of a major was a wise
decision.
7. There are lots of jobs for a person with a
degree in my field.
8. Based on my own experience, I would recommend
that other students pick the same major that
I did.
9. The salaries being offered to graduates in my
field look pretty good to me.
10. I am at least as interested in my field now as
when I picked my major.
11. If I were to go to graduate school, I would stay
in the same field.
12. Fundamentally, I am satisfied with my selection
of a major.

Respondents were asked to indicate the extent Of
their agreement or disagreement with each statement by
circling a number according to the following correspondence:

1. Very strong agreement

2. Strong agreement

3. Mild agreement

4. Neither agree nor disagree, don't know, not sure

5. Mild disagreement

6. Strong disagreement

7. Very strong disagreement
This seven point Lflmnt-type scale was used both for scale
items which were part of satisfaction measurement and for

AIO statements used as predictors. Likert scales typ-

ically are constructed with five or seven response

74

options although the number of categories may vary
widely. In this study the seven point scale was chosen
for consistency and ease of analysis. One group of items
used to develop predictors employed a semantic-differential
format which typically offers seven response options.
Also, the use of seven point scale leads to a broader
range of data points and facilitates the analysis des-
cribed in a subsequent section.
The twelve satisfaction items were then sub-
jected to cluster analysis via PACKAGE6 with the combined
data for all respondents. The analysis yielded one
cluster, including all twelve items, with an eigenvalue
greater than 1.0. A number Of items were ultimately
delected from the scale when it was examined against the
criteria of: (a) item content considerations, (b) the
pattern of intercorrelations with other items, and (c)
whether elimination of an item would result in a larger
coefficient alpha. The last two criteria were evaluated
with the sub-program RELIABILITY in SPSS.7
In successive steps the items eliminated were 1 and
11, 9, 8, 7, and 4. The resulting scale, which included
items 2, 3, 5, 6, 10, and 12, had a coefficient alpha of

.89987. This was the scale used to measure satisfaction

as the dependent variable in subsequent analysis.

Independent Variables

 

The satisfaction measurement employed as the

75

dependent variable in this study was tailored to the
specifics of the purchase situation. The selection of
independent or predictor variables, however, was influenced
by the findings of both segmentation research and psy-
chological research in general. In both cases there are
measures in general use although the nature of the pop-
ulation of students required some adaptation.

Marketing theory and business practice both in-
corporate the belief that personal characteristics in-
fluence purchase and, by extension, satisfaction. The
researcher's task in investigating these and related
issues is to incorporate appropriate personal character-
istics. The use of characteristics that transcend a
selected purchase situation, as has been the intention
in this study, increases the likelihood that findings are
not restricted and relevant only to that purchase situa-
tion. It may be, however, that the Opportunity to
establish stronger associations between the variables of
interest is traded off with this purpose in mind.

Three types of data were obtained from survey
respondents. The first section of the questionnaire, re-
produced in Appendix A, dealt with demographic and socio-
economic data. In most marketing studies, concern in this
area centers on age, race, family size and composition,
education, occupation, income, social class, and related
‘measures. The population for this study, however, is a

unique segment of the population at large, and the

76

measures in general use have limited applicability. The
student population is relatively similar in age, and the
response categories for age used, for example, by the
Census Bureau would allow little or no differentiation
of student respondents for this variable. Likewise, the
experience of a university education marks a major transi-
tion point in life at which point the conventional demo-
graphic and socio-economic measures that are premised
on greater stability lose their value. Where conventional
measures have been included, some instances of adaptation
of response categories were necessary (e.g., age). In
addition a number of other measures were included (e.g.,
grade point average and percentage of college expenses
earned) which involved respondent data associated
specifically with the university educational experience.
Because demographic and soci-economic data have
proved to be of limited value in the prediction of brand
selection decisions, two other types of data were sought,
and greater emphasis was attached to them in the analysis.
The first is data from a set of 26 pairs of self-image
semantic differential adjectives. This set, with some
slight differences, was used by Miller (1977) in his study
focusing on a search for satisfaction segments. Twenty-four
of the bipolar self-descriptor adjective pairs were drawn
originally from an analysis of impressions of personality
by Wishner and Asch (1969). Two which Miller added,

Easily Satisfied - Hard to Please and Satisfied with Life -

77

Dissatisfied with Life, were also included, and two others
which Miller added were excluded. In five of the ad-
jective pairs, one term.was changed to make the two terms
more clearly opposites. Respondents were asked to check
which of seven points between the two contrasting terms
most accurately described themselves.

The third category of data is based on thirty-
eight statements in the form of activity, interest, and
Opinion (AIO) statements which have become increasingly
common in marketing studies with the expansion of interest
in life-style or psychographic variables for segmentation
purposes. These statements were drawn from three sources.
Ten were selected from a set of forty-four statements
used by Miller (1977). They were judged to be of more
general value than the remainder which dealt in some re-
spect with food shOpping. On the questionnaire these are
item numbers: 60, 62, 63, 65, 67, 68, 70, 73, 75, and 78.
Another sixteen items were generated in the small group
discussions with ACC and HRI seniors referred to pre-
viously. These deal naturally with the subjects of
occupations, career Objectives, working conditions, work
attitudes, etc. On the questionnaire these are item
numbers: 61, 64, 66, 69, 70, 71, 74, 76, 77, 79-84, and 86.

The remaining group of twelve items, questionnaire
item numbers 48-59, constitute a dimension of values
scale developed to measure four basic value dimensions:

acceptance of authority, need-determined expression,

78

equalitarianism, and individualism. This scale was
developed by Withey (1965) as a shortened form of a scale
developed in the study of interpersonal relations research
by Bales and Couch (1969). This scale was attractive

as a means of incorporating these value measures into

the analysis in concise fashion. Generally speaking, it
would have been desirable to include in independent
variables a broad range of attitude measures from an in-
ventory such as that compiled by Robinson and Shaver (1973).
The problems which would result from lengthening the
questionnaire and complicating data collection made this
approach not feasible.

The response format for this group of thirty-
eight statements was the same as that for the satisfaction
variable used as the dependent variable. Respondents
were asked to circle a number from one to seven indicat-
ing the extent of their agreement with each statement.
Number one indicated very strong agreement and number
seven indicated very strong disagreement.

The emphasis in this study is on the possibility
of differences existing in the importance of respondent
characteristics for the prediction of satisfaction among
those selecting different brands (majors). It is important
that the variables employed as predictors have sufficient
predictive value to allow comparisions between the two
groups. Because the emphasis is on comparisons, it is not

necessary that the variables be selected in order to

79

maximize their predictive value. Indeed, if the intent
were to maximize the prediction of satisfaction, the range
of predictor variables employed would not be restricted to
consumer characteristics. The use of general measures
such as the self-image adjective pairs and Withey's
dimensions of values scales combined with items specific-
ally developed for this study provided assurance that an
acceptable level of prediction would be attained.

The absence of independent variables believed a
priori to be associated with satisfaction with major
selection and the volume of respondent data generated
created a need for data reduction and the generation of
more concise, parsimonious predictors. The procedure
employed was again similar to that used in developing the
satisfaction scale. The sixty-four questionnaire items
with seven-point scale responses (self-image adjectives
and A10 statements) were analyzed with the blind cluster
analysis procedure in PACKAGE. Seven clusters (eigenvalues
greater than 1.0) emerged from this analysis. Each of
these clusters was more sharply delineated by removing
selected items from the scale. As in development of the
satisfaction scale, the criteria for examining each
cluster were: (a) item content considerations, (b) the
pattern of intercorrelations with other items, and (c)
whether elimination of an item would result in a larger
coefficient alpha. Subprogram RELIABILITY in SPSS was

used in evaluating the last two criteria.

80

Items not fitting any cluster or removed from one
by this process were pooled and again subjected to cluster
analysis. This residual analysis of residual unused items
failed to reveal any additional clusters with eigenvalues
greater than 1.01. The eight clusters that resulted from
this analysis used forty-four of the original sixty-four
items. The basic analysis involving the prediction of
satisfaction employed these eight clusters as predictor
variables. A summary of information about the eight
variables is presented in Table 1, and more detailed
information about the composition of each variable is

presented in Appendix B

TABLE 1

PREDICTOR VARIABLES

 

 

 

VARIABLE' 'NUMBER OF ' COEFFICIENT
NUMBER COMMON THEME ITEMS ALPHA
1 Strength, high standards 8 .777
2 Socially active 8 . 7 31
3 Social desirability 9 .738
4 Platitudes 7 . 5 8 0
5 Need-determined 3 . 5 5 7

GNWESSMII
6 Acceptance of authority 3 . 7 0 2
Inportance Of noney 3 . 54 6
Individualism 3 . 4 8 3

 

 

 

 

The clusterhmjprocedure in PACKAGE used to develop
these predictor variables was a principal components

factor analysis employing oblique rotation. While this

81

makes it easier to insure that the resulting factors have
meaning, it increases the chances of multicollinearity
among the factors. Table 2 was constructed to assess

the correlations among clusters for respondents in ACC,

HRI, and for all respondents combined.

TABLE 2

CORRELATIONS BETWEEN CLUSTERS

 

 

 

 

 

CUEHER CLUSTER NUMBER
NUMBER. 1 2 3 4 5 6 7
.129 r
2 .039 r
.090 r
.175 -.242
3 .020 -.206
.105 -.218
.093 .083 .073
4 -.065 -.022 -.089
.030 .038 -.000
.004 .047 .147 .073
5 .061 -.007 -.047 .184
.017 .039 .075 .116
-.071 -.040 -.008 .025 .009
6 .062 .112 -.060 .008 -.155
-.015 .021 -.031 .017 -.063
-.l78 -.031 .050 -.011 -.204 .013
-.307 .034 .019 -.003 -.036 -.195
-.230 .001 .044 -.010 -.115 -.074
.185 .174 .017 .135 .154 .009 -.024
.264 .132 .008 .295 .219 .002 -.096
.201 .167 .028 .195 .202 .007 -.042

 

Visual examination of the correlations leads to the con-
clusion that the eight clusters used as the basic predictor

variables are relatively free of multicollinearity.

82

Analytical Procedures

This section deals first with the procedures
employed to evaluate the linkage between predictor and
criterion variables. The methods of determining whether
differences between groups are statistically significant
are then reviewed. This constitutes the core of the
analysis. A supplementary analytical procedure is then
described for the determination of whether respondents
perceive the two majors as dissimilar.

The initial phase of the analysis involved com-
putation of Pearson product-moment correlation coefficients
(r) for satisfaction and each of the eight predictor
variables. Subprogram REGRESSION in SPSS was used to
compute the correlation matrices for both ACC and HRI
students because of interest in the differences, if any,
among each of the eight pairs of coefficients.

The Fisher r to z transformation was used to
answer the question of whether the correlation coefficients
were significantly different.8 This transformation allows
the researcher to convert a sample r value to a Z value
on a one to one basis. Inferences may then be made in
terms of Z and restated in terms of correlation. The
hypothesis test of equal correlation then takes the form

of the ratio

 

83

where Z1 and 22 are transformed values of the correlation

coefficient for the ACC and HRI samples and

0(21-22) = l + l
N1-3 N2-3

A second and related approach to the same infer-

 

 

ential problem is offered by multiple regression procedures.
The confounding effects created when there is some covaria-
tion among the predictor variables are controlled by this
procedure. Although there is evidence of minimal multi-
collinearity, multiple regression techniques allow the
specification of the contribution of a given variable to
the prediction of the criterion variable when the other
variables are held constant. The two approaches would be
expected to yield similar results.

To verify this expectation, two regression
equations were computed using subprogram REGRESSION and
the stepwise procedure for inclusion of each variable.
One equation regressed satisfaction with choice of major
on the eight predictor variables for ACC students and the
other did the same for HRI students. In both instances,
the first concern is the determination of which, if any,
of the predictor variables have significant regression
weights. This question may be answered immediately by
reference to SPSS output. The next question is whether
significant differences exist between the regression weights

for ACC and HRI students. This test is feasible only when,

 

(n

(j

(J

I”)

84
for a given predictor, the regression weight is significant
in at least one of the two equations. The statistic which

facilitates this test is

 

 

where Bl and B2 are the unstandardized regression weights

and SEBl and SEB2 are their respective standard errors.

While this constitutes the basic test of the
central hypothesis, some additional analyses were con-
ducted. An additional regression equation was computed
for all respondents combined. The intent was to compare
this equation with those of ACC and HRI separately to see
if the intermingling of respondents selecting different
majors led to different conclusions about the relation-
ships between predictor and criterion variables.

Additional regressions were run for both groups
in order to determine if the demographic and socio-economic
data added significantly to the prediction of satisfaction
with choice of major. The emergence of one variable as
significant in the ACC group led to an additional analysis
of variance to explicate this finding better.

A final analysis was conducted to test the Operat-
ing assumption that the two brands (majors) at the center

of this research were perceived as dissimilar in the eyes

Of the subjects. The principal value of this analysis

85

lies in its use to minimize alternative interpretations of
the central research findings. The data needed for this
analysis were generated by questionnaire items 98-113.
Each respondent was asked to indicate his or her percep-
tion Of ACC and HRI as majors. Eight selected character-
istics or attributes of academic majors were expressed in
modified semantic differential form. A check mark in-
dicated the respondent's perception of the association of
each attribute with ACC and HRI, respectively. The
analysis of this data involves both independent and

paired sample t-tests of the difference between means.

Data Gathering Procedures

It was determined that accessing the population of
ACC and HRI seniors to ask their cooperation with the
study could be done fastest and at least expense through
scheduled classes. Accordingly, class lists were reviewed
in order to identify the classes which had high concentra-
tions of seniors in either of the two majors. Considera-
tion was given to the likelihood that some students might
be enrolled in more than one of the identified classes
and whether this might create problems. The conclusion
was reached that there appeared to be no feasible way of
avoiding this possibility. It was not judged to be an
issue, however, because the combined effect of instructions
to complete only one queStionnaire and the time required
would discourage subjects from completing more than one.

The professors responsible for each of the

86

identified classes were approached to explain the purpose
of the research, seek their cooperation, and specify the
arrangements for returning completed questionnaires. The
objective of the arrangements was to remove any barriers
of inconvenience which might raise the non-response rate.
Subjects were generally offered the Options of: (a)
returning them to a subsequent class at which time they
would be picked up, (b) leaving them with their professor,
or (c) depositing them in the researcher's departmental
mail box. On occasion a professor offered the use of
class time for completion of questionnaires, and there was
no need for special arrangements for returns.

Because the questionnaire distribution took place
near the end Of the fall term, 1977, some professors felt
that scheduled use of class time allowed insufficient flex-
ibility to distribute questionnaires. As a result of this
and because Of a desire to increase the size of the re-
spondent group, the same procedure for distribution and
collection of questionnaires was followed at the start
of the winter term, 1978.

For the fall term, this involved ten ACC classes,
257 questionnaires distributed, and 111 completed question-
naires returned. For the winter term, 130 questionnaires
were distributed through seven classes, and 69 were com-
pleted and returned. For the HRI students during the fall,
71 questionnaires were distributed in three classes, and

62 were returned. In the winter term 126 were distributed

87
in three classes and 64 were returned. Usable responses
totaled 306thh 180 of these being ACC majors and 126
being HRI majors.
The data gathering procedures ended with the
punching of the data on to standard 80-column IBM computer

cards.

FOOTNOTES: CHAPTER III

1John A. Howard, Consumer Behavior: Application

Of Theory (New York: McGraw-Hill, 1977), p. 9.

 

 

2John A. Czepiel, Larry J. Rosenberg, and Adebayo
Akerele, "Perspectives on Consumer Satisfaction," Pro-
ceedings of the American Marketing Association (Chicago,
Illinois: American Marketing Association, 1975),
pp. 119-123.

3Albert Aiello, Jr., John A. Czepiel, and Larry J.
Rosenberg, "Scaling the Heights of Consumer Satisfaction:
an Evaluation of Alternative Measures," Consumer Satisfac-
tion, Dissatisfaction and Complaining Behavior, ed. Ralph
L. Day (Bloomington, Indiana: Indiana University, 1977),
p. 44.

 

41bid.

 

51bid., p. 43.

 

6This computer program was authored by John E.
Hunter and Stanley H. Cohen and is available at the
Computer Institute for Social Science Research, Michigan
State University.

7Norman H. Nie et a1, Statistical Package for the
Social Sciences (2d. ed.; New York: McGraw-Hill, 1975), and
SPSS-6000 Suplement (East Lansing, Michigan: Michigan
State University, 1978).

 

 

 

8William L. Hays, Statistics for Psychologists
(New York: Holt, Rinehart, and Winston, 1963), p. 532.

 

88

CHAPTER IV

RESEARCH FINDINGS

In this chapter the results of the analytical pro-
cedures described in Chapter III are reviewed and discussed.
The bivariate relationships between satisfaction and the
eight predictor variables are reviewed first. Attention
is focused primarily on the differences between ACC and HRI
students with respect to the correlations between satis-
faction and each of the eight predictor variables. The dis-
cussion then advances to the multivariate procedures used.
The results are reviewed for the multiple regression
analysis including comparisons of regression weights for
the equations computed. The multiple regression analysis
is then expanded to include the demographic variables.
Finally, perceived attributes of the two majors are com-

pared for ACC and HRI students.

Bivariate Analysis

 

In the bivariate phase of the analysis, Pearson
product-moment correlation coefficients were computed be-
tween the eight predictor variables and satisfaction for
both ACC and HRI students. The principal value of the
correlation coefficient is that it shows the direction of
the relationship, or whether the variable is positively or

negatively associated with satisfaction.’ The coefficient

89

90

of determination (r2), however, is a more useful statistic
for prediction because it shows the percentage of the

total variance in satisfaction that is explained by a given
variable. It is, therefore, a direct indicator of the
strength of the association. The computations of r and r2
were made with subprogram.REGRESSION of SPSS and are sum-

marized in Table 3.

TABLE 3

BIVARIATE CORRELATIONS: SATISFACTION WITH PREDICTORS

 

 

 

 

, ACC Students HRI Students
Predictor
. 2 2
Variable r1 r1 n1 r2 r2 n2
1 .029 .000 177 -.052 .033 124
2 .281 .079 180 .305 .093 125
3 .374 .140 179 -.018 .000 125
4 -.025 .001 176 .021 .000 120
5 -.174 .030 179 -.019 .000 125
6 P.018 .000 179 .129 .017 126
7 .152 .023 180 .161 .026 125
8 -.133 .018 179 .307 .094 124

 

 

 

 

 

 

 

The values of the coefficients of determination for
ACC students ranged from .000 for predictor number 6
(acceptance of authority) to a high of .140 for predictor
number 3 (social desirability). For HRI students the values
ranged fn:n.000 for predictor number 3 (social desirability)
to .094 for predictor number 8 (individualism). Because
intercorrelations exist among the predictor variables,

the individual coefficients of determination cannot be

91

added to obtain an estimate of their joint predictive
value. Instead, it is necessary to use multiple regression
analysis. It is clear, however, from the low r2 values
that most of the variation in satisfaction is not explained
by this set of predictor variables.

An important qualification should be noted in inter-
preting the results of both bivariate and multivariate
analysis. Predictive efficacy was a secondary issue in the
research design. Had it been the major consideration, the
design would have incorporated other variables believed a
priori to be important determinants of customer satisfaction.
Specifically, the activities of sellers relating to trans-
actions with consumers would necessarily be included. This
would have included the ability of sellers to live up to
the claims of their Offerings and to deal with consumers
openly, honestly, and courteously.

The primary issue in this study was, however,
hypothesized differences in the influence of selected vari-
ables in the prediction of satisfaction between purchasers
of competing brands. Given this, correlation coefficients
are useful statistics. The signs of the correlation co-
efficients in Table 3 reveal that for five of the eight
variables the correlation with satisfaction was positive
for one group of students and negative for the other group.
These five variables were: number 1 (strength), number 3
(social desirability), number 4 (platitudes), number 6

(acceptance of authorityh and number 8 (individualism).

92

Nevertheless, the coefficients were small. Con-
sequently, it was desirable to test whether the differences
Obtained were merely due to chance. The statistical test
for determining the significance of such differences uses
the Fisher r to Z transformation. The test of significance

I
is based on the computation of Z as follows:

 

 

 

 

 

The one—to-one correspondence of correlation coefficients

I
to Z values and the computed Z values are shown in Table 4.

TABLE 4

TRANSFORMATIONS: CORRELATION COEFFICIENTS TO Z VALUES

 

 

 

 

Satisfaction
Predictor ACC Students HRI Students
Variable V ’
rl Zl r2 Z2 Z

l .029 .029 -.052 -.052 .69
2 .281 .289 .305 .315 .22a
3 .374 .393 -.018 -.018 3.48
4 -.025 -.025 .021 .021 .39
5 -.l74 -.l76 -.019 -.019 1.33
6 -.018 -.018 .129 .130 1.25
7 .152 .153 .161 .162 .08a
8 -.l33 -.134 .307 .317 3.82

 

 

 

 

 

 

asignificant at .001

93

The differences proved to be significant for only
two variables: number 3 (social desirability) and number 8
(individualism). The conclusion supported by this analysis
is that, other things being equal, the characteristic Of
social desirability contributed more to satisfaction among
ACC students while the characteristic Of individualism
contributed more to satisfaction among HRI students.

This may be viewed as offering evidence that sat-
isfaction is influenced by factors not relating to a
seller's actions, factors which may be outside the seller's
control. This suggests that the availability of choices
vamong brands contributes to consumer satisfaction if the
brands possess attributes well matched to differences in
relevant characteristics among consumers such as social

desirability or individualism.

Multivariate Analysis

 

Multiple regression can be used to attain two
different objectives. First, it Offers a means of evaluat-
ing the overall or combined contributions of independent
variables in predicting the dependent variable (satisfaction
with choice of major). Second, it Offers the opportunity
to evaluate the contribution of a given independent variable
when the effects of other independent variables are con-
trolled. For the purposes of this study, the second objec-
tive is more important although the first is not ignored.

The stepwise procedure employed enters variables

94

one by one into the regression equation with the order
determined by the contribution of each variable to ex-
plained variance in the criterion variable. The variable
with the highest contribution is entered first and the
remainder are entered in descending order. The procedure
also permits the deletion of variables that do not meet
pre-established criteria for inclusion.

Stepwise regression was run first for ACC students
with attention directed to the predictor variable re-
gression weights and whether they were statistically
significant. Regression weights proved to be significant
in the ACC group for four of the eight variables. They
were: number 3 (social desirability), number 2 (socially
active), number 7 (importance Of money), and number 8
(individualism). The E values or regression weights
ranged from -.31 to .47. The 95% confidence limits,
computed as BIZSEB where SEB is the standard error of B,
indicated statistical significance by the fact that the
upper and lower limits were either both positive or both
negative so that zero did not fall between them. Summary
statistics for the four significant variables are presented
in Table 5.

When a parallel regression analysis was performed
for the HRI students, only two variables had significant
regression weights. They were number 8 (individualism)
and number 2 (socially active). These two variables were

also significant in the ACC equation. Therefore, the

95

variables for which comparison of regression weights

could be made were limited to four variables previously
identified: number 3 (social desirability), number 2
(socially active), number 7 (importance of money), and
number 8 (individualism). Each of these four variables
produced a statistically significant regression weight,
that is, a weight different from zero, in one of the two
equations. This was a necessary condition for a test of
differences in weights between the two equations to be
meaningful. Table 6 presents summary statistics for these

four variables in the HRI equation.

TABLE 5

MULTIPLE REGRESSION RESULTS: ACC

 

 

 

Predictor Variable B SEB 95% Confidence Limits
3 .47 .10 .27, .69
2 .22 .07 .08, .36
7 .38 .15 .08, .67
8 -.31 .13 -.56, -.06

 

 

 

 

The confidence limits of variables 8 and 2 have
common signs. This indicates that the regression weights
are in fact not zero and that a relationship does exist
between the predictor variables and satisfaction. Likewise,
the contrasting signs on the confidence limits for vari-
ables 3 and 7 indicate that zero falls within the limits.

The regression weights are,therefore,not significant at

96

the a=.05 level.

TABLE 6

MULTIPLE REGRESSION RESULTS: HRI

 

 

 

Predictor Variable B SEB 95% Confidence Limits
8 .80 .22 .37, 1.22
2 .33 .11 .10, .56
3 -.ll .14 -.39, .17
7 .46 .24 -.01, .94

 

 

 

 

The central concern Of this research is whether
the regression weights for given variables are different
in the two equations. The test of the significance of
differences in regression weights utilizes the formula

B - B2

= 2
(SE ) + (SE )
\f B. .2

where B1 and B2 are the unstandardized regression weights

[—1

Z

NH

and SEBl and SEB2 are their respective standard errors.
This test was conducted for each variable which was

significant in at least one of the two equations, that is,

variables 3, 2, 7, and 8. Two variables were identified

for which the differences in regression weights were

significant. They were number 3 (social desirability)

and number 8 (individualism). The Z values were, respec-

tively, 3.37, significant at .005, and 4.36, significant

at .0001.

These same two variables were identified in the

97

bivariate analysis as having correlation coefficients which
differed significantly. The bivariate and multivariate
analyses thus yielded consistent results. Therefore, the
multiple regression results reinforced the conclusion that
the characteristic of social desirability contributed more
to satisfaction among ACC students while the character-
istic individualism contributed more to satisfaction

among HRI students. Satisfaction was clearly related to
brand choice. Other things being equal, an individual's
satisfaction with ACC was greater if that individual re-
flected the characteristic social desirability but was low
in individualism. At the same time, an individual high

in individualism could expect greater satisfaction with
the HRI major.

A slightly different perspective of the value of
considering brand choice in the prediction of satisfaction
may be gained by comparing not just equations for in-
dividual brands but also the equation for all respondents
combined. To facilitate the comparison an additional
regression analysis was performed on the total sample with
no distinction according to academic major. This yielded
three variables with significant regression weights. They
were: number 2 (socially active), number 7 (importance of
money), and number 3 (social desirability). Table 7
displays the summary results for these three variables.

These findings were then compared with those cited

previously for the ACC and HRI groups. Attention was

98

focused upon the differences among the lists of significant
variables and upon the coefficient of multiple determination
(R2). This statistic shows the percentage of variance in
the criterion variable explained by the combined effect of

the significant predictors. The significant variables and

R2 for each of the three regressions are shown in Table 8.

TABLE 7

MULTIPLE REGRESSION RESULTS: TOTAL SAMPLE

 

 

 

 

 

 

 

Predictor Variable B SEB 95% Confidence Limits
2 .27 .07 .15, .40
7 .35 .14 .09, .62
3 .22 .09 .05, .39
TABLE 8

SIGNIFICANT PREDICTORS AND R2 BY GROUP

 

 

 

Group Significant Predictors R2

(a = .05)
ACC 3,2,7,8 .236
HRI 8,2 .165
Total Sample 2,7,3 .121

 

 

 

The dissimilarities in the significance of pre-
dictors are more striking than the similarities. Variable
2 (socially active) was the only variable to appear in all
three equations. Its identification as a relevant issue

in satisfaction with academic major was,therefore,

99

independent of whether the analysis was focused at the
product class level (total sample) or at the brand choice
level (ACC or HRI). With respect to the other three
variables, analysis for the total sample either distorted
or concealed their importance for the two groups. Vari-
able 8 (individualism) was significant among both groups,
but because the signs of the weights were different, its
effects were concealed in the total sample. Likewise,
variable 3 (social desirability) appeared to be significant,
but the effect was confined totally to the ACC group. It
is clear that decomposing the total sample to examine the
brands separately yielded richer information about which
consumer characteristics were related to satisfaction.
The combined strength of these characteristics in
the prediction of satisfaction was also improved when the
two groups were considered independently. The percentage
of variance in satisfaction explained by the significant
variables (R2) was 23.6 for ACC and 16.5 for HRI but just
12.1 for the total sample. The lower R2 for the total
sample can be attributed to the offsetting effects of the
contrasting signs for variable 8 (individualism) and
variable 3 (social desirability) in the ACC and HRI
equations. Breaking down the analysis from the product
class level to that of individual brands produced two
related but distinguishable benefits. The first was more
accurate identification of variables associated with sat-

isfaction with the chOice of major. The second was the

100

improvement in the prediction Of satisfaction. The fact
of the improvement was more important than the size of the
improvement since predictive efficiency was a secondary
issue.

The essence of the analyses is that the belief
that consumer satisfaction is related to brand choice is
supported. The differences that exist among consumers are
important in the prediction of satisfaction. And studies
of satisfaction conducted at the level Of product class
tend to obscure the identifiable differences among brands.
This study was deliberately focused on consumer character-
istics rather than sellers' actions as sources of satis-
faction. This approach was adopted not because consumer
characteristics were preSumed to be the major causes of
satisfaction but because differences among consumers, ex-
pressed in their preferences, constitute a foundation of
market segmentation theory. Nonetheless, market segmen-
tation theory extended to its logical consequences in terms
of consumer satisfaction has not been subjected to em-
pirical tests.

Although the results discussed above constitute
the heart of the analysis, there are two additional aspects
to the study. First,the multiple regression analysis was
expanded to include demographic data. Several researchers
have published results of segmentation studies that support
the view that demographic data are less valuable than AIO

data in developing a better understanding of brand selection

101

decisions and consequences.

The only variables included in the stepwise re-
gression procedure were those for which the first order
correlation coefficient with satisfaction was significant
at a=.05. There were five variables in all. They are
questionnaire item numbers 5 (mfifl, 6 (sex), 12 (percentage
of college expenses earned), 15 (employment status), and
16 (hours employed per week). When these variables were
included in the analysis for ACC students, only the sex
variable had a statistically significant regression weight,
thereby improving the prediction of satisfaction. None of
the demographic variables had statistically significant re-
gression weights in the HRI equation. The difference be-
tween regression weights for sex in the two equations was

not significant at the a=.05 level.

Brand Perceptions

 

The second aspect dealt with perceptions of the
characteristics Of the two majors held by ACC and HRI
students. The analysis served two purposes. Most im-
portantly, it answered the question of whether the two
majors were perceived as dissimilar by the respondents.
It also provided an Opportunity to gain insight into some
of the processes at work in the formation of evaluative,
satisfaction judgments. The data used in this analysis
were generated by questionnaire items numbered 93 through

113. With these items respondents first reported their

102

perceptions of eight attributes or characteristics of
ACC as a major in modified semantic differential form.
They then did the same for HRI with the identical set of
attributes. The eight attributes were: (1) job Opportun-
ities, (2) starting salaries, (3) self-employment possi-
bilities, (4) size of employing firms, (5) quantitative
skills needed, (6) human relations skills needed, (7)
growth opportunities, (8) and difficulty Of program.

While the research hypothesis in this study is
that there are no differences in the role of variables
predicting satisfaction between competing brands, the ex-
pectation was that differences do exist. This expectation
was based on concepts of market segmentation theory and
research findings of segmentation studies dealing with
choice processes. Implicit in this theory is the pre-
condition that the brands be perceived by consumers as
being different in some key respects. If consumers do not
perceive differences among brands, the appearance of choice
is merely an illusion, and no differences would be ex-
pected in the role of predictor variables.

Consequently, the study assessed the comparative
perceptions held Of the two majors. Objective differences
between majors in ACC and HRI exist in the form of differ-
ences in courses, faculty, and associations with industries.
When making or evaluating choices of major, however, dif-
ferences that consumers perceive are more important than

objective differences. This is true whether or not the

103
perceptions are consistent with the objective differences.
The first question posed was: do ACC and HRI
students hold similar perceptions Of ACC as a major? The
question was answered by reference to the responses to
each of the eight attributes of ACC as a major. The mean
scores, T values, and significance levels for each of the

eight characteristics are summarized in Table 9.

TABLE 9

PERCEPTIONS OF ACC AS A MAJOR

 

 

 

 

 

Nbankaues
Characteristic ACC Students HRI Students T value

Assurance of a job 6.08 5.47 -3.94a
Level of starting salary 5.47 5.35 .94
Appropriateness fOr

self-enployrrent 2.91 4.24 -6.27al
LflqflihaXiofemqumEmt

‘with a large firnl 5.36 5.60 -l.43
Quantitative skills required 6.00 5.88 .92
Human relations skills required 5.67 3.28 13.225'
personal growth opportunities 6.03 4.15 10.77a
Difficulty of courses 6.34 5.99 2.73b

 

 

 

 

asignificant at .001

bshyfifharuzat.05

The mean scores differed significantly between the

two groups on five of the eight measures: assurance of a

104

job, apprOpriateness for self-employment, human relations
skills required, personal growth Opportunities, and dif-
ficulty of courses. ACC students perceived their major as
more difficult, offering better assurance of employment,
better preparation for self-employment, and more growth
Opportunities while requiring more human relations skills
than did HRI students. There were no differences in per-
ceptions of shading salary levels, quantitative skills re-
quired, and likelihood of employment with a large firm.
The perceptions of ACC held by the two groups were char-
acterized more by differences than similarities. The
common element of the differences was a more positive view
of ACC held by ACC students as compared to HRI students.

Respondents were then asked to indicate their per-
ceptions of the same eight characteristics of HRI as a major.
The responses are summarized in Table 10.

In this case the mean scores differed significantly
for seven of the eight characteristics. The exception was
human relations skills required, for which the two groups
were in agreement. HRI students saw their major as being
more difficult than ACC students did. They also saw it as
Offering better assurance of a job, better preparation for
self-employment, better opportunities for personal growth,
greater likelihood of employment with a large firm, as well
as requiring more quantitative skills than did ACC students.
They were more pessimistic than ACC students, however, about

the level of starting salaries.

105
TABLE 10

PERCEPTIONS OF HRI AS A MAJOR

 

 

 

 

Mean Scores
Cmnpcunfistk: NISStmknms HRIEkudams leflue
Assurance of a job 4.49 5.95 -9.11a
level of starting salary 3.51 2.84 4.21a
Appaxuiahamss fix: a
self—employment 3. 69 2 . 56 5. 34
IdkeLflrod<flfempknment a
‘with.a large firnl 4.34 5.23 -5.03
Quantitative skills required 3.87 4.77 -4.9za
Human relations skills required 6.51 6.60 - .90
Personal growth opportunities 5.20 6.05 -5.26‘31
Difficulty of courses 3.55 4.03 -2.98b

 

 

 

 

asignificant at .001

bsignificant at .005

As was the case with ACC, the students' views of HRI
were characterized more by differences than similarities.
In a pattern parallel to that of ACC, the common element of
the differences was a positive view of HRI among HRI students.
The sole exception to this general trend was the more pessi-
mistic view of starting salaries held by HRI students.

The similarity of the results for both ACC and HRI
majors suggests that students in both groups are subject to
a halo effect whereby they perceive their own major as be-

ing superior. A conclusion of this sort, however, would be

106

premature. A review of how ACC students perceive the two
majors in a direct comparison, attribute by attribute,

and a similar comparative review of HRI student perceptions
is required to reach such a conclusion. Such an analysis
was performed using paired Observations arranged casewise
within each group of students. Attention was focused on
differences in mean scores. Table 11 shows the results

for ACC students and Table 12 for HRI students.

TABLE 11

ACC STUDENTS:
PAIRED COMPARISONS OF ACC AND HRI AS MAJORS

 

 

 

 

 

Characteristic . Difference in Mean Scores T'Value
Assurance of a job 1.65 11.35"
level of starting salary 2.07 17.47a
AppnxuiaUxmss:fln: b

self-employment -. 70 -3 . 21
Lflqflihxxlofemphnment a
with a large firm 1.03 5.60
Quantitative skills required 2.17 14.803
Human relations skills required -.87 4.71“1
Personal growth Opportunities .93 5.97a
Difficulty of courses I 2.79 20.20‘31

 

asignificant at .001

bsignificant at .005

At first glance the results reported in Tables 11
and 12 appear to reinforce the earlier findings. Differences

appear to dominate as was the case previously. The

107

differences between means are significant in all instances
at the .05 level and in most cases at the .001 level.
Indeed, the condition that brands be perceived as being dis-

similar appears to have been met.

TABLE 12

HRI STUDENTS:
PAIRED COMPARISONS OF ACC AND HRI AS MAJORS

 

 

 

Chmrmterhfiic Difflnrmoe:hlybanthues 'rlkdue
Assurance of a jOb -.46 -3.01b
level of starting salary 2.51 14.47"”1
Appropriateness fOr

self-enployment l. 66 6. 69a
IdkeLflmodcflfempknment

'with a large firm .40 2.54c
Quantitative skills required 1.14 7.35a
HUman relations skills required -3.32 -l9.30a
Personal gmowth opportunities —l.96 -10.60a
Difficulty of courses 1.98 13.60a

 

 

 

asignificant at .001
bsignificant at .005

Csignificant at .05

The emphasis to this point has been placed on
differences, but it should be noted that some patterns of
similarity do exist. Both ACC and HRI students give the
ACC major higher ratings than HRI for the level of starting
salaries, likelihood Of employment with a large firm,

quantitative skills required, and difficulty of coursework.

108

Both groups also agree on a higher rating for HRI in terms
of human relations skills required. The three character-
istics which indicate disagreement between the two groups
are assurance of a job, appropriateness for self-employment,
and personal growth opportunities. In each case the
students reported a belief in the superiority of their own
majors. ACC students rated ACC higher, and HRI students
rated HRI higher.

If there was a tendency to perceive one's major in a
relatively positive light, it was weaker than appeared to be
the case earlier. It is possible that the tendency exists
and is strongest for such an attribute as growth opportun—
ities. They are experiential and personal. Information
about them is not easily available for dissemination to
students in other majors. Similarly, the tendency may be
weakest for those attributes which are more objective and
about which comparative information is readily available.

The scale item dealing with self-employment possi-
bilities is of particular interest. In the regression
analysis reported earlier, the most closely related pre-
dictor variable was number 8 (individualism). One of the
A10 items in this scale directly addressed the issue of self-
employment. The statement was: I would like to run my own
business. The individualism predictor was positively
associated with satisfaction with HRI and was negatively
associated with satisfaction With ACC. The result presents

something of a paradox. Students in HRI perceive relatively

109

.greater likelihood for self-employment with an HRI major.
Moreover, individualism, with which self-employment is
related, is a positive predictor of satisfaction. By way
of contrast, however, ACC students perceive the ACC major
as a better one for self-employment. Yet individualism is
negatively related to satisfaction.

There is a temptation to conclude that self-employ-
ment opportunities associated with HRI are determining
variables in the selection of the HRI major but are un-
important in the selection of ACC. Although it is consis-
tent with these results, it would be an inappropriate con-
clusion. The reason for this is that the study did not
deal directly with choice criteria. The emphasis in the
design of this study was placed on the prediction of satis-
faction. It was not placed on the prediction of choice.
The absence of data directly related to the choice process
at the time of major selection precludes this interpre-
tation.

A possibility that cannot be excluded because of
the nature of the cross-sectional design employed is that
the attitudes toward self-employment among the respondents
develOped subsequent to the selection of a major. The
interrelationships among variables important in the choice
process and those associated with satisfaction require long-
itudinal research designs to analyze them and draw meaning-
ful conclusions.

The link between predictor number 8 (individualism)

110

and the self-employment characteristic was interesting

and provocative in terms of "causal" relationships which
were not studied directly. But it is peripheral to the
fundamental purpose of this investigation which was to
determine whether satisfaction is related to or influenced
by brand choice.

The evidence generated in this empirical study
leads to the conclusion that, in a situation characterized
by perceived differences among brands, and one that con-
tains many of the attributes of extensive problem solving
behavior, satisfaction is indeed related to brand choice.
The mathematical association between satisfaction and the
personal characteristics employed as predictor variables
differed between purchasers of the two brands. Differences
were measured in terms of both strength and direction of
the association. The implication is that the differences
that exist among people may cause them to experience
more or less satisfaction with their purchase decisions

depending on the choices they make among competing brands.

CHAPTER V

SUMMARY AND CONCLUSIONS

Objectives

 

The central concern of this dissertation is an
empirical test of a fundamental element of market segmen-
tation theory - the belief that consumers benefit from
the ability to choose among marketplace Offerings. The
test does not take the form of a total cost-benefit analysis.
No attempt was made to assess the costs assoCiated with the
provision of choice alternatives or accrued as a result of
engaging in choice processes. Similarly, no attempt was
made to assess the gains from each of the several ways con-
sumers are presumed to benefit from the availability of
choice Options. For example, considerations of economic
structure dealing with the effect of substitutes on commodity
prices were not considered.

The study focused specifically upon consumers' eval-
uations of their satisfaction with a specific purchase de-
cision. Attention was then directed to the possibility that
a common set of predictor variables would differ in impor-
tance between purchasers of two brands within a given pro-
duct class. The presence of such differences would be
evidence of the social value of market segmentation. Var-

iation in the predictors of satisfaction would suggest that

111

112

the goal of consumer satisfaction is well served by an
abundance of differentiated market Offerings, particularly
if the attributes of the Offerings are well matched with
the differences in consumer motivations.

A necessary step in the resolution of uncertainty
in this area is the generation of empirical evidence as to
whether or not consumer satisfaction is independent of, or
unaffected by, brand choice. Such is the purpose of this

study.

Background

 

Recent years have witnessed the gradual evolution
Of marketing into a position of greater importance among
the functions of management. Simultaneously, consumer be-
havior has developed as an area of concern within marketing.
The shift away from an internal, production-oriented mode
of management toward an external, marketing-oriented
emphasis was consistent with the need to understand con-
sumer decision processes better.

Consumer satisfaction was not a new concept that
emerged with a new-found recognition of the importance of
consumers to the firm. The idea is of indeterminate age
and had been expressed for years by the statement that the
consumer was "king". Its impact upon the integration of
managerial activities, however, has been heightened in
the past quarter century by the frequent incorporation Of
the marketing concept into corporate policy statements.

Given the pragmatic tradition of marketing, it is

113

understandable that the focus of most research in consumer
behavior would be, and remains still, the determinants of
choice among market alternatives. Only within the past
decade has an interest developed in satisfaction as an
evaluative component of the post-purchase phase of the
consumer decision process.

It seems logical to expect research interest to
grow as evidence accumulates that increased knowledge would
be of value. There is evidence that the stakes marketers
have in understanding consumers' evaluative processes are
increasing. This evidence is directly related to the
growth in recent years of the influence of the consumer
movement. The movement can be traced at least to the turn
of the century, but its influence has been more cyclical
than steady. The most recent upsurge of interest in con-
sumer problems has also been the most wide-ranging, en-
compassing the general area of the rights and responsibil-
ities of buyers and sellers and such specific areas as
consumer safety, product liability, advertising puffery,
and others too numerous to mention. There can be little
doubt than the movement shares responsibility for the in-
creased interest in consumer satisfaction.

In recent years consumers have increasingly been
encouraged to be more aggressive in seeking redress for
market-related dissatisfaction. The advice has come from
formally organized consumer groups, self-appointed consumer

leaders, politicians, bureaucrats, the preSs, and increasingly

114

from business firms themselves. As consumer complaints

have escalated, governmental units and businesses alike have
become more sensitive to consumer problems, and the number
of consumer affairs departments has multiplied.

Legislative remedies are an increasingly common
response to growing volumes of complaints. Notwithstanding
efforts by the business community to resolve dissatisfaction
directly with its customers, incidents of dissatisfaction
are more often than in the past evolving into both public
and political issues. The stakes increase when this
happens, and the value of a better understanding of consumers'
evaluative processes also increases.

The small but growing body of published work in
marketing dealing with complaining behavior, satisfaction,
and related subjects is very eclectic in nature. One
approach that is apparent reflects a.shmnrrun, pragmatic,
managerial point-of-view. It is particularly evident in
many of the studies dealing with complaint handling. A
second approach addresses the problem at the aggregate
systems level far removed from the interests of most in-
dividual managers. These studies have a strong social wel-
fare flavor. A third approach, and one which is compatible
with the objectives of this study, attempts to answer rel—
evant theoretical questions without sacrificing the oppor-
tunity to produce results of managerial value.

The prevailing theory that is particularly relevant

for this study is market segmentation theory. Among the

115

premises of segmentation theory are heterogeneity of demand
and preferences. Different people buy different things.
Accordingly, it is considered wise strategy for marketers
to offer, where possible, a diversity of product service
combinations.

These premises have beengiven partial but in-
complete recognition in the studies of satisfaction in the
context of segmentation theory. As a result segmentation
theory has not contributed as much as is possible to
satisfaction research. At the same time it has not been
subjected to tests of its validity that can best be con-
ducted within satisfaction research.

A common approach employed in the studies in ques-
tion might be called the profile method. Researchers using
this method attempt to construct sets of descriptive char-
acteristics of complaining or noncomplaining and satisfied
or unsatisfied consumers. This is similar to work done with
purchasers of various products or brands and is based on the
assumption that groupings of similar customers exist. The
research task is to identify the characteristics associated
with the clusters. With the exception of a limited number
Of studies dealing with the role of expectations, most re-
cent satisfaction work has not employed a process approach
explicitly linking choice behavior with evaluative phenomena.

One consequence of research designs employed to
this point is the possibility of misinterpretation of satis-

faction segments. A second, more important point is that a

116

major premise of current marketing theory has gone untested.
That premise, which follows logically from heterogeneity of
demand, is that consumers not only buy different things but
are also satisfied by different things. To believe other-
wise is to suggest that the choices consumers exercise when
selecting among brands are little more than an illusion, that
random selection processes would yield equivalent satisfac-
tion. Because the premise is fundamental and contradictory
evidence would be disruptive, it is desirable to seek em-

pirical answers rather than depend upon assumption.

Methods

The purpose of the research design was to Operation-
alize the question of whether the prediction of satisfaction
differed among purchasers of two specific brands within a
product class. The product class selected for investiga-
tion was business school degree curricula at Michigan State
University. Two specific brands were chosen for direct
comparison: majors in accounting (ACC) and hotel and res-
taurant management (HRI). The research subjects were stu-
dents who had selected one of the two majors and had attained
senior status.

The source of the data for the analysis was a self-
administered questionnaire in the context of a cross sec-
tional ex post facto design. The questionnaire was distrib-
‘uted in classes which had substantial concentrations of
seniors in the two majors. Usable responses totalled 180

.from the 387 distributed to ACC students and 126 from the

117
197 distributed to HRI students.

The basic test of the central research hypothesis
required the computation of multiple regression equations
for both groups. Satisfaction with choice Of major was the
criterion variable and a common set of predictor variables
was used. Statistical tests were then conducted to deter-
mine which variables were statistically significant in the
two equations and whether the regression coefficients
differed significantly between the two equations for given
variables.

The dependent or criterion variable was a six item
scale which resulted from the application of cluster
analysis and scale reliability tests to the twelve ques-
tionnaire items dealing with satisfaction with selection of
major. The twelve original items were developed to reflect
aspects of the satisfaction construct which arose in small
group interviews with convenience samples of ACC and HRI
seniors.

The independent or predictor variables included dem-
Ographic data, self-image adjectives in semantic differen-
tial form, and AIO statements. The majority of the indi-
vidual items had been used by other researchers studying
satisfaction questions. Some were adapted from other studies
and some new items were developed from the small group in-
terviews.

Data reduction techniques similar to those used with

the criterion variable were employed with the predictor

118

variables. The self-image adjectives and A10 statements
which were hypothesized to have greatest predictive value
were reduced to eight predictor variable scales which in-
cluded 44 of the original 64 questionnaire items. These
scales were developed using the combined responses of ACC
and HRI students. It should be noted that the various items
used in the development of predictor scales were general
measures rather than being included because of a hypoth-
esized association with satisfaction with one or both of

the two majors.

Summary of Findings

 

Of the eight predictor variables employed in the
multiple regression analysis, four were noteworthy. All
four were statistically significant in the ACC equation,
and two of the four were significant in the HRI equation.

Of interest is the finding that the regression weights for
two variables Share the same sign in both equations, but
for the other two the signs are reversed. More importantly,
in two of the four comparisons the differences in regression
weights in the two equations were statistically significant.
In one of the two the signs were reversed. With regard to
the central research hypothesis, then, the findings is that
satisfaction is not independent of brand choice. There is
evidence in this case of differences in the variables pre-
dicting satisfaction between purchasers of competing brands
in a given product class.

An additional regression equation was developed

119
for ACC and HRI students combined. The purpose was to
compare with the other equations in order to determine
whether decomposition of product class purchasers into
brand categories in any way altered the perception of which
variables contributed to prediction of satisfaction. The
finding was that it very definitely did so. The most dra-
matic illustration involved a variable which was significant
in both ACC and HRI equations but did not appear in the
equation for the total sample. In another instance a vari-
able appearing in total sample equation did so because of a
predictive influence concentrated almost totally in just one
of the groups.

When the analysis was expanded to included demo-
graphic variables, the results were incidental as expected.
The last finding relates to the question of whether ACC and
HRI students see the two majors as different. For the eight
characteristics of the majors for which perceptions were
reported, T values for the differences in mean scores were
vartually all statistically significant. There is no ques-

tion that the majors are seen as dissimilar.

Implications For Marketing Theory and Practice

 

The results of this study are not revolutionary. Seg-
mentation theory creates certain expectations which were re-
alized by the results. The theory deals explicitly with the
association of consumer characteristics and the attributes
of differentiated products but only implicitly with the

evaluative judgments that follow purchase. This study

120

provides explicit evidence that the factors associated with
satisfaction with a purchase decision differs between brands.
As such it is supportive of prevailing marketing theory.

The implications for marketing practice are more
thought provoking. The notion of consumer satisfaction as
part of the marketing concept and good management in general
is hardly novel. This study tends to reinforce its value.
If a firm is astute in understanding its potential customers
and devising effective strategies for reaching well-chosen
target markets, it will undoubtedly develop some potential
business that is unsought or perhaps even unwanted. The
prospect of a sale may be higher than the likelihood the
buyer will be satisfied. The corollary of the statement
that different people are satisfied by different things is
that different people are dissatisfied with different things.
The focus here has been on satisfaction, but preventing
the development of dissatisfied customers may be equally as
difficult and more important in some respects.

Rather than accepting "fringe" customers and other
unplanned business unquestioningly, marketers may choose
to adjust their programs to attract desired customers while
at the same time repelling those who may generate immediate
sales but have a strong likelihood of dissatisfaction.

This implies that competing brands are dissimilar in their
need-satisfying potential and that the consequences of dis-
satisfaction are more serious than simply brand switching.

It will likewise be important for marketers to

121
determine the extent to which the criteria consumers use
in selecting products differ from the criteria used in
evaluative judgments. It is possible that buyer and
seller alike can benefit from the early provision of in-
formation known to be commonly used in post-purchase eval-
uation. As a very general statement, one might conclude
that the marketer who possesses a better understanding of
the determinants of satisfaction than exists among com-
petitors has an edge in designing marketing programs that

are effective and efficient.

Limitations and Future Research

 

This study is limited by the specifics of the pur-
chase situation studied and the cross-sectional design.
Consequently, it is far from a definitive study with respect
to questions of satisfaction vis-a-vis brand choice much
less the larger theoretical issues of segmentation and
choice theory.

Because the findings are tentative, a number of
possibilities for related research efforts are suggested.

A question that might arise upon examination of the results
of this study is whether the outcome will be different in
varied purchase situations. In the situation studied,
marked differences were perceived between the two choices
which were compared. The purchase also involved a sub-
stantial expense, time committment, and ego involvement

and may be viewed by some as irreversible. As these con-

ditions are relaxed, in particular the perception of

122
differences between brands, the relationship between
brand choice and satisfaction might be expected to weaken.
It may be argued that these conditions might indeed
diminish as a product market matures.

A significant question that remains for future
research turns on the strength of the predictive relation-
ship. Specifically, will the relationship be stronger if
predictor variables are not of general origin as was the
case here but are instead selected for testing as a re-
sult of a postulated association with satisfaction with the
Object or objects under investigation?

A research question involving analytical tech-
niques also remains to be answered. This study employed
R-factor analysis in construction of the predictor variables
because of a concern with the correlation between charac-
teristics. Alternatively, Q-factor analysis which deals
with the correlation between units (individuals) could be
employed to determine if similar individuals who make
different selections within a product class report diff-
erent levels of satisfaction.

Irrespective of the analytical technique considered
most appropriate, the analysis can be enriched in future
studies by expanding the number Of brands under study at
one time. This would be true for studies that vary widely
in terms of basic questions under study.

The most ambitious studies that remain to be con-

ducted are those incorporating longitudinal analysis. If

123
the nature of linkages between selection criteria and
evaluative processes are to be better understood, this is
essential. In this study, for example, the emphasis on
evaluation precluded dealing directly with choice processes.
As a consequence there was little basis for anything but
speculation about what they were. The brand choice de-
cision has been studied extensively but its relationship
to evaluative processes is largely unexplored. Ideally,
longitudinal studies would follow stable, representative
groups of consumers over a long enough period of time to
gather data extending across the processes from extensive
problem solving related to new ideas or products through
purchase behavior, consumption, and evaluation. A wide
variety of products and services would be studied relecting
differences in.gxfib classification, degree of competition,
and competitive strategies. The likelihood of such large-

scale projects is undoubtedly quite low.

Concluding‘Comments

 

The main stimulus for this study was a research
study which purported to identify satisfaction segments but
in the process appeared to disregard an implication Of seg-
mentation theory itself. Upon investigation, no research
could be found which directly addressed the point in ques-
tion, and it became the central hypothesis for this study.
The hypothesis was that the predictors of satisfaction vary
between brands within a product class. The specific sit-

uation chosen for investigation offered the appropriate

124
test conditions if the hypothesis was sound. The major
finding was that support for the hypothesis did exist in
the situation studied. It remains to be seen how general
or how strong the relationship is in other evaluative con-
texts. The major conclusion is that segmentation theory
and general marketing theory received an additional
accumulation of affirmative support in the findings of

this study.

APPENDIX A

APPENDIX A

RESEARCH INSTRUMENT

Your cooperation in completing the attached ques-
tionnaire is deeply appreciated. The validity of the study
is dependent upon completion of the questionnaire by a high
percentage of seniors such as yourself majoring in either
accounting or hotel and restaurant management. The focus
of this study is the evaluation of decision choice, in this
case the selection of an academic major, and its relation-
shop to personal characteristics.

While your assistance in this study is wholly vol-
untary, I would like to reemphasize its value and importance.
Also, each participant will remain anonymous.

James C. Cooper

Ph.D. Candidate
Michigan State University

5. Age: Under 21___, 21___, 22___, 23___, 24___, 25___,
26-30___ over 30___.
6. Sex: female___, male
7. Marital status: single___, married___, widowed___,
divorced___, separated___.

8. Expected term of graduation: F 77 , W 78 ,

Sp 78 , Sum 78 , after Sum 78 .

9. Major: accounting___, hotel and restaurant mgt.___
10. NO. of times major changed: 0___, l___, 2___, 3___,
4___, more than 4___
11. Have you earned a degree in another field? yes___ no___
12. Percentage of college expenses earned: 0<20___, 20<40___,

40<60 , 60<80 , 80-100

125

126
13. Current gradepoint average: under 2.0 , 2.0<2.5 ,
2.5<3.0 , 3.0<3.5 , 3.5-4.0

14. Do you have work experience in a job that is closely
related to your major field of study? yes , no

15. Are you presently holding a job? yes , no
16. If yes, how many hours per week do you work? less than
10 , 10<20 , 20<30 , 30 or more

17. What is your credit hour load this term? less than
7 , 7-10 , 11-13 , 14-16 , more than 16

18. Father's education in years: Elem.l-4 , Elem.5-7 ,
E1em.8 , High School 1-3 , High School 4 ,
College 1-3 , College 4 or more

19. Mother's education in years: Elem.l-4 , Elem.5-7 ,
Elem.8 , High Sohool, 1-3 , High School 4 ,
College I-3 , College 4 or more

20. Size of your high school graduation class: less than
50 , 50<100 , 100<300 , 300<500 , over 500

21. Present housing arrangements: MSU housing (dorm or
apt.) , fraternity, sorority, or co-Op , living
with parents , owned housing , off-campus rental

For each pair of contrasting adjectives check the point that
most accurately describes you.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

22. Stingy Generous

23. Inmnjve .mnjve

24. Easily sat- Hard to Please
isfied

25. Patnnm: Imgujent

26. Dominant submissive

27. Well Informed Uninformed

28. FboliSh Wfise

29. Sociable Unsociable

30. Satisfied with Dissatisfied

 

 

 

Idfe wiflalifie

31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.

47.

127

Defisiwa

 

Critical

 

thaufing

 

Uhhurdlhxmt

 

Omnioms

 

meflite

 

Self-centered

 

Infkadlfle

 

TaUatiwa

 

annical

 

 

 

Hmmue

 

Umhmt

 

Stdmxun

 

 

Agprssime

 

Sensitive

 

SUmfid

 

 

thmmest

 

 

 

 

Indecisive
Tbkaant
Yielding
Dnelkkmmt
nmnflsiwe
PoLHrz
(hherxenbard
Fhadifle
(huet
Lgnactkal
RuUfless
Just
(kmplfium
[magpresrwe
memsfijve
Ihfight

Emmet

Here are some statements which some people believe and others

don't.

48.

49.

1
2
3
4

\IONU'I

Very strong agreement
Strong agreement
Mild agreement

Indicate how strongly you agree or disagree with each
statement by circling the appropriate number.

Neither agree nor disagree, don't know,

not sure
Mild disagreement
Strong disagreement
Very strong disagreement

YOung people sometimes get rebelious ideas

butas'Umy'gnmvupijeycxphttr>getcweriiem.

YOUIEWE‘UDIESgXfilaUUKHity andxdmmgwmisoxa
rquxifingamuhorflqnjwmu:siUruion:unfltxmuth

main

1 2 13 4 ES 6 7

1. 2 13 4 5 £5 7

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

128

daedience and respect for authority are the
nost inportant things in character that
children should learn.

The solution to aluost any hunan problem
should be based on the situation at the
time, not on some general idea of right
or wrong.

Do what you want to do that's fun and
worry about the future later.

Since no vlaues last forever, the only
real values are those that fit the needs
of right now.

A group of people that are nearly equal will
work a lot better than one where people have

bosses and ranks over one another.

Everyone should have an equal chance and an
equal say in most things.

Everyone should have what he needs, the
inportant things we have belong to all Of
us.

We should all admire a man who starts out
bravely on his own.

In life a person should for the nost part
"go it alone", working on his own and
trying to make his own life.

One should not depend on other persons or
things, the center of life should be found
inside oneself.

I enjoy planning work carefully before
carrying it out.

The likelihood of a 9-5 job routine is
depressing.

The government has too much control over
business.

WhenIaminagroupthatisdiscussinga
problem, I seldom influence the solution
or action that is adopted.

I would like to run my own business.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

129

Students should have more power in university
affairs than they do now.

If you want something done right, you have to
do it yourself.

Businessmen in general are as honest as
other people

You can't change human nature.

I will accept the job that offers the
most money.

Business has too much influence on
government decisions.

I intend to work for a living rather than
live to work.

You can't get rich working on a salary.

Public welfare programs tend to make people
work less hard than they should

I would rather work with people than mostly
by myself.

You usually get what you pay for.

Having a job that is interesting is more
jlrportanttomethanhowmuchitpays.

In our society success is measured by how
much money you make.

You usually have to stand up for your rights
so people won't take advantage of you.

Job prospects were an important factor in my
selection of an academic major.

My family is supportive of my selection of a
major.

Making a lot of money is important to me.

The university is really not much help in
picking a major.

Before picking a major, I thoroughly checked
out a lot of other possibilities.

1

l

H

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

130

I've always known what kind of career I
wanted. 1

My major courses are pretty much what I
thought they would be like. 1

Nbst of my close friends at IVSU are not in
the college of business. 1

If I were starting college all over again,
I would pick the same major. l

IexpecttogetthekjndofjobthatI
want. 1

At this point, I would not consider
changing majors. l

I have no regrets about picking the major I
am now enrolled in. l

I think my selection of a major was a wise
decision. 1

There are lots of jobs for a person with a
degree in my field. 1

Based on my own experience , I would recommend
that other students pick the same major that
I did. 1

The salaries being offered to graduates in
my field look pretty good to me. 1

I am at least as interested in my field now
as when I picked my major. 1

If I were to go to graduate school, I would
stay in the same field. 1

Fundamentally, I am satisfied with my
selection of a major. 1

234567

Indicate your perception of accounting as a major by check-
ing the appropriate point on each scale.

98.

99.

 

 

Minimal assurance of a
job Lpon graduation

Starting salaries
are low

Substantial assur-
ance of a job Lpon
graduation

Starting salaries
are high

131

100. A good major if you

want to be self-em-
ployed __
Employment with a
large firm is un-
likely

101.

102. Ibquires minimal

quantitative skills _____

103. Ibquires minimal

human relations
skills

104. Offers limited per-

sonal growth
opportunities

105. Very easy course

of study

Indicate your perception of hotel

 

& restaurant mgt.

A poor major if you

want to be self-an-
__ Ployed

Employment with a

large firm is likely

Requires substantial
__ quantitative skills
Requires substantial
human relations
__ skills
Offers high personal
growth opportunities

Very difficult course
_ of study

as a major

 

by checking the appropriate point

106. Minimal assurance

Of a job upon

graduation
107. Starting salaries
are low

108. A good major if you

want to be self-
employed

Ehployment with a
large firm is un-
likely

109.

110. Requires minimal

quantitative skills___
lll. Requires minimal
human relations
skills

112. Offers limited per—

sonal growth
opportunities

113. Very easy course

of study

on each scale.

Substantial assurance

of a job upon
_ graduation

Starting salaries
__ are high

A poor major if you
want to be self-
__ employed

Employment with a
large firm is likely

Requires substantial
_ quantitative skills
Ibquires substantial
human relations
__ skills

Offers high personal
growth opportunities

Very difficult course
__ of study

APPENDIX B

APPENDIX B

TABLE 13

PREDICTOR VARIABLE COMPONENTS

 

 

 

Variable Questionnaire Descriptive Adjective
Number Item Number or AIO Statement
33 unbending
24 hard to please
32 critical
1 38 inflexible
25 impatient
43 stubborn
37 self-centered
26 dominant
29 sociable
39 talkative
30 satisfied with life
2 44 aggressive
23 active
31 decisive
74 I would rather work with
people than mostly by
myself.
63 When I am in a group that

is discussing a problem,
I seldom influence the
solution or action that
is adopted (negative).

 

46 bright
34 intelligent
42 just
41 humane
3 36 polite
28 wise
47 honest
40 practical

 

 

45 sensitive

 

132

133

TABLE 13 (cont'd.)

 

Variable
Number

Questionnaire
Item Number

Descriptive Adjective
or AIO Statement

 

57

56

66

68

55

65

78

We should all admire a
man who starts out
bravely on his own.

Everyone should have what
he needs, the important
things we have belong
to all of us.

If you want something
done right, you have to
do it yourself.

You can't change human
nature.

Everyone should have an
equal chance and an
equal say in most
things.

Students should have more
power in university
affairs than they do
now.

You usually have to
stand up for your rights
so people won't take ad-
vantage of you.

 

 

53

52

51

 

Since no values last for-
ever, the only real
values are those that
fit the needs of right
now.

Do what you want to do
that's fun and worry
about the future later.

The solution to almost
any human problem
should be based on the
situation at the time,
not on some general
idea of right or wrong.

 

134

TABLE 13 (cont'd.)

 

 

Variable Questionnaire 'Descriptive Adjective
Number Item Number or AIO Statement
50 Obedience and respect for

authority are the most
important things in
character that children
should learn.

49 You have to respect
authority and when you
6 stop respecting
authority, your sit-
uation isn't worth
much.

48 Young people sometimes
get rebellious ideas
but as they grow up
they ought to get

over them.

 

69 I will accept the job
that offers the most
money (negative).

76 Having a job that is

7 interesting is more
important to me than
how much it pays.

81 Making a lot of money is
important to me
(negative).

 

58 In life a person should
for the most part "go

it alone," working on
his own and trying to

8 make his own life.

59 One should not depend on
other persons or things,
the center of life should
be found inside oneself.
64 I would like to run my own
business.

 

 

 

BIBLIOGRAPHY

BI BLI OGRAPHY

Aaker, David A. and Day, George, S. (ed.) ‘COnsumerism:
‘Search for the Consumer Interest. 2nd ed.; New
York: The Free Press, 1974.

 

 

Aaker, David A. and Day, George S. "Corporate Responses to
Consumerism Pressures," Harvard Business Review.
40 (November-December, 1972), pp. 114-124.

 

Aiello, Albert Jr., Czepiel, John A., and Rosenberg, Larry
J. "Scaling the Heights of Consumer Satisfaction:
an Evaluation Of Alternative Measures,"“Consumer
Satisfaction, Dissatisfaction, and Complaining Be-
havior. ed. Ralph L. Day; Bloomington, Indiana:
Indiana University, 1977, pp. 43-50.

 

Alderson, Wroe. Dynamic Marketing Behavior. Homewood,
Illinois: Ricfidrd D. Irwin, I965.

 

. "The Analytical Framework for Marketing," Pro-
ceedings: Conference of Marketing Teachers from
Far Western States. ed. Delbert Duncan; Berkeley:
University of CaIifornia, 1958, pp. 15-28.

 

 

Anderson, Rolph E. "Consumer Dissatisfaction: the Effect
Of Disconfirmed Expectancy on Perceived Product
Performance," Journal of Marketing Research. 10
(February, 1973T77pp. 38-44.

 

Aronson, E. "Dissonance Theory: Progress and Problems,"
Theories of COgnitive Consistency: a Source book.
ed. R. P. Abelson et a1; Chicago: Rand McNally,
1968, pp. 5-27.

 

Bales, R. and Couch A. "The Value Profile: A Factor Ana-
lytic Study of Value Statements," Sociological

Inguigy. 39 (1969), PP. 3-17.

 

Barksdale, Hiram C. and Darden, William R. "Consumer
Attitudes toward Marketing and Consumerism,"
Journal of Marketing. 36 (October, 1972), pp. 28-35.

 

Bartlett, John. 'Bartlett's Familiar Quotations. ed. Emily
Morrison Beck; 14th ed.; Boston: Little, Brown and
Company, 1968.

 

135

136

Bell, Daniel. "Revolution of Rising Entitlements,"‘ Fortune.
91 (April, 1975), pp. 98-103.

Best, Arthur and Andreasen, Alan R. Talking Back to
Business: Voiced and Unvoiced Consumer Compiaints.
Center for Study of Responsive Law, 1976.

Birmingham, Robert L. "The Consumer as King: The Economics
of Precarious Sovereignty," Consumerism: Search for
the Consumer Interest. ed. David A. Aaker and
George S. Day; 2nd ed,; New York: The Free Press,
1974, pp. 181-192.

 

Blattberg, Robert; Buesing, Thomas; Peacock, Peter; and Sen,
Subrata. "Identifying the Deal Prone Segment,"
Journal of Marketing Research. 15 (August, 1978),
pp. 369-377.

Boschung, Milla. "Manufacturer's Response to Consumer Com-
plaints on Guaranteed Products," Journal of Consumer
Affairs. 10 (Summer, 1976), pp. 86-90.

Brehm, J.W. and Cohen, A.R. Explorations in Cognitive
Dissonance. New York: Wiley, 1962.

 

. "Re-evaluation of Choice Alternatives as a
Function of Their Number and Qualitative Similarity,"
Journal of Abnormal and Social Psychology. 58 (1959),
pp. 373-378.

Brock, T.C. "COgnitive Restructuring and Attitude Change,"
Journal of Abnormal and Social Psychology. 64
(1962): pp. 264-271.

 

Broffman, Morton H. "Is Consumerism Merely Another Marketing
Concept?" Business Topics. 19 (Winter, 1971),
pp. 15-21.

Cardozo, Richard N. "An Experimental Study of Consumer
Effort, Expectations, and Satisfaction," 'Journal
of Marketing Research. 2 (August, 1965), pp. 244-249.

 

 

Chapanis, N. and Chapanis A. "Cognitive Dissonance: Five
Years Later," Psychological Bulletin. 61 (1964),
pp. 1-22.

Cohen, A.R., Brehm, J.W., and Fleming, W.H. "Attitude Change
and Justification for Compliance," ‘JOurnal of
Abnormal'and Social PsychOIOgy. 56 71958),
pp. 272-278.

137

Cohen, A.R., Terry, H.I., and Jones, C.B. "Attitudinal
Effects of Choice in Exposure to Counter-Propaganda,"
Journal of Abnormal and Social Psychology. 58 (1959),
pp. 388-391

Collins, Barry E. "Four Components of the Rotter Internal-
External Scale: Belief in a Difficult World, a Just
World, a Predictable World, and a Politically Re-
sponsive World," Journal Of Personality and Social
Psychology. 29 (March, 1974), pp. 381-391.

 

Cravens, David W. and Hills, Gerald E. "Consumerism: a
Perspective for Business," Business Horizons. 13
(August, 1970» pp. 21-28.

 

Czepiel, John A., Rosenberg, Larry J., and Akerele, Adebayo.
"Perspectives on Consumer Satisfaction," Proceed-
ings of the American Marketing Association, 1975.
Chicago: American Mafketing Association, 1975,
pp. 119-123.

Day, Ralph L. (ed.) Consumer Satisfaction, Dissatisfaction,
and Complaining BehaviOr. Bloomington, Indiana:
Indiana University, 1977.

 

 

. "Toward a Process Model of Consumer Satisfaction,"
Conceptualization and Measurement of Consumer Satis-
faction and Dissatidfaction. ed. H. Keith Hunt;
Cambridge, Massachusetts: Marketing Science Institute,
1977, pp. 153-183.

 

 

Day, Ralph L. and Landon, E. Laird. "Collecting Comprehen-
sive Consumer Complaint Data by Survey Research,"
Advances in Consumer Research. ed. Beverlee B.
Anderson; 3 Tl976i, pp. 263-268.

 

Delozier, M. Wayne and Woodside, Arch G. MarketingManage-

ment. Columbus, Ohio: Charles E. Merrill Publishifig
Company, 1978.

 

Diamond, Steven L., Ward, Scott, and Faber, Ronald. "Con-
sumer Problems and Consumerism: Analysis of Calls to

a Consumer Hot Line," Journal of Marketing. 40
(January, 1976), pp. 58-62.

Engel, James F., Kollat, David T., and Blackwell, Roger D.
Consumer Behavior. 2nd ed.; New York: Holt,
Rinehart, and WidSton, Inc., 1973.

Engledow, Jack L. "Was Satisfaction a Pig in a Poke?" Busi-
ness Horizons. 20 (April, 1977), pp. 87-94.

 

138

Festinger, Leon and Carlsmith, J.M. "Cognitive Conse-
quences of Forced Compliance," Journal of Abnormal
and Social Psychology. 58 (1959), pp. 203-210.

 

 

Festinger, Leon. A Theory of CognitiVe Dissonance.
Evanston, Illinois: Row and Peterson, 1957.

 

Frank, Ronald E., Massy, William F., and Wind, Yoram.
Market Segmentation. Englewood Cliffs, New Jersey:
Prentice-Hall, Inc., 1972.

 

Gaedeke, Ralph M. and Etcheson, Warren W. (ed.) Consumer-
ism: Viewpoints from Business, Government and the
Public Interest. San Francisco: Canfield Press,
I972.

 

 

 

Gerard, H. B. "Basic Features of Committment,"‘ Theories
of COgnitive Consistency: A Source-book. ed. R.P.
Abelson et al; Chicago: Rand McNally, 1968, pp.
456-463.

 

Goldberg, M. E. "Identifying Relevant Psychographic Seg-
ments: How Specifying Product Function Can Help,"
Journal of Consumer Research. 3 (December , 1976),
pp. 163-169.

 

Granbois, Donald; Summers, John 0.; and Frazier, Gary L.
"Correlates of Consumer Expectation and Complaining
Behavior," Consumer Satisfaction, Dissatisfaction,
and Complaining Behavior. ed. Ralph L. Day;
Bloomington, Indiana: Indiana University, 1977,
pp. 18-25.

 

 

Grazin, Kent L. and Grikscheit, Gary M. "A Differential
Study Of Citizen Opinions Concerning Government Con-
trol of Business," Combined Annual Preceedings,
1973. ed. Thomas V. Greer; Chicago: American Market-
ing Association, 1974, pp. 476-481.

 

Greenwald, H. J. "Dissonance and Relative vs. Absolute
Attractiveness of Decision Alternatives," Journal
of Personality and Social ngchology. 11 (1969),
pp. 328-333.

Greyser, Stephen A. and Diamond, S.L. "Business is Adapting
to Consumerism," Harvard Business Review. 52
(September-October, 1974), pp. 38-40.

 

Grikscheit, Gary M. and Granzin, Kent L. "Who Are the Con-
sumerists?" 'Jgurnal‘of'BusineSS‘Research. 3
(January, 19753. pp. 1-12.

 

Hays, William L. ‘Statistics‘for‘Psychologists. New York:
Holt, Rinehart, add Wifiston, 1963.

 

139

Heilbroner, Robert L. The Economic Problem. 3rd ed.;

Englewood Cliffs, New Jersey: Prentice-Hall, Inc.,
1972.

 

Herrmann, Robert O. "Consumerism: Its Goals, Organizations,
and Future,"' Journal of Marketing. 34 (October,
1970). PP. 55-60.

 

Hicks, J.R. A Revision of Demand Theory. Oxford: Clarendon
Press, 1956.

 

Hollander, Stanley C. "Measuring the Cost and Value of
Marketing," Business Topics. Summer, 1961,
pp. 17-27.

 

Howard, John A. Consumer Behavior: Application of Theory.
New York: McGraw-Hill, 1977.

 

Howard, John A. and Sheth, Jagdish N. The Theory of Buyer
Behavior. New York: John Wiley & Sons, Inc., 1969.

 

Hunt, H. Keith. (ed.) Conceptualization and Measurement of
Consumer Satisfaction and Dissatisfaction. Cam-
Bridge, Massachusetts: Marketing Science Institute,

 

 

1977.
Hunt, H. Keith. "CS/D - Overview and Future Research Direc-
tions," Conceptualization and Measurement of Con-

 

sumer Satisfaction and Dissatisfaction. ed. H.
Keith Hunt; Cambridge, Massachusetts: Marketing
Science Institute, 1977, pp. 455-488.

 

Hunter, John E. and Cohen, Stanley H. "Package: A System of
Computer Routines for the Analysis of Correlational
Data," Education and Psychological Measurement.
29 (1969), pp. 697-700.

 

Hustad, Thomas P. and Pessemier, Edgar A. "Will the Real
Consumer Activist Please Stand Up: An Examination
of Consumers' Opinions about Marketing Practices,"
Journal of Marketing Research. 10 (August, 1973),
pp. 319-324.

 

Kendall, C.L. and Russ, Frederick. ”Warranty and Complaint
Policies: An Opportunity for Marketing Management,"
Journal of Marketing. 39 (April, 1975), pp. 36-43.

 

Kiesler, C.A. "Committment," Theories of Cognitive Con-
sistency: a Sourcebook. ed. R.P. Abelson et a1;
Chicago: Rand McNally, (1968), pp. 448-455.

 

 

Kotler, Philip. Marketing Management: Analysis, Planning
and Control. 3rd ed.; Englewood Cliffs, New Jersey:
Prentice-Hall, Inc., 1976.

 

 

140

. "What Consumerism Means for Marketers," ‘Harvard
Business Review. 50 (May-June 1972), pp. 48-57.

 

Kraft, Federick B. "Characteristics of Consumer Complain-
ers and Complaint and Repatronage Behavior," Con-
sumer Satisfaction, Dissatisfaction and Complaining
Behavior. ed. Ralph L. Day: Bloomington, Indiana:
Indiana University, 1977, pp. 79-84.

 

Lancaster, Kelvin. Consumer‘Demand. New York: Columbia
University Press, 1971.

 

Landon, E. Laird, Jr. and Emery, Donald R. "Causal Attri-
butions Of Consumer Dissatisfaction as a Predictor
of Consumer Complaint Action: a Preliminary Report,"
paper presented at the Association for Consumer Re-
search meetings, Chicago, November 9, 1974.

Landon, E. Laird, Jr. "A Model Of Consumer Complaint Be-
havior," Consumer Satisfaction, Dissatisfaction,
' and Complaining BehaviOr. ed. Ralph L. Day;
Bloomington, Indiana; University of Indiana, 1977,
pp. 31-35.

 

 

Leiss, William. The Limits to Satisfaction: an Essay on the
Problems of Needs and Commodities. Toronto:
University of Toronto Press, 1976.

 

 

Lewin, Kurt; Demb, Tamara; Festinger, Leon; and Sears,
Pauline 8. "Level of Aspiration," Personality and
the Behavior Disorders: a Handbook oiExperimental
and Clifiical Research. ed. J. McV. Hunt; V01. 1,
New York: The RonaldiPress, 1974.

 

 

 

Liefeld, J.P., Edgecombe, F.H.C., and Wolfe, Linda. "Demo-
graphic Characteristics of Canadian Consumer Com-
plainers," ‘Journal‘of'Consumer Affairs. 9 (Summer,
1975), PP. 73-80.

 

Maddox, R. Neil. "Consumer Satisfaction with Supermarkets:
A Factor Analytic Study," Consumer Satisfaction,
Dissatisfaction, and Complaininngehavior. ed.
Ralph L. Day: Bloomington, Indiana: Indiana
University, 1977, pp. 163-166.

 

 

Mansfield, Edwin.‘ Microeconomics. New York: W.W. Norton &
Co., 1970.

 

Margulis, S.T. and Songer, E. "Cognitive Dissonance: a
BibliOgraphy of Its First Decade,"' Psychological
Reports. 24 (1969): PP. 923-935.

 

141

Mason, Joseph B. and Himes, Samuel H., Jr. "An Exploratory
Behavioral and Social-Economic Profile of Consumer
Action about Dissatisfaction and Selected House-
hold Appliances," Journal of Consumer Affairs.
7 (Winter, 1973), pp. 121-127.

 

McGuire, William J. "The Current Status of Cognitive Con-
sistency Theories," Behavioral Science Foundations
Of Consumer Behavior. ed. Joel B. Cohen; New York:
The Free Press, I972, pp. 253-274.

 

 

SPSS-6000 Supplement. East Lansing, Michigan: Michigan
State University, 1978.

 

Miller, John A. "Data Reduction Techniques and the Explor-
ation of Satisfaction Segments," Consumer Satis-
faction, Dissatisfaction, and Complaininngehavior.
ed. Ralph L. Day; Bloomifigton, Indiana: Indiana
University, 1977, pp. 102-114.

 

 

. "Studying Satisfaction, Modifying Models, Elicit-
ing Expectations, Posing Problems, and Making Mean-
ingful Measurements," Conceptualization and Measure-
ment of Consumer Satisfaction and Dissatisfaction.
ed. H. Keith Hunt; Cambridge, Massachusets: Market-
ing Science Institute, 1977, pp. 72-91.

 

Monsen, R. Joseph. Business and the Changing Environment.
New York: McGraw-Hill, 1973.

 

.Murray, Barbara B. (ed.) Consumerism: the Eternal Triangle.
Pacific Palisades, California: Goodyear Publishing

 

Co., 1973.
Iiabers, Lawrence. "Veblen's Critique of the Orthodox Eco-
nomic Tradition," Thorstein Veblen: A Critical

 

Reappraisal. ed. Douglas F. Dowd; Ithaca, N.Y.:
Cornell Ufiiversity Press, 1958.

 

Iiie, Norman H., Hull, C. Hadlai, Jenkins, Jean G., Stein-
brenner, Karen, Bent, Dale H. Statistical Package
for the Social Sciences. 2nd ed.; New York: McGraw-
HilI) 1975.

 

 

()‘Keefe, M.T. "The Anti-Smoking Commercials: a Study of
Television's Impact on Behavior," Public Opinion
Quarterly. 35 (1971), pp. 242-248.

 

 

(blander, Folke. "Consumer Satisfaction a Skeptics View,"
Conceptualization and Measurement of Consumer Satis-
faction and Dissatisfaction. ed. H. Keith Hunt;
Cambridge, Massachusetts: Marketing Science
Institute, 1977, pp. 409-452.

142

Oliver, Richard L. "A Theoretical Reinterpretation of Expec-
tation and Disconfirmation Effects on Posterior Pro-
duct Evaluation: Experiences in the Field," 'Con-
sumer Satisfaction, Dissatisfaction andCOmplaining
Behavior. ed. Ralph L. Day; Bloomington, Indiana:
Univers1ty of Indiana, 1977, pp. 2-9.

 

Olshavsky, Richard W. and Miller, John A. "Consumer Expec-
tations, Product Performance, and Perceived Product
Quality," Journal of Marketing Research. 9
(February, 1972), pp. 19-21.

 

Pearson, Michael M. "A Note on Business Responses to Con-
sumer Letters of Praise and Complaint," Journal of
' Business Research. 4 (February, 1976), pp. 61-68.

 

 

Resnik, Allen; Gnauk, Brain; and Aldrich, Rodney. "Corpor-
ate Responsiveness to Consumer Complaints," Con-
sumer Satisfaction, Dissatisfaction and Complaining
BehaviOr. ed. Ralph L. Day; Bloomington, Indiana:
Indiana University, 1977, pp. 148-152.

 

Richards, Elizabeth A. and Sturman, Stephen S. "Life-style
Segmentation in Apparel Marketing," Journal of
Marketing. 41 (October, 1977), pp. 89-91.

 

 

Robinson, John P. and Shaver, Phillip R. Measures of Social
Psychological Attitudes. Rev. ed.; Ann Arbor,
Michigan: Institute for Social Research, 1973.

 

 

Sahlins, Marshall. Stone Age Economics. Chicago: Aldine,
1972.

 

Scitovsky, Tibor. The Joyless Economy. New York: Oxford
University Press, 1976.

 

Smith, Wendell R. "Product Differentation and Market Seg-
mentation as Alternative Marketing Strategies,"
Journal of Marketing. 21 (July, 1956), pp. 3-8.

 

Stampfl, Ronald W. "Structural Constraints, Consumerism,
and the Marketing Concept," ‘Business Topics.
Spring, 1978, pp. 5-16.

 

Steinmetz, Charles P. "The World Belongs to the Dissatis-
fied," The American Magazine. 85 (May, 1918),
pp. 38-40.

 

Swan, John E. "Consumer Satisfaction with a Retail Store
Related to the Fulfillment of Expectations on an
Initial Shopping Trip," Consumer Satisfaction,
'Dissatisfaction,'and COmpIaining Behavior. ed.
Ralph L. Day; Bloomington, Indiana: University of
Indiana, 1977, pp. 10-17.

 

143

Thomas, William R. and Shruptine, F. Kelly. "The Consumer
Complaint Process: Communication and Resolution,"
Business and Economic Review. 21 (June, 1975),
pp. 13-22.

 

Thorelli, Hans B. and Puri, Yash R. "On Complaining in
Norway and the Role of the Information Seekers,"

 

plaining Behavior. ed. Ralph L. Day; Bloomington,
Indiana: Indiana University, 1977, pp. 130-138.

 

veblen, Thorstein.‘ The Theory of the Leisure Class. New
ed.; New York: Modern Library, 1918.

 

Venkatesan, M. "Cognitive Consistency and Novelty Seeking,"
'Consumer Behavior; Theoretical Sources. ed. Scott
Ward and Thomas S. Rdbertson; Englewood Cliffs,
New Jersey: Prentice-Hall, Inc., 1973, pp. 354-384.

 

Wall, Marjorie; Dickey, Lois E.; and Talarzyk, W. Wayne.
"Predicting and Profiling Consumer Satisfaction and
Propensity to Complain," Consumer Satisfaction,
Dissatisfaction,'and'Compldining Behavior. ed.
RaIph L. Day; Bloomington, Indiana: Indiana Univer-
sity, 1977, pp. 91-101.

 

Warland, Rex. H., Herrmann, Robert O., and Willits, Jane.
"Dissatisfied Consumers: Who Gets Upset and Who
Takes Action," Journal of Consumer Affairs. 9
(Fall, 1975). PP. 148-163.

 

Warland, Rex. "A Typology of Consumer Complaints," Con-
sumer Satisfaction, Dissatisfaction, and Complaining
Behavior. ed. RaIph L. Day; Bloomington, Indiana:
Indiana University, 1977, pp. 144-146.

Webster's New world Dictionary. ed. David B. Guralink;
42nd Coll. Ed.; Cleveland: World Publishing Co., Inc.,
1974.

 

Weisenberger, Terry Mathew. "Generalized Market Segments:
a Study Using Selected Convenience Goods in Vigo
County, Indiana." Unpublished Ph.D. dissertation,
Michigan State University, 1977.

Westbrook, Robert A. and Newman, Joseph W. "An Analysis of
Shopper Dissatisfaction for Major Household
Appliances," ‘Journal 0 ‘Marketing Research. 15
(August, 1978), PP. 456-466.

 

Wind, Yoram. "Introduction to Special Section of Market
Segmentation Research," ‘JOurnal of Marketing Re-
search. 15 (August, 1978Y] pp. 315-316.

 

144

. "Issues and Advances in Segmentation Research,"
Journal of MarketingAResearch. 15 (August, 1978),
pp. 317-337.

 

 

Wishner, Julius. "Reanalysis of 'Impressions of Person-
ality'," Psychologic Review. (March, 1960),
pp. 96-112.

Withey, S. "The U.S. and the U.S.S.R.: A Report on the
Public's Persepective on United States-Russian
Relations in Late 1961," Components of Defense
Policy. ed. D. Bobrow; Chicago: Rand McNally, 1965,
pp. 164-174.

 

Zaichkowsky, Judy and Liefeld, John. "Personality Profiles
of Consumer Complaint Letter Writers,” Consumer
Satisfaction, Dissatisfaction, and Complaining
Behavior. ed. Ralph L. Day; Bloomington, Indiana:
Indiana University, 1977, pp. 124-129.

 

"‘WWW“