ACCOUNTS, MEDIA, AND CULTURE: AN INTERNATIONAL IMPRESSION
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
This dissertation is dedicated to my husband, Lee, with love and thanks.
This dissertation would not have been possible without the help, dedication and
inspiration of many wonderful people, both at the University of Florida and much further
afield. Lynda Lee Kaid, my chair, has taught me more about doing quality research than I
ever thought I would know. It has been a privilege to work with her on this and many
other projects during my time here, and I thank her for her encouragement and support.
My other committee members from the College of Journalism and
Communications, Juan Carlos Molleda, Marilyn Roberts, and Leonard Tipton, have each
contributed time, energy and great ideas to this dissertation. All are superb teachers and
scholars, and I thank them for all the information, exasperation, and camaraderie they
have shared with me during my time here at Florida.
A seminar taught by my outside member, Barry Schlenker, provided the idea for
this dissertation. His feedback and guidance have been essential to the completion of this
study, and for that I offer him my sincere thanks.
Of course, having worked on both my master's and Ph.D. degrees here at Florida, I
have had the pleasure of working with many of Florida's
fine faculty members at one
time or another. Two in particular, John Sutherland and Kathleen Kelly, deserve warm
thanks for being extremely helpful and understanding bosses.
Juan Carlos Molleda, Antoni Castells i Talens, and Julio Cesar Herrero all helped
limited budget. This dissertation truly would not have been possible without their
generous help, and I thank them with all my heart.
Because communication is an applied field, my work has benefited tremendously
from the years I spent in the publishing industry. For giving a medievalist a chance in a
very competitive field I owe three people in particular: Nick Blenkey and the late Ran
Slater at Simmons-Boardman Publications, and Melissa Snyder at USA Today. I hope the
completion of this dissertation serves to justify in part the faith they showed in me so
many years ago.
Many friends have cheered me on and cheered me up, when need be during this
mid-life career change, and for that I thank them all. Three of them, Camille Broadway,
Janet Brownstein and Sandra Friedrich, have been particular sources of encouragement
during my graduate work. I am confident they will all be relieved that we can talk about
something other than my dissertation from this time forward.
My in-laws, Patrick and Kay Ahem, provided many hours of babysitting during
their trips here to Florida, so that I could get some guilt-free work accomplished. They
even helped with translation. I appreciate all their help and support.
My parents, Michael and Maureen Connolly, have never stopped encouraging me
to pursue my dream of an academic career. They have shared their time, energy, and
resources without hesitation to help make that dream a reality. They are an inspiration to
me on every level and I appreciate them more than they will ever know.
My daughters, Abigail Catherine and Eleanor Rose, have been a source of great
joy, cheerleading and diversion throughout my graduate career. They have been patient -
instead of playing Cariboo and making minestrone like a proper parent. I thank them for
them for every hug, kiss, note and poem that helped keep my priorities straight
throughout this process.
Finally, I thank my husband, Lee, for his unwavering belief in me. His ability to
take any situation and turn it into an opportunity never ceases to amaze me. He has been,
quite simply, the most wonderful partner anyone could ever hope to have.
TABLE OF CONTENTS
ACKN OW LEDGM ENTS ................................................................................................. iv
LIST OF TABLES ............................................................................. ..................... .............. x
LIST OF FIGURES ............................ ................ .............................. .................................. xi
1 PURPOSE AND SIGNIGICANCE OF1 THIS STUDY ...............................................1
Purpose of the Study .......................... ............................................. ........... .............6
State ent of the Problemi ........................................ ...... ............. ....... ....................7
REVIEW OF LITERA TURE .......................................................................................8
M ass Communications and Impression M management ............................... ................20
Organizational Communications Studies ........................................, em............... ...... 20
M ass Communications Studies ............ ................... ......................................... 21
Corporate Identity................................................................................................. 23
Advertising and Impression M anagement....................................................................25
Problems with Advertising................................... .................................................25
Advocacy Advertising................. ............................................ .............. .............26
Advertising Versus Public Relations................................................................ ........28
Predicaments anrd Impression M management ................... .............. ....... ..................30
Corporations and Predicaments ..............................................................................31
The Triangle M odel of Responsibility ........................................ .......................31
Types of Excuses ..................................................... ....g............ ..... ....................... 34
Problems with Excuses.......................................................................................36
Culture and Im pression M anagement....................... .... ....................... ....................38
Cultural Dimensions......................................... .................................................38
Criticisms of Hofstede's Approach ........................... ............. ...... ..................41
1N I1 I FM C Aje
Hypothesis and Research Questions ........ ........... ....... ............ .... ......................... 47
3 M ETHODOLOGY .... ................. ...... .......... ..................................... ..........................50
Research Design................. .. .......... .................... c...... .......................... ..... ...... ..51
Stimulus M aterials................................. ..............................................................51
Crisis Communication Elements ................ .........................................................53
Implementation of the Triangle M odel.................. ...... .................... ................... 54
M measurement ............. ....................................................................................... 55
Sample l e.S...S ................... ...................... ............................................................... 57
Procedure .......................................................................................................... 58
Analysis s .......... ..................................................................................................58
4 FINDINGS.................... ............................ ............................................................60
Factor Analysis and Scale Creation ................... ......................... .................................. 61
Preliminary Analysis E .............................................. ............................................. .......65
Primary Analysis ..................................... .. ..................................................66
Effects of Culture .................................................................. .................................66
Effects of Excuse Type................................................................ .........................68
Effects of Delivery Format..................................................................................70
Credibility of Advertising in Different Cultures ........... ...................... ...........71
Honoring of Excuses ..........................................................................................72
5 DISCUSSION AND CONCLUSIONS ........... ................... ...... ....................... .........74
The Impact of Culture .........................................................................................75
The Impact of Excuse Type .......................... ................... ....... ... ..................77
The Lack of Impact of Delivery Format..............................................................79
Implications ...................................... ............................ ......................................79
Limitations and Future Research.......... ......................... ............................ ..............81
Future Research......... ...... .............. u......... ............... te Sc............................... .......... 85
Conclusion ..........s ...................... .............. c.. .................... ........ ........ .......................c i86
A MULTICHEM CORPORATE WEB SITE ...... ................................ ........................88
B EXPERITISE STIM ULUS M ATERIALS ............................................................90
C TRUSTWORTHINESS STIMULUS MATERIALS................... ............................ ... .92
EVENT REPORT STIMULUS MATERIALS ..........................................................94
CLARITY EXCUSE STIMULUS MATERIALS............. .................. ..... ................. 96
OBLIGATION EXCUSE STIMULUS MATERIALS ..................................... ....... 100
CONTROL EXCUSE STIMULUS MATERIALS ............... .. ............ ....................104
CONNOLLY-AHERN'S CORPORATE CREDIBILITY MEASUREMENT .......108
BERLO, LEMERT, AND MERTZ'S (1970) SOURCE CREDIBILITY SCALE ..110
MCCROSKEY'S (1966) SOURCE CREDIBILITY SCALE..................................111
GAZIANO AND MCGRATH'S (1986) NEWS CREDIBILITY SCALE ..............11
FERGUSON'S (2003) ATTITUDE TOWARD THE BRAND SCALE .................113
NEWELL AND GOLDSMITH'S (2001) CORPORATE CREDIBILITY SCALE 11
LIST O~F~ REEENE .....S......a....S............mS....S..S...*......S..t.St....S*.. .ttaSe.S...S...S.S.S.... ......... ..116
LIST OF TABLES
Hofstede's cultural values indices and ranks for Spain and the United States.........50
Control Group Means for All Scale Items ... ..... ................ .......................... ............60
Factor Analysis: Total Variance Explained ........... ............. ............................ ...... ...61
Rotated Component Matrix Scores for Corporate Credibility Scale Items..............62
Confirmatory Factor Analysis: Total Variance Explained.......................................63
Rotated Component Matrix Scores for Confirmatory Factor Analysis of Corporate
Credibility Scales .....
Factorial Analysis of Variance for Trustworthiness ............ .................... .... ...........67
Factorial Analysis of Variance for Expertise ............. ..................... .......... ......... .. .. 67
Corporate Credibility by Culture................... ........... .... ................. .... .. ........68
Types of Excuses and Corporate Credibility ...... ............. ............. ...... .....................70
Effect of Excuse and Format on Corporate Credibility..... ................... ........... ... ..7 1
Corporate Credibility of Advertising Condition Subjects by Culture ......................71
LIST OF FIGURES
The Triangle M odel of Responsibility... .. ....................i............... ........4m.............................33
Estimated Marginal Means of Trustworthiness by Excuse Condition.....................69
Estimated marginal means of control and experimental groups on trustworthiness
Estimated marginal means of control and experimental groups on expertise by
nationality .........,..,.. ...........,..,, ....
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
ACCOUNTS, MEDIA, AND CULTURE: AN INTERNATIONAL IMPRESSION
Lynda Lee Kaid
Although multinational corporations operate in multiple cultural environments,
they currently have no way of anticipating how audiences in different cultural settings
will receive, evaluate and respond to their corporate communications in the wake of a
crisis. Using impression management as a theoretical framework, specifically the triangle
model of responsibility, this dissertation employed a quasi-experimental design in two
countries, Spain and the United States, to study the impact of culture, types of excuses
and media format on corporate credibility.
Credibility was measured on a 16-item semantic differential scale, with two
constituent factors: trustworthiness and expertise. Data were analyzed using a 2 (country-
Spain or United States) x 3 (excuse-clarity, obligation, control) x 2 (format-ad or
newspaper story) factorial ANOVA.
The results of the study reveal that nationality is the most significant predictor of
incident. Spanish subjects found excuses offered by a fictitious corporation less credible
than American subjects.
Data further reveal that some excuses offered in response to an adverse incident
may diminish credibility more than others. Judgments of expertise were more damaged
by a prescription clarity excuse (in which the corporation had no way of knowing that the
explosion would occur because no scientific literature had predicted it) than by an
organizational control excuse (in which the explosion was beyond the organization's
control because of a faulty valve).
Finally, results of the study indicate that issuing an excuse in an advertisement or
within the context of a news story makes no difference in corporate credibility. This
finding implies that organizations wishing to have complete control of their message may
wish to choose advertising as a vehicle for delivering their advocacy message.
PURPOSE AND SIGNIGICANCE OF THIS STUDY
Perhaps no one has worked harder to build and maintain a particular kind of
identity, and make a particular kind of impression, than TV personality and business
tycoon Martha Stewart. In fact, Ms. Stewart became so synonymous with that identity-
the upscale, stylish, do-it-yourselfing iiberhousewife-that the former stockbroker and
divorcee from Connecticut was able to parlay it into a vast media empire. Her three
magazines and her Emmy-award winning television show all bear her name, a testimony
to the powerful association between the woman and the lifestyle she has come to
In cultivating her image, and that of her company, Martha Stewart Living
Omnimedia (MSLO), Ms. Stewart was not so very different from the Chief Executive
Officer any other U.S. corporation. According to Barton (1993)
U.S. corporations spend millions of dollars on a daily basis to project a positive
image with the consuming public. By placing corporate spokespeople on talk
shows, mailing press kits, participating in trade shows, and making charitable
donations, organizations hope to build and maintain an image that will enhance
their overall performance. (p. 128)
In fact, the only difference between Ms. Stewart and all but a few CEOs around the world
was the almost complete identification of her with both the corporation and the brands it
a l S I1
i n '~ _.. ItL. Lt rI r L~r t rY ru nj:~
U.S. attorney for the Southern District of New York indicted the celebrity on charges of
conspiracy, obstruction of justice and securities fraud emanating from a 2001 stock trade.
Ms. Stewart pleaded not guilty to all the charges. However, she did announce she was
stepping down as chairwoman and chief executive of MSLO (Hays, 2003).
Ms. Stewart did much more than simply vow to fight the charges against her,
however. On June
Ms. Stewart took out a full-page advertisement in USA Today, the
national newspaper, proclaiming her innocence and outlining her side of the story ("Open
2003). The ad,
"An Open Letter from Martha Stewart,
" was addressed to her
"friends and loyal supporters." Ms. Stewart did not apologize for her actions in the
advertisement. Instead, she sought to excuse and justify them. In the ad she reiterated her
contention that the sale of her ImClone stock on December 27
,2001, a day before the
company revealed that the Federal Drug Administration had failed to accept an
application for a new cancer drug, was based on price considerations, and not insider
information (Hays, 2003). She closed by inviting readers to visit her Web site,
www.marthatalks.com, to contact her and receive "current information on the case"
Why did Stewart turn to advertising at this critical time? Why not simply release a
statement to the press, or consent to an interview with a sympathetic, high profile
network reporter? While organizations and individuals use advertising for a wide variety
of reasons, the most widely used forms of advertising give the public information about
specific products and services (Belch & Belch, 1998). However, advertising can also be
used to create images for companies, to attract potential employees, to announce changes
in an organization's
structure or operations, to influence social debate and public policy,
and to offer accounts of corporate or individual actions.
decision to pay for an advertisement in the country's
newspaper illustrates a dramatic attempt to restore her image, which had been tarnished
by legal proceedings. In averring her innocence, Martha Stewart sought to reclaim a part
of her identity-that of an honest businesswoman--that was denied her by the allegations
of illegal activity. In so doing, Ms. Stewart was engaging in an extremely important and
strategic means of impression management.
All organizations, like all individuals, engage in impression management, which
Schlenker, Britt, and Pennington (1996) define as "the goal-directed activity of
controlling information about some person, object, idea, or event"
(p. 118). Impression
management has its roots in the work of Erving Goffinan (1959), a sociologist who
described a dramaturgical view of social interactions, in which all people are considered
and those with whom they are interacting-even the self-are considered
" Current impression management study deepens and expands the life-as-
theater analogy, suggesting that, just like actors, all people seek to influence how an
audience, either real or imagined, perceives their "personality traits, abilities, intentions,
behaviors, attitudes, values, physical characteristics, social characteristics, family,
friends, job, and possessions" (Schlenker, 1980, p. 6).
From an organizational point of view-and Martha Stewart can hardly be separated
from her eponymous corporation-the decision to use advertising as a public relations tool
is an important one from an impression management perspective. By taking out an
individual is assured that a message is transmitted exactly as desired. It is reasonable to
assume that precise message transmission is particularly important when an individual or
corporate entity is facing legal action for alleged wrongdoing.
However, such precision may come at a cost: credibility. Americans, in particular,
are savvy media users. They know there are no "scrubbing bubbles"
in their can of
bathroom cleanser. They know that their cookies are not made by "elves."
Americans enjoy advertising (De Mooij, 1998a), they may not take it too seriously.
Knowing this, however, and well aware of the benefits of public relations in building
positive images for organizations of all kinds, organizations and individuals continue to
pay for issue advertising, to ensure their complete control over a message they wish the
public to receive, often in the wake of some kind of corporate crisis.
Looking at organizational communications during crises, Kaufmann, Kesner, and
Hazen (1994) suggest, "A careful and thoughtful program of matching the
communications method to the crisis may be the answer to surviving the thorniest of
With this in mind, the Martha Stewart advertisement
opens a number of interesting lines of study. First, what type of impression does an
individual or corporation make when it chooses to account for activities in advertising?
Second, is it different from the impression made by the same information conveyed by
another media form, such as within the body of a newspaper article? Does the type of
excuse offered change the way individuals respond to the event?
Such questions become even more complex when one considers the global nature
of today's business environment. For example, transnational corporations, which have
the wake of any crisis: knowing what is the most appropriate communications tool for
important messages in multiple cultures. The problem is compounded by the fact that
many organizations manage their transnational communications efforts from home
offices that are often far removed-geographically and culturally-from the communities
that must be reached with a corporation's message (Molleda & Quinn, 2004).
Large multinational corporations are not the only organizations that must deal
effectively with publics from a variety of cultural backgrounds. International audiences
are increasingly important even to companies that largely operate within one country, in
the form of foreign shareholders. The value of U.S. assets held by foreign entities in 2001
was $9.4 trillion (Quinlan & McCaughrin, 2002). Clearly, few public corporations
operate in a truly monocultural environment at the beginning of the 21st century.
However, the implications of the multicultural business environment on corporate
communications are only beginning to be explored.
Ware and Linkugel (1973) underscore the importance of choosing a "culturally
acceptable stance from which to speak on [one's] own behalf' (p. 283) to a successful
self-defense. This implies that some defenses are more acceptable in some cultures than
in others. However, there is little information on how self-defenses are viewed in other
cultures. In Martha Stewart's
case, for example, how would an Indonesian shareholder of
MSO stock react to her defiant stand against the authority of the government? How
would a Costa Rican housewife view her account of the stock trade? Even assuming that
advertisement could be translated perfectly into another language, which many
scholars doubt (de Mooij, 1998a), the reaction of the individual receiving the message
whether Stewart was able to give off the desired impression in a cross-cultural context
would be related not just to the message sent, but also to the worldview of the person
reading the advertisement.
This suggests more avenues of inquiry. First, do cultural values affect the
impression given off by a corporation using advertising to make an excuse? Second,
which cultural variables are related to those impressions, either positively or negatively?
Third, are cultural variables related to the kinds of excuses that are considered acceptable
in the wake of an adverse incident?
Purpose of the Study
The purpose of this dissertation is to extend the body of work already completed in
the social psychology field of impression management in two areas where there has been
relatively little study thus far: mass communication, specifically corporate crisis
communication, and cross-cultural comparative study. Using a quasi-experimental
design, the researcher tested the effects of different types of excuses and different
delivery mechanisms for those excuses on the perceived credibility of a fictitious
Additionally, this dissertation begins to test a specific hypothesis related to culture
and communications, specifically that cultural values are related to the perceived
credibility of corporations in the wake of a crisis. This was accomplished by conducting
the experiment in two different cultural contexts: Spain and the United States.
Although cross-cultural communication is an emerging field in mass
communication (Molleda & Connolly-Ahern, 2002), most of the work done in the area
communication (Molleda & Connolly-Ahern, 2002), most of the work done in the area
This study employs an interdisciplinary approach, using an existing theoretical
framework from the field of social psychology impression management to test the
effects of different crisis communication strategies on audiences from different cultures.
In this way, the dissertation makes an original contribution to the body of knowledge in
the mass communication field. An extensive search of the literature in the areas of
advertising, public relations, organizational communications, and social psychology
identified no similar study.
Statement of the Problem
Although multinational corporations operate in multiple cultural environments,
they currently have no way of anticipating how audiences in different cultural settings
will receive, evaluate and respond to their corporate communications, especially in the
wake of a corporate crisis. This dissertation employs a quasi-experimental design to
determine which kinds of excuses are most credible with audiences from different
The sample consisted of undergraduate students from large, academically
competitive universities, one in the southeastern United States and two in central Spain.
(Every effort was made to match the samples.) The results from a 16-item semantic
differential scale were subject to factor analysis, and the variables that were derived from
that analysis were analyzed using a 2 (country-Spain or United States) x 3 (excuse-
clarity, obligation, control) x 2 (format-advertisement or newspaper article) factorial
REVIEW OF LITERATURE
Defending oneself verbally in the wake of an attack on one's nature or actions has
deep roots in Western society (Ware & Linkugel, 1973). Socrates, Martin Luther, Richard
M. Nixon, and John F. Kennedy are just a few of the historical figures who have made
speeches of self-defense in response to criticism of their actions or even the very nature
of their existence.
While apologia have traditionally been studied within the realm of
interpersonal and public communication, Hearit (1996) notes that in the era of big
business, it is often corporations that feel the need to "clear their name"
(p. 234). This
dissertation will serve to expand the ancient study of self-defense communications
through an empirical examination of the affects of a sub-category of such speeches-
corporate excuses-in different cultural and mediated contexts.
In rhetorical terms, the defense of one's reputation or actions is known as apologia
(Aristotle, trans. 1954; see also Benoit, 1995; Hearit, 1996, 1999; Ware & Linkugel,
1973). The goal of all apologiae is simple: to bring resolution to an existing conflict
between the speaker and society.
Many different typologies have been suggested for studying apologiae, from a
variety of academic disciplines, including speech communication, sociology and
psychology (Benoit, 1995; Benoit & Drew, 1997
Schlenker, 1980). Typologies vary
predicaments, to detailed, such as Scott and Lyman's
as in Schdnbk
(1968) 10-part typology of accounts
(1995) 14-item list of image restoration strategies, to exceedingly complex,
ich's (1990) taxonomy of almost 150 reactions of actors during account
phases. This complex list provides some indication of the wide range of ways in which
people and corporations seek to explain their behavior once their reputation is threatened.
In fact, Sch6nbach'
vast taxonomy serves to underscore directly the importance of
reputation management in modem,
Typologies create a starting point for empirical research by creating a framework
for future analysis (Coombs & Schmidt, 2000).
Working in the area of speech
communication, Ware and Linkugel (1973) used Abelson's
(1959) four basic strategies of
conflict resolution to create their typology of apologetic postures. From Abelson's
they identified the following basic apologetic factors: denial, in which the accused
completely disavows a charge or event; bolstering, in which the accused tries to link his
own identity to some positive event, object, information, or person; differentiation, in
which the accused tries to separate some event, object, information, or person from the
larger context of the discussion; and transcendence, in which the accused attempts to join
some event, information, object, or relationship to a larger context that has not previously
been part of the discussion. These strategies fall into two distinct types (Ware &
Linkugel, 1973). Denial and bolstering are reformative, because they do not attempt to
change the audience's
meaning of an incident; differentiation and transcendence are
transformative, because their goal is to change the audience's meaning of an incident.
Ware and Linkugel (1973) then combined the apologetic factors to classify four
and differentiation, with the goal of acquittal for the speaker.
combine denial and transcendence, with the goal of illustrating the moral rectitude of the
accused with regard to his accusers. Explanative speeches combine bolstering and
differentiation, with the goal of giving the audience information about the event that will
improve their understanding and make condemnation of the speaker less likely.
Jusitificative speeches combine bolstering and transcendence, with the goal of seeking
not only understanding, but outright approval, from the audience.
While apologiae can, and often do, employ elements of each of the conflict
resolution strategies, Ware and Linkugel (1973) suggest that most apologiae derive their
persuasive power from one primary posture. However, they offer no empirical evidence
to substantiate that claim. In fact, most of the speeches included in their study include
examples of at least two postures, and there are no corroborating data on audience effects
to confirm which of the postures was most effective in gaining audience acceptance of
the apologia. However, many rhetorical critics have employed Ware and Linkugel'
typology in their analyses, especially in the area of political discourse, such as
(1985) analysis of Nixon's
Vietnam speeches. From a corporate apologia
perspective, Benoit and Lindsey (1987) used Ware and Linkugel'
the restoration of Tylenol's
typology to examine
image in the wake of a notable poisoning scare.
Working in the field of sociology, Scott and Lyman (1968) defined an account as,
"a statement made by a social actor to explain unanticipated or untoward behavior-
whether that behavior is his own or that of others, and whether the proximate cause for
the statement arises frnm the actnr himself or frnm someone else" (n. 46. Thev identify
are a number of sub-categories. Excuses, which are defined as "socially approved
vocabularies for mitigating responsibility when conduct is questioned" (Scott & Lyman,
1968, p. 47), come in four types of appeals: appeals to accidents, appeals to defeasibility,
appeals to biological drives, and scapegoating. Of these, appeals to biological drives,
which normally refer to an excuse based on an uncontrollable appetite, such as addiction
or sexual deviance, is not likely to transfer from the interpersonal to the corporate realm.
It is interesting, however, that the authors link these biological drives excuses to culture,
suggesting that some excuses may be more acceptable in some cultures than in others. As
an example, they offer an ethnography of a Mexican woman who was angry not because
her husband had lipstick on his shirt, but because he lied about it. The woman explains:
And he had me almost believing it was red paint! It was not that I am jealous. I
realize a man can never be satisfied with just one woman, but I cannot stand being
made a fool of. (Lewis, as cited in Scott & Lyman, p. 50)
Scott and Lyman note that failure to invoke the culturally appropriate account may result
in the failure of the reparation attempt.
Justifications, according to Scott and Lyman (1968), are defined as,
approved vocabularies that neutralize an act or its consequences when one or both are
called into question" (p. 51). In justification, an actor attempts to convince the audience
of the positive aspects of an event or its consequences. The authors identify six different
types of justifications: denial of injury, denial of victim, condemnation of condemners,
appeal to loyalties,
and self-fulfillment. Again, the typology has limited use in
the corporate environment, since while corporations might use some of the justifications,
others clearly would not be plausible. For example,
seeks to justify an event
part of an attempt to experience everything possible in the world, such as illicit drugs.
Scott and Lyman's
typology, in fact, seems very rooted to the time and place where they
collected some of their data: interviews in the Haight-Ashbury district of San Francisco
(1995) theory of image restoration strategies is a 14-part typology rooted
in the field of critical analysis. The book analyzes both personal and corporate apologia,
indicating that organizations, like individuals,
"must attempt to restore their reputations
after alleged or suspected wrong-doing" (Benoit, 1995, p. 1). The author identifies five
main categories of image restoration strategies: denial, evading of responsibility,
reducing offensiveness of the event, corrective action and mortification. Sub-categories
of denial include simple denial and shifting of blame. Sub-categories of evading
responsibility include provocation, defeasibility, accident and good intentions. Sub-
categories of reducing offensiveness include bolstering, minimization, differentiation,
transcendence, attacking the accuser, and compensation.
list is extensive, but it is also somewhat confusing because the categories
are not mutually exclusive. For example, shifting blame seems to fit as easily with
"evading responsibility" as with "denial,
" despite Benoit's assertion that shifting blame
"can be considered a variant of denial, because the accused cannot have committed the
repugnant act if someone else actually did it" (p. 75). Despite this, image restoration
theory has supplied a theoretical framework for a number of corporate crisis case studies
(Benoit, 1995; Brinson & Benoit, 1996, 1999). Coombs and Schmidt (2000) used the
C.. fh fvlr T 7r nr lr- ,, nm rfl's4 tm annC+<- + In 1 A ar. n a* 4 A l Cnana nn nn n rn-n nr: nl ra" n+<^+<< n
different image restoration strategies, however their findings indicated no significant
differences among the strategies.
Working in the field of social psychology, Schlenker (1980) identified two general
types of responses to predicaments: accounts and apologies. Accounts are "explanations
of a predicament designed to minimize the apparent severity of the predicament" (p.
136). Accounts come in three basic forms: defenses of innocence, which absolve the
actor from responsibility for the event; excuses, which allow that a negative event took
place, but reduce the actor's
responsibility for the event; and justifications, which allow
the actor to take responsibility for a negative event, while minimizing the negative
repercussions of the event (Schlenker, 1980).
Schlenker (1980) defines an apology as a response to a predicament in which
"actors admit blameworthiness for an undesirable event, but concomitantly attempt to
obtain a pardon or reduce the negative repercussions from real or imagined audiences"
154). A complete apology includes five basic elements: an admission of guilt or remorse,
a recognition of the normally prescribed conduct in the situation, rejection of the conduct,
a vow to use the prescribed conduct, and some form or penance or offer of compensation.
Schlenker notes, however, that the categories overlap: in response to a predicament,
actors will often employ a combination of accounts and an apology in an attempt to
remedy an adverse situation. Schlenker's
typology is general, but comprehensive. As
such, it presents a useful framework for the study of the effects of accounts on various
While tvoologies vary, the literature on the subject consistently notes that in
help them mitigate the impact of the event on future relationships. On the broadest level,
this can be accomplished either by distancing oneself from the event partially or
completely, minimizing the importance of the event, or simply apologizing for the event.
For the purposes of empirical inquiry, the Schlenker (1980) typology provides a good
framework, because of its small number of broad, comprehensive categories. It is
possible to imagine a corporation invoking any one of Schlenker's
categories; it is
difficult to imagine a corporate defense that would not fit into one of Schlenker's
According to Coombs and Holladay (1996),
"What organizations say to their
various publics during a crisis should influence the extent of the reputational and
financial damage a crisis can inflict on the organizational image" (p. 292). Of course,
circumstances will to a large extent determine the appropriateness of any given self-
defense to the audience for which it is intended and its success in influencing public
opinion (Coombs & Holladay, 1996; Schlenker, 1980; Scott & Lyman, 1968; Ware &
When there is physical evidence of bad environmental practices in the
form of contaminated drinking water, for example, the public is not likely to accept a
's attempt to distance itself completely from the problem.
Corporations, like individuals, must issue self-defenses appropriate to the known
facts of the case in order to have them accepted by the public (Benson, 1988; Coombs &
Holladay, 1996; Martinelli & Briggs, 1988). In a corporate context, apologiae are
normally issued in response to some kind of crisis. A corporate crisis is defined as "an
rwnt that i'Tinrnc nvr ba th*h nnt vanfal fahr arinmnT an nrvnl2tciafnn intrn licreniit n d
p. 4). Crisis communications are the communications between an organization and its
relevant publics that are designed to minimize the damage to an organization's
before, during and after a negative event (Fearn-Banks, 1996). In most corporations,
crisis communications is regarded as a public relations function (Coombs, 1999; Fearn-
Banks, 1996; Lerbinger, 1997). However, with increased media scrutiny and the threat of
lawsuits in the wake of corporate crises, legal counselors have begun to play an
increasingly important role in crisis communications (Fitzpatrick, 2000; Fitzpatrick &
Rubin, 1995). Fitzpatrick and Rubin (1995) used content analysis to illustrate the
difference between public relations and legal messages in the wake of a crisis.
Many case studies have examined corporate defenses in response to crisis.
Dionisopoulos (1986) examined the nuclear power industry's
campaign, which was designed to calm audiences in the wake of the Three Mile Island
disaster. Dionisopoulos suggests a causal link between the advertising campaign and the
relative unimportance of nuclear power as an issue to voters in the subsequent elections.
However, there is no way for a rhetorical essay to substantiate such a claim empirically.
Hearit (1996) discussed the effectiveness of a particular form of apologia, the
"kategoria" or countercharge, in a business-to-business exchange, between General
Motors and Dateline NBC. Hearit suggests that the use of countercharges are an
important means of challenging inaccurate or unfair media coverage, since other media
outlets are likely to cover charges leveled against their competition. However, he warned
that because of the "heavy-handedness"
of the strategy, it is best used only when the
accuser is of the same or greater size as the accused that chooses to countercharge; the
use of kategoria by corporations against smaller accusers, such as individuals or non-
profit organizations, might be construed as bullying by the public.
Hearit (1999) looked at the impact of new media-and the culture of new media
users-on acceptance apologia, in a study of Intel's
online apologia in the wake of the
discovery that its Pentium 586 chip made errors on high-level calculations. The study
documents problems corporations have in cyberspace that are not present in traditional
media. In particular, the "culture" of the Internet can become difficult to navigate for
corporate executives who attempt to make apologia in a new media forum if they are not
in the language and customs of the online environment. In the case of
Intel, newsgroup users refused to believe an apology actually came from Intel'
because the return e-mail address on the posting was not his. For the newsgroup users,
the lack of personalization of the posting translated into a lack of authenticity and,
therefore, a lack of legitimacy and credibility for both the CEO and his corporation.
Ironically, the computer company ultimately turned to more a traditional media venue-
an advertisement in the Wall Street Journal-to make a successful apologia.
Researchers have begun to examine crisis communication using empirical
methodologies. Marcus and Goodman (1991) investigated changes in stock price in
response to different types of corporate communications (defensive and accommodative)
during and after different kinds of crises (accidents, scandals and product safety and
health incidents). The field experiment found differences in the shareholder response, as
measured by price of stocks, after defensive or accommodative statements by
nt -nnr n +^tj rlnnc.Q fwla anli i or r*r tMo in iornn in itrhi r-h an a rrsm~nt ttlrC
strategies, including "denials of intention, volition, and agency" (p. 291), served
shareholders better after accidents. The authors found no significant difference in
shareholder response (as measured by stock prices) between the two strategies after
product safety or health incidents.
Coombs (1999) tested the effect of different components of crisis management
strategies on corporate reputation among a group of crisis managers. The results indicated
a positive result of compassionate methods, as opposed to informative methods, on
corporate reputation. Unfortunately, the subjects used, namely veteran crisis management
professionals, drastically reduce the generalizeability of the study. The 114 managers
who participated in the study had worked in the crisis management area for an average of
about 10 years; the reactions of such a highly trained group of professionals in response
to a crisis simulation can hardly be generalizeable to the reactions of the general
population. They do serve to gauge the importance placed on compassionate
communication by the professionals who are likely to respond to a corporate crisis
situation, however, and underscore the likelihood of them employing such a strategy in
the wake of a crisis.
Coombs and Holladay (1996) examined the fit between crisis type and crisis
communication strategy, based on Benson's
(1988) symbolic approach. The symbolic
approach posits that a corporation experiencing a one-time accident should attempt to
distance itself from the accident (i.e. make an excuse), while a corporation experiencing
repeated accidents or corporate transgressions should attempt to rebuild its image through
remedial measures, such as victims being offered compensation for their injuries. Using
corporate image of a fictional corporation as the dependent variable, the results indicate
that matching response type to crisis type improves reputational image among audiences.
Siomkos and Shrivastava (1993) used an experimental design to test the effect on
reputation of two fictitious corporations of four types of crisis responses: denial,
involuntary product recall, voluntary product recall and "super-effort," in which a
corporation goes well above the legal requirements of a product recall to protect
customers. The study found that existing company reputation influenced post-crisis
attitudes, with high reputation companies viewed more positively after a crisis than low
reputation companies. The results of the study indicate that subjects viewed crisis
responses negatively (denial and involuntary product recall) or positively (voluntarily
product recall and super-effort). However, there was only a minimal difference within the
positive and negative categories: an involuntary product recall was almost as negative as
a denial, while a voluntary product recall was almost as positive as a super-effort. The
study underscores the importance of honestly assessing-and improving-corporate
reputation in advance of a crisis, since low reputation companies need to use a very costly
super-effort response in order to improve consumer attitudes toward the company after a
crisis, while high reputation companies can use the less costly voluntary product recall
Arpan and Pompper (2003) examined the effect of a particular crisis
communication strategy-"stealing thunder"--on a specific group of individuals, namely
reporters and journalists. Stealing thunder is a proactive form of crisis communication, in
which a corporation makes "an admission of a weakness (usually a mistake or a failure)
media" (Arpan & Pompper, 2003, p. 294). The researchers created a fictitious corporation
and asked professional journalists and journalism students to respond to different
corporate crisis scenarios in which public relations professionals communicated
differently with journalists. Their findings indicated that while stealing thunder showed
promise as a tool for building credibility for public relations professionals among an
important target audience-the media-by increasing journalist'
perceptions of public
relations practitioners as credible sources of information, it also had the effect of imbuing
the story with greater newsworthiness, thus assuring prominent coverage. This study is of
particular interest because it looked at the impact of crisis communication strategies on a
particular group, or culture. However, like the Coombs (1999) study, the subjects of the
experiment limit the generalizeability of the results. That is not a fault of the research,
since such generalizeability was not the authors' intention. The researchers were
interested in the relationship between public relations practitioners and journalists, not in
the reactions of the public to the media produced by those journalists. The fact remains,
however, that the empirical literature on public reaction to corporate apologia remains
Although there is a growing body of literature on corporate apologia, mainly in the
form of case studies of corporate crises, only a small portion of that research has been
empirical in nature. The few empirical studies have confirmed some important aspects of
crisis management theories, including the importance of appropriate self-defenses in
rebuilding corporate image in the wake of a crisis. However, while theoreticians and
researchers have nosited a link between culture and acceptance of excuses (Hearit. 1999:
area. This dissertation begins to fill a significant gap in the literature by examining the
impact of culture on acceptance of corporate excuses.
Mass Communications and Impression Management
While initial work in the area of impression management, like that of apologia,
came from the area of interpersonal relations, it has proven an effective framework for
the study of organizational communications as well (Giacalone & Rosenfeld, 1989; see
also Arndt & Bigelow, 2000; Bromley, 1993; Hooghiemstra, 2000). Studies in the area of
impression management and organizational communications are varied, and have
described both external and internal corporate communications,
as well as target reaction
to such communications.
Organizational Communications Studies
Researchers have employed a broad range of methodologies in the field of
organizational communication and impression management, including case studies,
content analyses, surveys, and experiments. Most of these have focused on self-
presentation within the context of an organizational setting. Greenberg (1990) conceived
of organizational justice as a function of self-presentation, suggesting that fairness is a
"desired social identity" in the context of an organization. In addition to the intrapsychic
benefits of creating a good self-image, Greenberg notes that the perception of fairness has
the important benefit of enhancing a manager's power base within an organization.
Additionally, an overall image of corporate fairness may serve to attract and retain
qualified employees. Haas-Wilson (1990) used survey research to examine the
association between well-informed recommendations and fees paid to psychotherapists.
the effects of different self-presentational strategies, gender, and position in an
A few studies have looked at entities as impression managers. For example, Amdt
and Bigelow (2000) used qualitative content analysis to examine the way hospitals
present accounts of corporate restructuring in their annual reports. Hooghiemstra (2000)
used a case study method to examine the rationale behind one corporation's
corporate social reporting to minimize the negative effects of environmentally suspect
Mass Communications Studies
Although there is a great deal of work in the field of mass communications based
on the dramaturgical work of Goffinan, particularly Kaid's
studies in the
field of political communication (Kaid & Johnston, 2001; see also Kaid, 2002; Kaid &
Davidson, 1986), only a small body of literature exists in the field of mass
communications that is directly informed by the theoretical framework of impression
management (Sallot, 2002). Dominick (1999) studied self-presentational styles on
personal Web pages using quantitative content analysis. Coding for Jones' (1990) self-
presentation strategies-ingratiation, competence, intimidation, supplication and
exemplification-he found that most personal Web pages could be categorized by the
dominant self-presentation strategy utilized by the author. Ingratiation and competence
were the self-presentation strategies employed most often by Web authors; supplication
and intimidation were almost never used. Dominick concluded that self-presentation
explained many of the cyber-behaviors in the study, including the use of feedback
Web pages, the study is still closely linked to the interpersonal roots of impression
management, studying self-presentational tactics of individuals in a mass mediated
Papacharissi (2002) used content analysis to examine the design tools people use to
achieve impression management goals on personal Web pages. For example, she viewed
feedback mechanisms such as e-mail and guestbooks as a gauge of desire for social
approval and hyperlinks as a gauge for need for affiliation and belonging. She noted
differences in creativity and use of hyperlinks among different Web communities, and
suggested that portals such as Geocities or Earthlink, may actually serve to mediate
impressions by making templates available to community members. By utilizing all the
"bells and whistles"
of a high-end template, a person projects a competent, high-tech
This study touches on the impression management of an entity in a mass mediated
context, to the extent that the portals are responsible for the impressions made by a
particular Web page. However, the focus of the study remained primarily individuals.
Sallot (2002) used an experimental design to test the effects of perceived motives
(none, selfish, prosocial, altruistic), communications styles (one-way and two-way), and
professionalism (licensed and unlicensed public relations professionals) on the reputation
of public relations professionals. She found that perceived motives made a difference in
how public relations professionals were regarded, with significantly better evaluations for
those engaging in activities for altruistic reasons than for selfish ones.
when participating in projects for altruistic reasons, public relations professionals were
still viewed a as advnoctes fnr their r.rnorate snnnsnrs Sallnt mlrroefts that this is related
advocacy" (p. 163). In other words, the general public already views public relations
professionals as self-serving and strategic in their actions on behalf of their companies, so
it is difficult to convince subjects that they are capable of working for purely altruistic
reasons. Again, this study is more interpersonal than corporate in nature, albeit focusing
on the impressions made by members of a mass media profession.
From an organizational standpoint, impression management is closely related to
corporate identity. Hooghiemstra (2000) describes corporate identity as "the way the
organisation presents itself to an audience" (p. 57). This is analogous to the concept of
self-presentation in interpersonal communications, which is an attempt to control
impressions of the self to an audience (Schlenker, 2002). An organization presents itself
to an audience every time that someone uses the company's
products or services, talks
with one of its employees, or encounters any consequences from its corporate actions.
Each of these interactions, therefore, involves some form of impression management.
Corporate communications are one tool organizations use to build corporate
identity (Hooghiemstra, 2000). According to Van Riel (as cited in Hooghiemstra),
corporate communications are "an instrument of management by means of which all
consciously used forms of internal and external communication are harmonised as
effectively and efficiently as possible, so as to create a favourable basis for relationships
with groups upon which the company is dependent" (p. 57). Goodman (1998) suggests
corporate communications includes such diverse communications functions as public
relations. investor relations. emolovee relations, media relations, labor relations,
Esrock and Leichty (1998) did look at the impressions made by entities-
specifically members of the Fortune 500-on their corporate Web pages. Their content
analysis found that corporate Web sites were used more for highlighting corporate good
deeds than for promoting specific policy decisions. This indicates that the desired
corporate identity is a socially responsible one. Their findings also suggest that few
corporations used their Web sites to actively communicate with relevant publics.
Apparently, the Web sites were not used to gather the feedback that is critically needed
for successful impression management (Schlenker, 1980). The authors note, however,
that their findings are likely "time-bound,
"evolutionary" stage in the late 90s. With
" since Web use by corporations was still at an
improved Web site design, such feedback may
now be easier to collect.
Most organizational and mass communications studies in the area of impression
management continue to look at individuals within organizations, rather than the
impressions managed by the organization as a whole. However, it is usually the corporate
entity, and not the individuals within that corporation, that makes the most direct
impression on the consumers. This dissertation fills a gap in the organizational
impression management literature, using an empirical design to study the effects of
excuses offered by a corporate entity in the wake of an adverse incident.
One thing that is clear from the literature is that most of the studies in the area of
organizational and media impression management have examined corporate
communications, usually as the subject of content analysis. In essence, corporate
nn^^^'^-*y'nnmfln; nnm +; -...i n nC 41-. -n .w n+-nc.- I i A y w +, eT f laiC
studies of corporate impression management were found in the literature, this, along with
the growing empirical literature in the field of crisis communications, suggests the
appropriateness of using corporate communications materials as the stimulus material for
investigating impression management in the corporate context.
Advertising and Impression Management
Advertising is a unique and very specific form of corporate communications.
Advertising is defined as "any paid form of non-personal communication about an
organization, product, service, or idea by an identified sponsor" (Belch & Belch, 1998, p.
14). To use the life-as-theatre analogy, an advertisement functions as a prop for an
organization, which helps an audience understand who (or what) the organization is, what
it stands for, and why it behaves in a particular way. Organizations choose to present
themselves to an audience with advertising because (1) it gives them complete control
over the message that they wish to communicate to an audience, and (2) it allows them to
decide into which media vehicle, or setting, they put that message.
When organizations use advertising, they pay money to describe themselves in the
exact way they wish to be described to a particular audience. Schlenker (1980) describes
written self-descriptions as a direct way of claiming specific images that an actor wishes
to claim. Advertising, therefore, must be seen as a very direct and not very subtle form of
corporate identification. As such, it represents an important tool for the study of corporate
identification and organizational impression management.
Problems with Advertising
From an organizational standpoint, advertising is an imperfect impression
numbers), a corporation gets no immediate audience feedback from its advertising
messages. Martinko (1991) states,
"Successful IM attempts require constant assessment
of audience reactions and feedback" (p. 269).
While the organization's
goals for the
advertising interaction are set in advance, advertising does not allow the organization to
modify its message to ameliorate an unsuccessful interaction-a bad or offensive ad
remains bad or offensive.
This potential problem is accentuated by the "permanent" nature of the advertising
interaction, since a record of the interaction, usually in print or video form, can remain
long after the interaction occurs.
We can therefore presume that organizations go to great
lengths to manage impressions right the first time when it comes to advertising. That is,
we can assume that organizations are highly motivated to achieve desired corporate
identifications through their advertising messages.
Second, advertising may suffer from the same problem that Sallot (2002) noted for
public relations namely, a lack of credibility. U.S. citizens comprise a sophisticated
media market they are highly aware that advertising is a paid form of communications,
and they are increasingly wary of advertising claims (Belch & Belch, 1998). Therefore, a
paid advertisement from a sponsor may seem less credible to the desired audience than a
newspaper article containing similar information about the organization
There is perhaps no time when it is more important for an organization to manage
impressions right in an advertisement than when an organization is running advocacy
rlxriorr Arrfrarlrtrna Arnr ctn vAmthec rnalled icnp a dvprhtiinO iS. "conrnCmed
such, it is a subcategory of corporate advertising, which is any form of advertising
designed "to establish, alter or maintain a corporation's
identity" (Schumann, Hathcote,
& West, 1991).
Sethi (1979) originally used the term "advocacy advertising" to describe
concerned with the propagation of ideas and the elucidation of controversial social
issues of public importance. It does so in a manner such that supports the position
and interests of the sponsor while expressly or implicitly downgrading the
sponsor's opponents and denying the accuracy of their facts. (p. 70)
However, Fox (1986) disagreed with Sethi's adversarial approach to the subject, and
suggested that advocacy advertising be defined as "advertising which addresses an issue
that can affect and can be influenced by actors beyond the advertising's
sponsor" (p. 62).
definition of "advocacy advertising,"
which encompasses a wide range of corporate
advertising designed to express corporate opinions on a broad array of subjects with
many relevant publics, will be used throughout this dissertation.
Because it is not product-oriented, Belch and Belch (1998) view advocacy
advertising as an extension of an organization's public relations activities, rather than as
an advertising function. Public relations is defined as "the management function which
evaluates public attitudes, identifies the policies and procedures of an organization with
the public interest, and executes a program of action (and communication) to earn public
understanding and acceptance" (p. 514). Advocacy advertising is a very direct attempt to
earn public acceptance of an organization's views. Therefore, previous public relations
research, more than advertising research, provides an appropriate framework for the
According to Dardenne (1981), corporations employ advocacy advertising most
often to correct public misconceptions in the wake of negative coverage by the press, and
to give the public information that may help alter public policy decisions. Corporate
advocacy advertising may be run proactively, in an attempt influence the opinions of
important audiences in advance of some action or event. This type of advertising might
be employed when a corporation wants to make its opinions known to legislators in
advance of some vote that may impact its operations. Corporate advocacy advertising
may also be run reactively, in response to some negative publicity or event. This type of
advertising is often run in the wake of corporate actions that have impacted its publics in
some negative way. For example, Connolly-Ahern and Kaid (2002) documented the use
of this form of advertising by a number of corporations in the wake of the tragic events of
September 11, 2001, and by the U.S. Postal Service after the 2001 anthrax attacks.
advertising is run in response to some negative corporate event or predicament, it can be
called "corporate account advertising." It is this type of advertising that is the subject of
Advertising Versus Public Relations
From a professional standpoint, one of the key benefits of public relations as a
marketing communications tool is said to be increased credibility. Beckwith (2003), a
well-known publicist writes,
"...publicity is more powerful and more influential than
because publicity is usually linked to newsworthy events or
information that make it useful or interesting" (p. 3).
College-level textbooks offer similar oninions.
Wilcox. Cameron. Ault and Agee.
magazines, radio, and television through news releases about a company's products or
services, community involvement, inventions and new plans" (2003, p. 16).
These third-party endorsements come in the form of media gatekeepers, such as
reporters, editors and producers, who transfer some of the credibility of the news vehicle
to a news release once it has been vetted and included in a newspaper, magazine or news
broadcast. Speaking about the relative credibility of new and traditional media, Hearit
(1999) concludes the traditional media gatekeepers are an important element of
credibility, suggesting that "when defending their reputations, companies are wise to
couple their use of the new media with the use of traditional media because its
employment of media gatekeepers confers to their message a legitimacy and credibility
not yet available in cyberspace" (p. 303).
Scholarly work has made the same assumption: "An advocacy message, appearing
in a print medium as purchased space, is likely to appear less credible than the same
message appearing in the news columns as news" (Salmon, Reid, Pokrywckzynski, &
Willet, 1985, p. 553).
However, the small body of work has examined the credibility and effectiveness of
public relations tactics as opposed to advertising for advocacy messages is inconclusive.
In an experimental study of the relative effectiveness of advertising and public relations
tactics, Salmon et al (1985) reported that while news stories were judged less biased than
advertisements for advocacy messages, in one case-that of the American Cancer
Society-advertising was actually judged more trustworthy, interesting, informative,
aorrPPhl1e and mnrp. in keeninp with behavior intentions than the same information
Working in the area of political communication, Zhao and Chaffee (1995) used
survey data to study the relative effects of television news and campaign advertising on
U.S. voter knowledge. Again, results were mixed.
While the effects of advertising were
generally less than the effects of news on issue knowledge, in one race--a high-profile
race fought on either side of a deep ideological divide--the effect on issue knowledge
attributable to advertising was greater than that attributable to television news.
Although organizations increasingly use advocacy advertising to take their
messages directly to relevant publics, there is only a small body of work comparing
advocacy advertising with more traditional public relations tactics.
While public relations
is generally believed to be more credible with audiences than advertising, there is no
definitive empirical study stating that news stories are superior to advertising in
delivering advocacy messages. Additionally, a search of the literature uncovered no study
that looked at the topic of advocacy advertising from an international standpoint. This
dissertation seeks to expand the small body of literature in this area of advocacy
advertising and public relations effects, while expanding the discussion of the topic to the
Predicaments and Impression Management
In the wake of a negative event, an organization is faced with a predicament. From
the impression management perspective, organizations are faced with predicaments
anytime in which "events have undesirable implications for the identity-relevant images
[they] have claimed or desire to claim in front of real or imagined audiences"
1 oQA n 1 9o\ A nradionnamnt an effect enmnrtte identity tn the extent that the
responsible a corporation is seen by itself and audiences for the negative event in
question, the greater the threat to its corporate identity. Schlenker et al. (1994) suggest
"responsibility is a necessary component of the process of holding people accountable for
their conduct" (p. 634). In the organizational context, it is the conduct of the corporation
itself that is in question.
Corporations and Predicaments
When faced with an identity-threatening predicament, organizations, like
individuals, are likely to attempt to avoid the negative consequences of their actions by
employing impression management tactics such as accounts and apologies (Schlenker,
1980). In response to predicaments, a corporation has five communications choices in the
wake of an adverse incident. The first, of course, is not to communicate with any of its
relevant publics. This is the classic legal approach of "no comment" (Fitzpatrick &
Rubin, 1995). Once the organization decides to communicate, however, it can either issue
some form of account (a defense of innocence, an excuse, or a justifications) or an
From a communications perspective "corporate accounting"
can be accomplished
through a variety of communications channels, depending on the audience that the
organization is targeting with its account: internal communications, annual reports, press
releases, advertising, or any combination of these.
The Triangle Model of Responsibility
Deciding whether or not a corporation is responsible or accountable for some
negative action involves "evaluative reckoning" in which the corporation is judged by its
specific domains to assess responsibility for an event (Weiner, Perry, Magnusson, 1988).
The triangle model of responsibility offers one model for the information needed to judge
responsibility for a negative event, based on information
about three main components and the linkages or connections between them (Schlenker,
Pontari, & Christopher, 2001; Schlenker et al., 1994).
According to Schlenker et al. (1994), the three main components of the triangle
(a) the prescriptions that should be guiding the actor's conduct on the occasion, (b)
the event that occurred (or is anticipated) that is relevant to the prescriptions, and
(c) a set of identity images that are relevant to the event and prescriptions and that
describe the actor's
roles, qualities, convictions, and aspirations. (p. 634)
The relevant linkages of the triangle model, according to Schlenker et al. (2001),
(a) prescription clarity, the clarity of the prescriptions (goals, rules, and scripts)
that are applicable in the situation, linking the prescriptions and the event; (b)
personal obligation, the extent to which the actor appears to be bound by the
prescriptions because of duty or other requirements, linking the prescriptions and
the actor; and (c) personal control, the extent to which the actor seems to have
control over the outcomes in the situation, linking the actor and the event. (p. 18)
The three components and the links between them form a triangle (Figure 1). The
stronger the links between the three components of the model are, the greater the
's responsibility for the event in question will be judged by its audiences.
Excuses are an impression management tactic employed by corporations in an attempt to
weaken the links between the components, thereby lessening corporate responsibility for
the events in question in the minds of the relevant audiences (Schlenker et al., 2001).
A number of studies have incorporated the triangle model to assess responsibility
related to the perceived strength of the links of the triangle model, with a combination of
strong prescription-identity and strong prescription-event links producing the highest
perceived responsibility ratings.
One study related a common American cultural characteristic, Protestant Work
Ethic (PWE), to attributions of responsibility using the triangle model (Christopher &
Schlenker, in press). The researchers found that high PWE is related to holding people
personally responsible for their actions, regardless of the excuses offered, as well as
negative reactions after failure. Although this was not an intercultural study, per
suggests the triangle model is appropriate for measuring differences in reactions to
excuses among different cultural groups.
Note. From "Excuses and Character: Personal and Social Implications of Excuses," by
B.R. Schlenker, B.A. Pontari, and A.N. Christopher, 2001, Personality and Social
Psychology Review, 5
p. 18. Copyright 2001 by Lawrence Erlbaum Associates, Inc.
Types of Excuses
Each of the linkages in the triangle model can be weakened by specific kinds of
excuses. Prescription clarity excuses claim that the goals, rules and standards pertaining
to the event in question were unclear to the organization (Schlenker et al., 2001).
Microsoft employed this kind of excuse in the aftermath of the Justice Department's
trust lawsuit against it. Microsoft claimed that there were no laws governing the creation
of proprietary software, and that it therefore had no way of knowing that it might be
considered illegal to create software that automatically disabled competitor's
Organizational obligation excuses claim that because of some attribute of the
corporation, the normally expected prescriptions for the event do not apply to the
corporation, regardless of how they may apply to others in a similar situation (Schlenker
et al., 2001).
While the basic values of fairness and equality might seem to weaken the
effectiveness of this kind of excuse with audiences, it is possible to imagine a scenario
where a corporation might employ such an excuse.
For example, a pharmaceutical corporation that produces mass quantities of vaccine
for children might employ such an excuse.
While it would generally be contemptible to
create a product that might harm a child, vaccine producers are aware that there is a small
risk that a child will become ill, or even die, as a result of being vaccinated with one of
their products. However, because the vaccinations are required by the state, vaccinations
are generally credited with saving far more lives than they harm, and the risk is widely
understood by health professionals, pharmaceutical companies continue to make vaccine.
company could not be judged by conventional standards of product liability, since it was
obligated to produce vaccines that save countless lives.
Organizational control excuses claim that the corporation had no way to manage or
direct the event in question (Schlenker et al., 2001). Usually, the event in question is
attributed to external forces. United Airlines famously employed a series of control
excuses in the aftermath of the September 11,
2001, terrorist attacks. Having lost two of
its planes, along with crew and passengers, during the event, United Airlines responded
to the negative publicity surrounding the event, and the subsequent fear of flying by the
American people, with a series of advertisements that blamed the tragic events on an
" who had attempted to take the joy of flying away from the highly
In rhetorical terms, excuses fall under the differentiation strategy of conflict
resolution, because they seek to separate some piece of information-information on
prescription clarity, personal obligation or personal control-from the larger context of
the adverse incident (Ware & Linkugel, 1973). The goal of separating this information
for the audience is to create audience understanding, which will decrease the likelihood
of audience condemnation of the person or entity offering the excuse. If the person or
entity offers the excuse along with positive information about itself or its connections
with other entities, it is taking an explanative excuse posture. If the excuse is offered in
combination with some kind of denial, the person or entity is taking an absolutive excuse
WhilP mmranrntinn mav pmnlnv any nf the tvnes nf Yr1.n1es mentionedn some mav
literature contains a number of studies measuring the effectiveness of various accounting
strategies (e.g., Benoit & Drew, 1977; Kane, Joseph, & Tedeschi, 1977; Riordan, Marlin,
& Kellogg, 1983), the results have been mixed. According to Benoit (1995), only
apology has been found to be generally effective as an image restoration strategy.
Furthermore, without a uniform typology of self-defense mechanisms, the results of
these studies are difficult to compare (Benoit, 1995). The proposed experiment seeks to
extend the literature in the area of credibility of accounts, by testing the relative
credibility of narrow sub-category of self-defense strategies, excuses, which are likely to
be employed by corporations in the wake of crises. This study further extends the
literature through its examination of the acceptability of an account by a corporate entity,
rather than an individual.
Problems with Excuses
People employ excuses when confronted with predicaments because excuses can be
helpful in minimizing the negative repercussions of their mistakes (Schlenker et al.,
2001). Snyder and Higgins (1988) suggest a number of intrapersonal benefits to using
excuses, including helping those who employ them protect self-esteem and improve task
effectiveness, while reducing anxiety and depression. Studies have also noted
interpersonal benefits of excuses, including minimizing damage to the personal identity
of the person making the excuse, as well as the reduction of sanctioning after an event,
including reduced punishment in the wake of a crime (Critchlow, 1985). Relating the
practice of public relations to excuse making, Schlenker et al. (2001) suggest, "The recent
feature of the 'sPin doctor' position in most laree comPanies illustrates the importance of
feature of the 'sum doctors position in most large companies illustrates the importance of
However, while excuses may serve to protect identity in the event of a predicament,
there are also risks associated with using them (Schlenker et al., 2001). In particular,
Schlenker et al. suggest that excuses can undermine reliability of the actor using the
excuse, which may lead to impressions of deceit, ineffectualness or self-absorption.
However Keller (1998), who defines corporate credibility as "the extent to which
consumers believe that a firm can design and deliver products and services that satisfy
customer needs and wants" (p.426), claims that corporate credibility depends on three
factors: trustworthiness, expertise, and likeability. Yet these are precisely the factors
which may be undermined by excuses, as noted by Schlenker et al (2001).
Both Ohanian (1990) and Newell and Goldsmith (2001) suggest that of those three,
expertise and trustworthiness are the most salient dimensions of corporate credibility.
Excuses of any kind may therefore be related directly to diminished corporate credibility,
with those that undermine expertise and trustworthiness the most damaging to corporate
credibility. Coombs and Holladay (1996) suggest that expertise may be especially
"Organizations that might be perceived as being able to prevent a
crisis should have a more negative image [in the wake of a crisis] than an organization
perceived to have little or no control over a crisis" (p. 293).
When making an excuse, especially in the advertising medium, it appears that
corporations walk a thin line between wishing to minimize their responsibility for a
particular incident or event, while protecting their corporate credibility and, ultimately,
their corporate identity. United Airline's
post 9-11 advertisements, for example, were
very careful to leave the other two linkages extremely strong, however, emphasizing both
the airline industry (obligation). The ads featured actual employees of the airline giving
personal testimonials about their love of flying and their commitment to the corporation,
making United-and its employees-seem as trustworthy and likable as possible. This
served to underscore United's
corporate credibility, and to salvage as much as possible of
corporate identity, in spite of facing a predicament that was in direct conflict
with its desired corporate identity as an extremely safe and efficient airline.
Culture and Impression Management
Appropriate self-presentation is related to an individual'
cultural mores. Leary
"Confusion and misunderstanding can arise when people from one culture engage
in self-presentations that do not conform to the norm of another culture" (p.1996, p80).
He says the subject has been inadequately explored in the impression management field,
"as our global society expands, we have increasing reason to understand
cultural differences in self-presentation" (p. 80). Bond (1991) suggested,
approach to examining cross-cultural variation in self-presentation is to examine
differences in values around the world" (p. 197). However, while the need for
intercultural research in the cross-cultural area is clear, a review of the literature could
find no cross-cultural studies using the impression management theoretical perspective.
This dissertation begins to fill an existing gap in the literature, improving knowledge of
cultural differences in effects of impression management used by corporations in the
wake of adverse incidents.
The nature of culture is a matter of great debate, even within the field of
most comprehensive definitions in the field of anthropology comes from Kluckhohn
Culture consists in patterned ways of thinking, feeling and reacting, acquired and
transmitted mainly by symbols, constituting the distinctive achievements of human
groups, including their embodiments in artifacts; the essential core of culture
consists of traditional (i.e. historically derived and selected) ideas and especially
their attached values. (p. 86)
A few frameworks exist for the comparative study of culture (De Mooij, 2001).
One of the most widely used is that of Dutch organizational anthropologist Geert
Hofstede. Hofstede (2001), who based his cultural theories on a worldwide survey
administered by IBM to its employees, describes culture as the "collective programming
of the mind that distinguishes the members of one group or category of people from
another" (p. 9). Culture includes language, attitudes toward elders, the physical distance
preferred between individuals, and perceptions of basic human activities such as eating,
sexual relations, and even defecation (p. 2). According to Hofstede,
itself in visual elements,"
such as symbols, heroes, rituals and other practices.
these elements may be visible to people outside of the group, only people inside the
group fully comprehend the cultural meanings of these elements, which remain invisible
to a casual observer (p.10).
Advertising and journalism are visible elements of culture. Like any form of
expression, they are the end product of a long process of message encoding and decoding
that reflects the attitudes and values of the source that creates it-and is viewed through
the attitudes and values of those who encounter it. Advertising and journalism can never
be totally separated from the culture that produces them. They are social artifacts, and it
Hofstede (2001) identified five independent dimensions of culture:
"Power distance (PDI), which is related to the different solutions to the basic
problem of human inequality" (Hofstede, 2001, p.29). Hofstede says that within
organizations, "Power distance is a measure of the interpersonal power or influence
between [an employer and an employee] as perceived by the less powerful of the
two" (p. 83). He suggests that each culture has a norm for this level of interpersonal
power, which is measured by PDI.
"Uncertainty avoidance (UAI), which is related to the level of stress in a society in
the face of an unknown future" (Hofstede, 2001, p.29). Since uncertainty leads to
anxiety, each society or culture must find a means for effectively dealing with
uncertainty. According to Hofstede, "Ways of coping with uncertainty belong to
the cultural heritages of societies, and they are transferred and reinforced through
basic institutions such as the family, the school, and the state" (p. 146). Essentially,
UAI is a societal measure of discomfort with ambiguity.
"Individualism versus collectivism (IDV), which is related to the integration of
individuals into primary groups" (Hofstede, 2001, p.29). Basically, this refers to the
relative importance of individual or group achievement in society. According to
Hofstede notes, "The relationship between the individual and the collectivity of
human society is not only a matter of ways of living together, it is intimately linked
with societal norms" (p. 210). It is the balance of this relationship that is measured
"Masculinity versus femininity (MAS), which is related to
emotional roles between men and women within a society'
The author explains that masculinity and femininity "refer
role patterns in the vast majority of both traditional and me
patterns of male assertiveness and female nurturance" (p.
level to which gender role patterns differ within countries,
the division of
'(Hofstede, 2001, p.29).
to the dominant gender
!84). MAS measures the
with high MAS meaning
great differentiation, and low MAS (or femininity) meaning little differentiation.
"Long-term versus short-term orientation (LTO), which is related to the choice of
focus for people's efforts: the future or the present" (Hofstede, 2001, p.29). Long
term orientation includes values such as persistence, observing status in
relationships, thrift and having a sense of shame; short-term orientations includes
values such as personal stability, protecting "face," respect for tradition and
reciprocation of greetings, favors and gifts. This dimension is based on Bond's
Confucian Values Survey, and has only been measured for a few non-Eastern
cultures. Because information on LTO is available for only a few countries, it was
not considered in this dissertation.
Each of the 53 countries in Hofstede's sample were positioned on a scale for each
dimension, with the dimensions occurring "in all possible combinations, although some
significant correlations between Hofstede's
cultural dimension scores and other external
variables, including consumer attitudes and attitudes toward advertising.
Criticisms of Hofstede's Approach
A number of interpersonal and mass communications studies, as well as a wide
range of studies in other areas, have employed Hofstede's
cultural value system module
as a framework for cross-cultural research (e.g., DeMooij, 1998b; Gudykunst et al., 1996;
Tak, Kaid, & Lee, 1997
validity of Hofstede's fi
Taylor, 2000). However, some scholars have questioned the
Spector, Cooper and Sparks (2001) were unable to reproduce Hofstede's
consistency statistics in their 23-country sample on either the individual or
country/province level. Furthermore, their ANOVA results indicated only one of
accounted for more than 10% of the variance in the replication
of the values survey. Schimmack, Oishi, and Diener (2002) found that cultural values of
individualism-collectivism were moderated on the individual level by gender, but gender
differences are not considered within the structure of the Hofstede values.
The Hofstede Values System Module cannot be considered a perfect measurement
tool of culture. First, because the Values Systems Module reports country-level means, it
does not take into account the wide range of individual variation that exists within any
country (Hofstede, 2002). Any convenience sample taken from a country is likely to
differentiate extensively from the reported country means. This differentiation may be
exacerbated when samples cannot be secured from different geographic areas within a
of criteria, including but not limited to age, gender, education level, occupation, and
socio-economic status, can prove daunting in a cross-cultural context, especially when
international researchers are likely to have limited access to local populations, and
samples are based on convenience rather than chosen in a systematic, random fashion.
Despite the problems of using the Hofstede measurements in cross-cultural
research, hundreds of scholars from around the world have documented strong,
significant correlations between the cultural values measurements and a diverse range of
outside criteria, from purchases of mineral water to average waiting times for doctors and
even the Nobel Prize index (Hofstede, 2001). The measurement is not perfect, but it the
best one currently available, and its use is justified in the present study, especially in light
of the growing body of communications and marketing literature utilizing the scale.
Culture and Communications
Noted U.S. American anthropologist Edward Hall said,
and communication is culture,
"Culture is communication
" (1959, p. 169). It is clear that communication, both
interpersonal and mediated, is inseparable from culture-both of the sender and of the
receiver. According to Gannon,
"perhaps the most interesting feature of culture is that is
triggers unconscious values leading to action" (2001, p.18). He indicates that culture is an
important variable in the communication process, particularly in the realm of persuasive
communication, such as business negotiations, where one party desires a specific
outcome as a result of the reception of a communicated message.
Working in the field of interpersonal communication, Ting-Toomey and Kurogi
(1998) introduced a model ofintercultural conflict management based on the culture-
"The dimension of individualism-collectivism serves as a conceptual
framework in explaining why the meaning of "self' and hence,
(p. 189). The essay includes suggestions for creating "intercultural facework
competence" to enhance human dignity in the increasingly global environment.
Gudykunst (1997) notes that while most anthropological research takes an "emic"
or specific approach, and most psychological research uses an "etic" or universal
approach, communications research is derived from both the emic and etic viewpoints. In
his introduction to a special issue of Communication Research, Gudykunst (1997) stated
cultural values provided a means for comparing similarities and
differences in national communication styles. However, he cautioned that researchers
should also attempt to take into account individual-level factors that might impact the
influence of culture on communication style. Some individual-level measures that might
account for differences in communications style include personal values, personality
orientation, egalitarianism, and sex role orientation.
Culture and Advertising
According to Tak, Kaid, and Lee (1997)
Many researchers agree that culture does affect the reception and acceptance of an
advertising message; the advertising message therefore reflects the culture in which
it appears. In communication...a send
message form, whereas the receiver's
her's cultural background affects the overall
cultural background determines message
perception. ( 2)
Their content analysis research found that, among other things, UAI was a good
determinant of Korean and U.S. American political advertising styles.
De Mooij (2001) also used a content analysis of international advertisements to
significant correlations between consumer attitudes toward advertising and IDV and
MAS (Hofstede, 2001). This indicates that these two values may be useful tools for
developing advocacy advertising research.
According to De Mooij (2001), highly individualistic cultures, such as the one in
the United States, tend to be universalistic, whereas collective cultures, like most of those
in Latin America, tend to accept cultural differences of other groups, because they are
comfortable with the idea that their group is different from all others. In terms of
communications, De Mooij says this can be reflected in "assuming certain truths as self-
evident and commenting on events in other parts of the world from [a] universalistic and
ethnocentric point of view"
(p. 79). She notes that this is often manifested in international
advertising and marketing campaigns.
High MAS societies are ones in which performance and achievements are prized.
These are the classic "bigger is better countries," including Japan, Germany, and the
While the United States is often thought to be a very masculine society, in
actuality fourteen countries, including four Latin American countries-Venezuela,
Mexico, Colombia and Ecuador-score higher than the United States on the MAS scale.
Low MAS societies are service- and people-oriented (De Mooij, 2001). All the
Nordic countries fall into this category. In the Americas, Costa Rica, and Chile are found
on the extreme low end of the MAS scale, and seven of the 13 Latin American countries
score below the mean.
De Mooij (1998b) suggests that culture is not merely reflected in advertising.
1I ^.4. a -- lS-l .^...l t^^ nO nkJ4-Ah J5. Anl An nl A n A k-J> 445 fl. a. tA nA~qS in int nnl ,. In 'L *U nn| nu ,A li."/ rt....I. .,t/-H1- W
In particular, De Mooij found strong positive correlation between confidence in the
advertising industry, Masculinity and Uncertainty Avoidance in a survey of European
nations. In particular, she found that feminine countries weak in uncertainty avoidance
reflect a deep skepticism of the exaggerations in advertising.
Content analysis reveals different advertising executions are correlated with the
MAS dimension (De Mooij, 2001):
The masculine/feminine dimension discriminates between cultures particularly with
respect to values related to winning, success and status, which are much used in
advertising appeals. It is therefore an important dimension for marketing and
selling. It reflects the division between countries in which hype and the hard sell
prevail and countries with a soft-sell, more modest approach. (p. 82)
Not all scholars have considered MAS critical to the understanding of advertising,
however. Zandpour et al. (1994) and Zanpour and Harich (1996), for example, offered a
model to assess cultural fit for advertising using Hofstede's
cultural values, omitting the
MAS dimension. However, Hofstede (1998) called the model incomplete because of the
omission, which Hofstede considered a critical error in the area of advertising. He
suggested the dimension was omitted from the model because of an American cultural
bias against discussing gender, and noted,
"It seems paradoxical that academics operating
within a culture accustomed to focusing on facts, data, and truth can be so biased in their
research approach" (p. 72).
Culture and Public Relations
According to Sriramesh and VerciC (2003),
"Logically, culture should affect public
relations and public relations helps alter culture" (p. 7). However, the authors note that
researchers have only recently begun to explore the subject of culture and public
Because public relations in general, and international public relations specifically,
is a relatively new area of communications study, much of the work that has been done in
the area of culture and public relations has focused on case studies and ethnographies
(e.g. Sriramesh, 1996). In fact, The Global Public Relations Handbook (Sriramesh and
VerCiC, 2003) is written with one chapter about each country, with very little comparative
work attempted. Some comparative work has been done (e.g. Sriramesh, Kim, &
Takasaki, 1999; Taylor, 2000), but an extensive review of the literature revealed no
cross-cultural public relations research that is experimental in nature.
Culture and Excuses
The MAS dimension may be important in a culture's
relative willingness to accept
excuses. According to Hofstede (1998), feminine cultures are more permissive than
masculine ones. In particular, Hofstede found a strong negative correlation between an
index for permissiveness, based on the European Value Systems study, and masculinity.
That is, people in feminine countries were found to be more accepting of a list of vices
that included joyriding, soft drugs, accepting bribes and prostitution than their
counterparts in more masculine countries. It seems reasonable, therefore, that individuals
from feminine countries might accept excuses more willingly than their counterparts in
Additionally, masculine countries are known to hold success and winning in high
While feminine countries show sympathy for the weak, masculine countries show
sympathy for the strong (Hofstede, 2001). Due to their success orientation, masculine
piilhiiyr o ra~r nt nrnvirs jlace FIIoo-RInrr kM I fhRnn+'~r onAnntnn af c ffjoraA11 r nnrnnratnnc in
Of course, it is logical to assume that other cultural values are also related to the
acceptance of excuses. For example, Hofstede (2001) notes that one characteristic of high
uncertainty avoidance cultures is pessimism about organizations' motives. Taylor (2000)
linked uncertainty avoidance, power distance and organization-public relationships in a
cross-cultural case study about the aftermath of a Coca-Cola tainting scare in Western
Europe. High uncertainty avoidance/high power distance countries, including Belgium,
France, and Spain, were distrustful of the corporation after the scare, and quickly pulled
Coca-Cola from the shelves, while weak uncertainty avoidance, low power distance
countries such as Sweden, Norway, and Denmark waited for conclusive evidence of
tainting before recalling Coca-Cola'
products. Taylor's observations have implications
for this dissertation, suggesting that Spanish audiences may be particularly distrustful of
large corporations. It should be noted that while Spain does rank very high in uncertainty
avoidance, its power distance ranking (31) is considerably lower than that of either
France (15/16) or Belgium (20).
A review of the literature indicates that there is reason to believe that cultural
values will influence acceptance of excuses. Additionally, culture has already been
shown to influence the reception of corporate communications messages emanating from
the advertising and public relations tradition. However, an extensive review of the
literature uncovered no studies testing either of these propositions empirically.
Hypothesis and Research Questions
Given the research reviewed in the areas of impression management, dimensions of
culture and mass communications, two hypotheses and a number of research questions
Hofstede (2001) suggests that people who live within a group are "programmed" to
react to events or actions in similar ways. Scott and Lyman (1968) specifically suggest
that acceptance of excuses may be related to culture. However, to date no empirical
studies have analyzed the relationship between cultural values and corporate credibility.
RQ1: What is the relationship between Hofstede's cultural values and the
credibility of corporate excuses?
Schlenker et al. (2001) offer a typology that includes three kinds of excuses.
However, there is no evidence in the impression management literature indicating which
kinds of excuses are most likely to be credible with audiences.
RQ2: Which types of corporate excuses-prescription clarity, personal obligation
or personal control-are most credible with audiences?
Authors such as Belch and Belch (1998) and Beckwith (2003) indicate that
advertising may suffer from a general lack of credibility. However, a search of the
literature revealed mixed results that fail to validate this claim.
RQ3: Are corporate excuses offered in advertising less credible than the same
excuses offered within the context of a news article?
RQ4: What is the relationship between Hofstede's cultural values and the relative
credibility of excuses offered in advertising and within the context of a newspaper
(1998a) research indicates that high MAS cultures view advertising
more positively than low MAS cultures. Hofstede (2001) indicates that masculine
cultures have more confidence in advertising than feminine cultures.
HI: Overall, members of higher MAS cultures will find excuses offered in
advertising more credible than members of lower MAS cultures.
Conversely, Hofstede (2001) indicates that feminine cultures are more permissive
H2: Overall, there will be less difference in credibility scores after an adverse
incident in low MAS cultures than in high MAS cultures.
This posttest only with control group (Babbie, 1998) experiment was designed to
determine which kinds of excuses are most credible with audiences from different
cultural backgrounds. This was done by exposing subjects to three corporate accounts of
an adverse incident that were manipulated to give a different kind of excuse for the same
event-one of clarity, one of obligation, and one of control. Audiences received the
corporate account in one of two print media forms: an advertisement or in the body of a
newspaper article. To begin the exploration of the relationship between Hofstede's
dimensions of culture and corporate credibility, the experiment was run in two different
countries, the United States and Spain. As illustrated by the cultural dimensions of these
countries, which are outlined in Table 1, the two countries offer wide contrast, in terms of
rank-order, on three of the four variables under study; the country ranks differ by at least
twenty places in all categories except for power distance, where the difference in rank is
only 7 places.
Table 1. Hofstede's
cultural values indices and ranks for Spain and the United States
Power Uncertainty Individual/ Masculinity/
Country Distance Avoidance Collectivism Femininity
Spain 57/31 86/10-15 51/42 42/37-38
United States 40/38 46/43 91/1 62/15
From Culture's Consequences, by Geert Hofstede, pp. 500-501. Copyright 2001 by Sage
The cover story for this experiment was based on a corporate crisis-an
explosion-at the Winnipeg, Canada (for U.S. participants) or Cordoba, Argentina (for
Spanish participants) production facility of a fictitious multi-national chemical
corporation called "MultiChem Corporation." The two cities were chosen because they
were deemed to be secondary, industrially oriented cities in countries with which
participants would be familiar but were not so large as to have many participants have
personal acquaintances in either of them, as might have been the case in Toronto or
Buenos Aires. The experimental materials consisted of a "scrapbook" of media related to
MultiChem. The first three pieces in each scrapbook included materials designed to
create a credible image for the fictitious corporation; the fourth item informed
participants about an adverse incident at a MultiChem facility in a foreign country; and
the fifth item offered a corporate excuse for the incident. Recruited participants from each
country were randomly assigned to one of six experimental conditions. Each scrapbook
was followed by a one-page questionnaire consisting of 16 semantic differential scale
items designed to assess the credibility of the fictitious corporation.
Materials were first created in English, translated into Spanish by a native Spanish
speaker fluent in both English and Spanish, and then back-translated by another fluent,
tri-lingual speaker to ensure the translations were sound and that the original meaning
was not lost.
All participants received corporate communications and news materials that
to be a printout of the company's Internet homepage. This homepage was designed by a
commercial designer who modified the elements from the Web sites of major chemical
corporations, including BASF, Dow Chemical, DuPont and Hoechst, to increase the
authenticity of the stimulus. The text and graphics were designed to create an image of a
highly credible corporation, highlighting the traits of trustworthiness and expertise. To
highlight trustworthiness, the site contained links for "Environmental Policies,"
a link for "Philanthropy."
as well as
The mission statement highlighted the fact that the company
was "dedicated to discovering and implementing the most environmentally friendly
" To highlight expertise, the site referred to the corporation as an
"innovator" operating on six continents. The Latest News link included headline about
new marketing initiatives in New Zealand, and a new research facility in India.
The next two pieces were mock articles from fictitious newspapers: the Winnipeg
Times or Cordoba Hoy. The first article, entitled "CAS honors MultiChem Scientists,"
("ACA honra a cientificos de MultiChem,
" in Spanish), focused on the expertise of the
corporation, listing the numerous awards received by MultiChem scientists, including a
"Nobel Prize for their identification of an enzyme that aides in composting of agri-
industrial waste," (Appendix B). The second article, entitled "City greets new MultiChem
" ("Se inaugura nuevo centro de investigaci6n de MultiChem," in
Spanish), focused on the company's trustworthiness, detailing its 30-year commitment to
the Winnipeg or Cordoba community, through charitable donations totaling more than
1 ( million fnr "htident dlisahled ritizen.ns inle mothers. need children." among
A control group went directly from the corporate communications materials to the
corporate credibility scales to determine the baseline credibility of the fictitious
All participants aside from those in the control group next read what appeared to be
a copy of a newspaper article from the same fictional newspaper entitled "Explosion
injures 3, evacuates 100s,
" ("Explosi6n deja a 3 heridos, cientos de personas evacuadas"
in Spanish). (Appendix D). The article described the "late night" explosion in neutral
terms, giving factual data about injuries and evacuations. No sources are cited directly; a
statement from MultiChem was anticipated at a morning news conference. An accident
was chosen because it was deemed to be relevant to testing issues of responsibility and
credibility arising from excuses (Appendix D). The scenario was similar to the one used
by Coombs (1999) to test the effectiveness of instructing information and compassionate
messages in crisis communications.
Next, each participant received a corporate account of the incident. These were
either in the form of advertisement from the MultiChem (the advertising condition) or
contained within a newspaper article in which a corporate spokesperson was quoted (the
news story condition), offering an identical excuse.
Crisis Communication Elements
The elements included in the crisis communications were derived from Barton's
(1993) list for using communication as a management tool. The audience for the
communication was local citizens and opinion leaders in the affected communities of
Winnipeg or Cordoba. The goal of the communication was to restore MultiChem's
community, as well as assurances that the incident would not be repeated. The messages
assumed that the receivers would have had knowledge of both the corporation and the
incident. The source for the communication was a corporate "vice president." No
additional credentials were given, aside from those of the company in the earlier
corporate communications materials.
The materials were designed to minimize the influence of external factors on the
honoring of the excuses. Therefore, no comments were offered by either community
leaders or governmental agencies either against or in support of the corporation.
Implementation of the Triangle Model
To ensure that the excuse treatments did not differ in tone, the same introductory
paragraphs and closing paragraphs were included in each excuse. Using Ware and
(1973) typology, the company's
posture was explanative. That is, it sought
both to bolster its image, reminding the audience that the corporation has been a good
corporate citizen, improving quality of life for citizens, and to put distance between the
corporation and the event. The excuses were the distancing mechanism.
Newspaper stories included introductory paragraphs noting that company officials
had "identified the cause" of the explosion, along with a statement that the company had
assured a local governmental agency that there was "no health risk to the community"
and no chance that the incident would be repeated. The excuse appeared as a statement
from a company official made to "reporters."
A concluding paragraph gave the reader
some general information about MultiChem, including the global scope of its operations
and its long history in either Winnines or Cordoba.
obligation. The prescription clarity excuse stated that the explosion was the result of an
unanticipated interaction between two chemicals that were previously thought to be safe
when used together, but which the company has now determined are only safe when used
below a certain temperature (the clarity excuse condition). The personal obligation
excuse stated that the company uses volatile chemicals because of important work they
do under contract for the government following strictly prescribed procedures, and that
the government had assured them of the safety of the chemicals before they undertook the
work (the obligation excuse condition). The personal control excuse stated that the
company has determined that the cause of the explosion was a faulty valve in a holding
tank, which caused a chemical reaction between two chemicals that never should have
been allowed to interact (the control excuse condition) (Appendices E-G).
After reading the corporate excuse, each participant was asked to answer a 16-item
semantic differential scale constructed for this dissertation (Appendix H). The items
chosen were derived from previously used credibility scales. The items included in the
current measurement instrument were those that were deemed to best represented the
trustworthiness and expertise measures in the various scales. Duplicate measures were
eliminated and wording was modified to create a scale of bi-polar items, rated on a 7
The first was scale used was Berlo, Lemert, and Mertz's
(1970) Source Credibility
Scale (Appendix I). The scale employs a 15-item randomly ordered and reversed bi-polar
adjective scale. renresenting three distinct factors: safety. qualification. and dynamism
scholars in both the interpersonal and mass communication fields since the 1970s, and
has also served as a starting point in the creation of other scales (Rubin, 1994a).
The second scale used was McCroskey's (1966) 12-Item Semantic Differential
Scale (Appendix J), which was derived from both rhetorical and social psychological
frameworks (Rubin, 1994c). This scale was chosen because it has been used in numerous
communication studies, including some that examined the impact of culture-related
variables on source credibility, including rate of speech and gender (Wheeless, 1971),
nonverbal cues (Arnold, 1973) and social status and dialect (Bochner & Bochner, 1973).
Coombs and Holiday (1996) adapted this scale to measure corporate image for an
experiment assessing the fit of crisis type and crisis communication strategy.
The third scale used was Gaziano and McGrath'
(1986) News Credibility Scale
(Appendix K). Unlike the Berlo and McCroskey scales, which were developed for the
study of interpersonal communication, Gaziano and McGrath developed this instrument
for the study of mass communication messages, specifically news articles. The scale was
originally created for the American Society of Newspaper Editors, who sponsored a study
of local print and television journalism (Rubin, 1994b). A modified version of this scale
was recently used by Ferguson (2003), to gauge the relative credibility of alignment
advertisements of for-profit and not-for-profit organizations (Appendix L).
The fourth scale used was Newell and Goldsmith'
(2001) Corporate Credibility
Scale. This is an eight-statement, 7-point Likert-like scale, specifically designed for the
study of corporate image. Newell and Goldsmith describe corporate credibility as "the
^,,r*n.J 4a w,1.Lt nnn. tan1r +lnn+ itta Cwn' knn +lnn ljrnn'e rrlaArlr a a,. 0kiltr fn nfhlfl11 1
researchers sought to remove elements from previous credibility scales that made little
sense in a corporate context, such as "sexiness"
and "handsomeness." Therefore, this
scale focuses on two components of source credibility: expertise and trustworthiness
The time frame for recruiting and data collection was between March and April of
2004. The Spanish data was collected on March 9 and 10, 2004. The dates of this
collection are particularly important in light of the March 11, 2004, terrorist bombings at
the Atocha train station in Madrid, Spain. Since all of the data were collected before the
bombings occurred, it does not serve as an intervening variable in this study (Agresti &
Finlay, 1997). However, as will be discussed in the Limitations section of this
dissertation, the events of"l11-marzo," as it is known in Spain, may have significantly
affected the attitudes of Spaniards toward the rest of the world in general, and foreign
corporations specifically, in the short time since the data were collected.
The sample consisted of undergraduate students from large, academically
competitive universities, one in the southeastern United States (n
= 214) and two in
central Spain (n=223). Although every effort was made to ensure the equivalency of the
two groups, it must be acknowledged that this research, like most cross-cultural research,
suffers because it is almost impossible to ensure group equivalence across cultures.
Under these circumstances, this dissertation methodology is appropriately termed a quasi-
experiment (Stacks & Hocking, 1992).
Students were recruited from communications classes in both locations and offered
the sample, Babbie (1998) indicates that this is an acceptable form of recruiting for
exploratory research like that undertaken in this study.
Students were randomly assigned to one of the six excuse conditions upon arrival at
the testing site. Cells were not conducted individually. Instead, scrapbooks were
randomly ordered using a random number table (Table of Random Numbers, n.d.) prior
to conducting the experiments, and distributed to students at the test site. The full
experiment took approximately 15 minutes to complete; the control group took only
about ten minutes.
Because the control group was to fill out the credibility scales directly after the
corporate communication and mock news articles underscoring MultiChem's
and trustworthiness, the control groups were conducted separately from the other cells.
While this may introduce the possibility of selection bias into the results (Babbie, 1998),
there is no reason to believe that the group of students selected to complete the control
questionnaire in either country was materially different from the other students recruited
to fill out the other conditions.
First scale items were subjected to factor analysis. The resulting credibility scores
were analyzed using a 2 (country-Spain or United States) x 2 (format-ad or
newspaper) x 3 (excuse-clarity, obligation, control) factorial ANOVA.
To assess the impact of the cultural variables on credibility scores, Hofstede's
cultural values scores from the country of origin of the participants
- MAS, IDV, UAI
States. All Spanish participants, therefore, were considered low masculinity, low
individuality, high uncertainty avoidance, and high power distance condition, by virtue of
their country of origin, without any individual-level testing for those variables.
The reliability coefficient for the sixteen-item corporate credibility scale was .9540
a). Control group (n=57) means for scale items ranged between 4.77
(SD=1.36) ("doesn't/does make truthful claims") and
on a 7-point scale, indicating that the stimulus materials succeeded in creating a
moderately credible image for the fictitious corporation. Table
2 summarizes the control
group means and standard deviations for all scale items.
Control Group Means for All Scale Items
Mean Std. Deviation
Amateur/Expert 5.72 1.21
Inexperienced/Experienced 5.70 1.18
Unqualified/Qualified 5.68 1.02
Passive/Active 5.65 0.88
Poorly Trained/Well Trained 5.58 1.08
Unconcerned/Concerned About Community 5.58 1.13
Bad/Good 5.40 1.05
Unjust/Just 5.11 1.06
Don't Like/Like 5.05 1.19
Dangerous/Safe 5.04 1.19
Poor/High Quality Goods 5.02 1.13
Unbelievable/Believable 5.02 1.40
Dishonest/Honest 4.98 1.17
Not Trustworthy/Trustworthy 4.93 1.21
Chief Concerns Profit/Well-being 4.89 1.40
Doesn't/Does Make Truthful Claims 4.77 1.36
Factor Analysis and Scale Creation
All scale items were subjected to principal components factor analysis, employing
a VARIMAX rotation with Kaiser normalization. Factor analysis of the sixteen items
yielded two distinct factors. Eigenvalues for the factor analysis appear in Table 3a; the
rotated component matrix scores are contained in Table 3b. The Kaiser-Meyer-Olkin
measure of sampling adequacy was .957
, indicating a high degree of intercorrelations
among the data. This was expected, since the instrument created for this study was
derived from a number of other scales that had already exhibited high reliability in
measuring credibility. According to Hair, Anderson, Tatham, and Black,
of sampling adequacy] measure can be interpreted with the following guidelines: .80 or
above, meritorious; .70 or above, middling; .60 or above, mediocre; .50 or above,
miserable; and below .50, unacceptable" (p. 99). The results indicate that factor analysis
is appropriate for the data set.
Table 3a. Factor Analysis: Total Variance Explained
Component Total % of Variance
1 9.569 59.804
2 1.119 6.995
3 .768 4.799
4 .640 4.000
5 .554 3.464
6 .489 3.058
7 .453 2.829
8 .397 2.478
9 .333 2.081
10 .330 2.065
11 .309 1.933
12 .250 1.562
13 .231 1.444
14 .199 1.245
Table 3b. Rotated Component Matrix Scores for Corporate Credibility Scale Items
Factor 1 Factor 2
Not Trustworthy/Trustworthy .804 .376
Bad/Good .773 .365
Dishonest/Honest .768 .339
Don't Like/Like .766 .348
Doesn't/Does Make Truthful Claims .745 .358
Unjust/Just .717 .486
Chief Concerns Profit/Well-being .708 .201
Unbelievable/Believable .696 .459
Dangerous/Safe .649 .297
Unconcerned/Concerned .637 .465
Amateur/Expert .316 .812
Unqualified/Qualified .411 .766
Inexperienced/Experienced .295 .764
Poorly Trained/Well Trained .297 .731
Poor/High Quality Goods .377 .647
Passive/Active .346 .640
The two factors that emerged from the dataset were anticipated by the theoretical
framework of the study (Newell & Goldsmith, 2001; Ohanian, 1990). Factor 1 was
characterized as "trustworthiness." Bi-polar scales items that loaded high in
trustworthiness included "not trustworthy/trustworthy" (.804),
"doesn't make truthful claims/makes
"unjust/just" (.717), and "is concerned about making profits/is
concerned about the public interest"
was characterized as "expertise.
" Bi-polar scale items that loaded high in
expertise included "amateur/expert" (.812),
"inexperienced/experienced" (.764), and "poorly trained staff well trained staff (.731).
To validate the results of the factor analysis, a second factor analysis of a random
sample of 50% of the original sample was performed (H-air et al 1 998R This confirmed
Kaiser-Meyer-Olkin measure of sampling adequacy of .947
. Eignenvalues for the
confirmatory analysis appear in Table 4a; the rotated component matrix scores for the
confirmatory factor analysis are contained in Table 4b.
Table 4a. Confirmatory Factor Analysis: Total Variance Explained
Component Total % of Variance
1 9.753 60.957
2 1.291 8.071
3 .799 4.993
4 .587 3.669
5 .524 3.277
6 .456 2.849
7 .409 2.557
8 .350 2.188
9 .346 2.164
10 .275 1.720
11 .248 1.548
12 .230 1.435
13 .221 1.384
14 .207 1.294
15 .159 .995
16 .144 .899
Table 4b. Rotated Component Matrix Scores for Confirmatory Factor Analysis of
Corporate Credibility Scales
Factor 1 Factor 2
Not Trustworthy/Trustworthy .824 .324
Doesn't/Does Make Truthful Claims .807 .224
Dishonest/Honest .799 .309
Chief Concerns Profit/Well-being .760 .183
Good/Bad .756 .398
Don't Like/Like .749 .397
Unjust/Just .732 .476
Unbelievable/Believable .714 .407
Unconcerned/Concerned .682 .439
Dangerous/Safe .660 .428
Amateur/Expert .254 .838
Inexperienced/Experienced .217 .821
Unqualified/Qualified .428 .776
Poorly Trained/Well Trained Personnel .334 .763
The results of the factor analysis were used to create two summated scales. Benefits
of summated scales include their ability to represent multiple dimensions of a construct
with one measure and their generalizeability (Hair et al., 1998).
While many different
methods of creating summated scales have been suggested, the most straightforward
method is to add the responses of variables that loaded highest on a factor and using the
total as a replacement variable with which to measure the desired construct; this was the
method employed in this dissertation. Scale items with loadings greater than
.7 on each
factor were therefore included in the scales; no scale items had loadings greater than
both factors. Each subject's
responses to each of the scores included in either factor were
added together to create two variables: trustworthiness and expertise. A preliminary
analysis of the Trustworthiness scale indicated that reliability would increase with one
"is concerned about making profits/is concerned about the public interest,"
removed. The item was removed from the summated scale.
The reliability coefficient for the 6-item Trustworthiness scale was .9314
a). A split-half reliability test yielded an alpha of .8661 for the 3 items in
part one, and an alpha of .8821 for the 3 items in part two.
The reliability coefficient for the 4-item Expertise scale was .8836 (Cronbach'
A split-half reliability test yielded an alpha of .7378 for the 2 items in part one, and an
alpha of .8487 for the 2 items in part two.
Reliability of both summated scales was deemed good. The summated scales
derived for the factor analyses were deemed to have both content and construct validity,
rP'nlilw mPi^aciurina turn actahfliaA~o/l aolnnntoi nf rnrnnrot nrawk/lh'.t, vni+x1rTlhfO OrrintyklAl r
" and "expertise." Constituent measures of the two summated scales are
listed in Table
Constituent Measures of Summated Scales
Constituent items Bad/Good Inexperienced/Experienced
Don't Like/Like Poorly Trained/Well
Dishonest/Honest Trained Personnel
Doesn't/Does Make Unqualified/Qualified
F 28.1313 12.0918
Prob. .0000 .0000
Cronbach's a .9314 .8836
Scale mean 26.5423 19.7757
To test for discriminant validity, the trust and expert sums were subject to bivariate
correlation. The discriminant validity test was inconclusive, indicating that the measures
were positively, but not perfectly, correlated, r (N=437)=.728, p=.01 (two-tailed). This
correlation was expected, since trustworthiness and expertise are known to be correlated
aspects of corporate credibility (Newell & Goldsmith, 2001).
Because the experiment was meant to test differences between Spanish and
American subjects, all subjects who listed their nationality as "other" were eliminated
from the subsequent analysis (N=420). The remaining subjects consisted of 220
Spaniards (42% males and 58% females) and 200 Americans (37% males and 53%
One-way ANOVA between the control group and all experimental subjects
indicated that experimental subjects judged MultiChem both less trustworthy, F (1,
= 15.881, p
.000, and less expert, F (1,
= 27.144, p
.000, than the members of
the control group.
indicates an effect of the experimental treatments.
Another preliminary one-way ANOVA indicated no main effect of gender on both
trustworthiness, F (1,
and expertise, F (1
Overall, males rated MultiChem as no more trustworthy (n=220, M=26.16, SD= 7.74) or
more expert (n=220, M=19.77
, SD= 4.63) than females (n=200,
, SD= 4.68, respectively).
Therefore, gender was not included as a factor
in the primary analysis for either dependent variable1.
Both trustworthiness and expertise were analyzed using a 2 (country-Spain or
United States) x 3 (excuse--clarity, obligation, control) factorial ANOVA x 2 (format-
advertisement or newspaper article).
Table 6 illustrates the results of the factorial
ANOVA on trustworthiness, including significant interactions. Table 7 illustrates the
results of the factorial ANOVA on expertise.
Effects of Culture
Research question 1 asked about the relationship between Hofstede's
values and the credibility of corporate excuses.
With regard to trustworthiness, Tables 6
and 8 show a main effect on nationality, with Spanish subjects reporting lower
= 21.882) for MultiChem than American subjects (M
I Suhsenuent analysis nf variance on hath the trnstwnrthinep. and exnertise variahlen. incr.ldin gPender
= 182.896, p=.001. Nationality had no significant interactions with any other
It appears that culture does impact the perceived trustworthiness of a corporation in
the wake of an adverse incident. In Hofstede's
terms, it appears that members of a low
masculinity, low individuality, high uncertainty avoidance, high power distance culture,
such as Spain's, perceive a corporation as less trustworthy in the wake of an adverse
incident than members of a culture that is high in masculinity, high in individuality, low
in uncertainty avoidance, and high power distance, such as the United States.
Table 6. Factorial Analysis of Variance for Trustworthiness
.360 (Adjusted R Squared
. Factorial Analysis of Variance for Expertise
R Snlnared =
214 (Adiiiqted R Snmiaretd =
With regard to expertise, the ANOVA analysis (Tables
effect of nationality (F (1, 365)
and 8) indicates a main
= 84.831, p = 000), without any significant interactions.
Thus it appears that, regardless of excuse or delivery format, culture impacts the
perceived expertise of a corporation in the wake of an adverse incident. Table 8 indicates
that the respondents in Spain (a low masculinity, low individuality, high uncertainty
avoidance, high power distance culture) perceive a corporation as less expert (M =
17.452) than do respondents in the United States (M
= 21.481), which is rated as a high
masculinity, high individuality, low uncertainty avoidance, high power distance society.
Table 8. Corporate Credibility by Culture
(n = 220) (n = 200)
Trustworthiness* 21.882 30.626
Expertise* 17.452 21.481
* Analysis of variance indicates that the difference between Spain and U.S.
Effects of Excuse Type
The second research question asked which kinds of corporate excuses were most
credible with audiences. As illustrated in Table 6, there is no main effect of excuse type
on trustworthiness (F (2, 365)
= 1.421,p =
.243). However one interaction, between
excuse and format, approached significance (F (2, 365)
Controlling for nationality, subjects in the control excuse condition who read the
excuse in an advertising format found MultiChem more trustworthy (M= 27.570, SD =
.814) than those who saw the same excuse in a news format (M
ihhi;P tc in tth 1nlt d hvnincs rnrmretinn ,l-in rHrl th'a puvrmnc in an *rlT aicmc frrat
who saw the same excuse in a news format (M
.718). There was also little
difference between the advertising and news conditions for subjects in the obligation
excuse condition (M= 25.721, SD =
, and (M= 25.636, SD
.761, respectively). The
interaction is illustrated in Figure
Figure 2. Estimated Marginal Means of Trustworthiness by Excuse Condition
In a scaled-down model, controlling only for nationality, ANOVA indicates a main
effect of excuse on trustworthiness (F (2, 365)
= 3.115, p = .046). However, ANOVA
indicates no main effect of medium, controlling for nationality only (F (2, 365)
p = .985). Nationality achieves significance at the p
< .001 level in both of the smaller
Table 7 indicates a main effect of excuse type on expertise (F (2, 365)
p = .048). Controlling for nation, gender and medium, subjects' judgments of
more expert than those exposed to prescription clarity or personal obligation excuses.
Table 9 illustrates the means for excuse type on trustworthiness and expertise.
Further one-way ANOVA of the 16 original scale items by excuse type showed
only two significant differences in overall means: "inexperienced/experienced" (F (2,
= 3.127, p = .045) and "amateur/expert" (F (2, 364)
= 3.450,p = .033). Subsequent
t-tests revealed that subjects exposed to the personal control excuse judged MultiChem
more experienced, t (240)
= 2.371,p=.019 (two-tailed), and less amateur, t (240)
p=.009 (two-tailed), than subjects exposed to the prescription clarity excuse.
Table 9. Types of Excuses and Corporate Credibility
Prescription Clarity 26.990 19.043a
Organizational Obligation 25.679 19.128b
Organizational Control 26.094 20.229a,b
a, b items with same subscript are significantly different from other items in that column,
according to Tukey's HSD test
Effects of Delivery Format
The third research question asked if corporate excuses offered in advertising are
less credible than the same excuses offered within the context of a news article. One-way
ANOVA indicated no main effect of delivery format on either trustworthiness (F (1, 364)
= 1.666, p =
.198) or expertise (F (1, 364) = .039,p =
.843). The results of the ANOVA
analysis fail to indicate that overall advertising is less credible than a news story for
delivering a corporate advocacy message.
The fourth research question asked about the relationship between Hofstede's
cultural values and the relative credibility of excuses offered in advertisinQ and within the
excuse type, there is no evidence of a main effect of delivery format on either
trustworthiness (F (1, 365)
.216) or expertise (F (1, 365)
.000, p = .990) of
corporations based on advocacy messages in the wake of an adverse incident. The data
indicate no relationship between Hofstede's
cultural variables and the relative credibility
of excuses offered in advertising or within the context of a newspaper article.
Table 10. Effect of Excuse and Format on Corporate Credibility
News Story Advertisement
Prescription Clarity 27.307 26.673
Organizational Obligation 25.636 25.721
Organizational Control 24.617 27.570
Prescription Clarity 19.883 18.703
Organizational Obligation 18.825 19.431
Organizational Control 20.200 20.527
Subsequent t-tests on the 16 original scale items by delivery medium reveal only
two significant differences in format means: subjects in the advertising condition judged
MultiChem more likeable, t (378) = 2.322, p=.021 (two-tailed), and more active,
t (378) = 2.699, p=.007 (two-tailed), than subjects exposed to the news story condition.
Table 11. Corporate Credibility of Advertising Condition Subjects by Culture
(n = 90) (n = 81)
Trustworthiness* 22.73 30.57
Expertise* 17.67 21.27
* Analysis of variance indicates that the difference between Spain and U.S. is significant
Credibility of Advertising in Different Cultures
MAS cultures, such as Spain. This hypothesis was supported. One-way ANOVA of those
subjects in the advertising condition (n=171) indicates that, as expected, U.S. subjects
found excuses offered in advertising both more trustworthy (F (1, 171)
= 63.745, p =
.000) and more expert (F (1, 171)
= .000) than Spanish subjects.
Honoring of Excuses
Hypothesis two suggested there would be less difference in credibility scores after
an adverse incident in low MAS cultures, such as Spain, than in high MAS cultures, such
as the United States. To test this hypothesis, subjects were recorded as either "control
group" or "experimental group.
A 2 (condition-control or excuse) x 2 (nation-Spain
or U.S.) ANOVA was used to determine whether or not there was a significant
interaction between the two national groups before the adverse incident (the control
group) and after the excuse was offered (the experimental group), indicating a difference
in excuse honoring.
- -. .-
Hypothesis two was not supported. Although the data indicate a main effect of both
nation (F (1,
= 82.988, p
= .000) and condition (F (1,
= 17.201,p = .000) on
trustworthiness, there is no indication of an interaction between condition and nation on
trustworthiness (F (1,
420) = .819,p =
Likewise, while the data indicate a main effect of both nation (F (1,
420) = 46,385,
p = .000) and condition (F (1,
.000) on expertise, there is no indication
of an interaction between condition and nation on expertise (F (1,
As Figures 3 and 4 indicate, the lines between the two nations are close to parallel,
signifying that members of the low MAS Spanish culture did not honor excuses better
than members of the high MAS American culture.
Figure 4. Estimated marginal means of control and experimental groups on expertise by
DISCUSSION AND CONCLUSIONS
The MultiChem scenario used in this dissertation was designed to mirror the
corporate events that are routinely covered in local, national and even international
media. MultiChem, a large but unassuming multinational chemical corporation, used
corporate communications-a Web site, press releases, and local events-to create a
moderately positive image for itself. It was regarded as both trustworthy and expert in its
field of endeavor, due to its environmentally conscious policies, its philanthropic efforts,
and its many awards for scientific accomplishment.
However, despite the company's
best efforts, there was an explosion in an
industrial facility. Some employees were injured, and many people were inconvenienced.
The media immediately reported the events. In impression management terms,
MultiChem faced a difficult predicament.
In the wake of the disaster, the company investigated the explosion, found out why
it happened, made sure it would never happen again, and then responded to the
predicament by offering an account to the public, in an attempt to recover a portion of its
corporate identity-that of a trustworthy, expert member of the community-that was
denied it by the adverse incident.
The most important question for MultiChem is, what is the likely response of
incident, the extent to which its credibility is diminished hinges both on the cultural
values of the individuals being asked to accept the excuse, the type of excuse the
corporation offers, and to a lesser extent, the media format in which the excuse is
The Impact of Culture
Findings indicate that the fictitious company began with lower credibility in Spain
than in the United States. This finding was not surprising. Taylor's (2000) study of
reactions to The Coca-Cola Corporation after a tainting scare
indicated that the combination of high PDI and high UAI translated to a distrust of
authority figures in the wake of a potentially harmful incident.
It is easy to imagine that a powerful, multinational corporation, especially in the
chemical industry, where there is large opportunity for harm, could be an object of
distrust after a damaging incident. However, this experiment, unlike Taylor's
approach, extends this information. It appears that corporations may not need to have an
adverse incident to warrant distrust in certain cultural contexts-they actually come into
any given situation with a presumed lack of credibility, relative to other cultural contexts,
such as that of the United States.
The findings reveal that there was an overall significant effect of knowledge of the
adverse incident on credibility, even after an excuse was made, in both Spanish and U.S.
samples. Both trustworthiness scores and expertise scores were diminished among
members of the experimental conditions, relative to those in the control group. Any
knnwledce of an adverse incident seems to lessen credibility, no matter what excuse is
achievement" (p. 21). This study indicates the types of excuses may have a large impact
on organizational reputation and achievement as well.
This study found that nationality was the most significant predictor of evaluations
of trustworthiness and expertise. The Spanish subjects found MultiChem less credible
than their American counterparts, controlling for both the type of excuse given for the
explosion and the media format in which the excuses were delivered. Although some
scholars, such as Schimmack et al. (2002), have indicated gender effects mediating
culture, this study offered no evidence of any main effect of gender on corporate
credibility in the wake of an adverse incident.
In its support of hypothesis one, the study indicated that excuses made in
advertising would lead to greater credibility in high MAS cultures than in low MAS
cultures. This substantiates DeMooij's (1998b) claims. However, it must be noted that
this finding is tempered by the fact that Americans found MultiChem more credible than
Spaniards in every condition.
While it is confirmatory, it is certainly not conclusive.
The second hypothesis, that excuses would be better honored in low MAS Spain
than in the high MAS United States, was not supported. There was no interaction
between the control or experimental conditions and nationality on either trustworthiness
or expertise, indicating that MultiChem did not experience a better recovery in either
cultural setting after it offered its excuse. In fact, these findings indicate a remarkably
similar pattern of declining credibility in the wake of an adverse incident and excuse that
appears more dependent on a corporations' baseline credibility going into a corporate
a.. nt a, +r1 4*n +la r, l rm 14 ,.. 1 rnt rwn n2f l+ AZa *t f nt +IZ A AnW4 n .. ..A ,4 j
The Impact of Excuse Type
The findings indicate that excuse type affected the two dimensions of corporate
credibility-trustworthiness and expertise-differently. There was no main effect of
excuse type on trustworthiness, controlling for nationality and medium. However, there
was a near-significant interaction of excuse and medium, controlling for nationality. In
the control condition, in which subjects were told that the explosion was beyond the
control because of a faulty valve, those subjects who received the excuse
in the form of an advertisement found the corporation more trustworthy than those who
received the excuse in the form of a newspaper article.
Since the findings indicate that subjects exposed to the advertising condition judged
MultiChem more "active"
than those exposed to the news story condition, it may be that
the proactive move of taking out an advertisement had the effect of making the
corporation less "helpless,"
and therefore more solid and stable, despite its lack of control
over a very dangerous situation. Conversely, passively waiting for news coverage of the
event may have exacerbated the perception that the corporation was not in control of the
situation, and should not be viewed as trustworthy.
The interaction also indicates that subjects exposed to the clarity excuse (in which
the corporation had no way of knowing that the explosion would occur because no
scientific literature had predicted it) in a news story format found the company more
trustworthy than those exposed to the control excuse in the news story format, although
this difference only approached significance. This may be an indication of the media
gatekeener effect nosited by Hearit (1999). Because the clarity excuse was dependent on
which the corporation blamed a faulty valve made by another company, may not have
benefited as much from the traditional media gatekeepers, appearing as a case of"he-
said-she-said" reminiscent of the Ford/Firestone tire public relations disaster, in which
the companies used the press to blame each other for malfunctioning tires on some car
The study did find a main effect of excuse on expertise. Those in the organizational
control condition found MultiChem more expert than those in the prescription clarity and
organizational obligation conditions, controlling for nationality and delivery format.
It seems logical that having built an expert corporate identity, there was more
damage to MultiChem's perceived expertise from an excuse claiming a lack of
knowledge about a scientific concept than from an excuse claiming that the explosion
was beyond the control of the corporation. Schlenker et al. (2001) suggest that "excuse
makers run the risk of appearing ineffectual when their excuses deny personal control
over events or they claim to be unknowledgeable about or indifferent to relevant goals,
rules, and scripts, especially under circumstances where such control, knowledge, and
commitment are normally expected" (p. 23). In MultiChem's case, it appears that the
result of the corporation's
excuse claiming lack of knowledge was reduced credibility, in
the form of diminished expertise by those exposed to that excuse, relative to those
exposed to the other two excuse types.
What remains unclear, however, is why scapegoating another corporation (the
organizational control excuse) was more effective than scapegoating the government (the
aa*-^rt-nn 4l atnrl r InIt e-(rn+ an n rn l n na t In F t fl nn. n- +i, a-a nrn n Ia i nl If 0 flr rrfl^rfltiny T
a belief that the government's scientists shouldn't make mistakes. However, it may be
that there is simply a lack of credibility in doing government work.
The Lack of Impact of Delivery Format
Perhaps the most surprising finding of this study was a lack of superiority of the
news story condition for creating a credible corporate image in the wake of an adverse
event. Controlling for country and excuse type, there was no main effect of delivery
format on either trustworthiness or expertise. Of the two items in the original scale that
evidenced a significant difference between delivery formats, the study found that
advertising created a more credible corporate image than the context of a news story.
Especially interesting was that subjects exposed to the advertising condition found
MultiChem significantly more likeable than those in the news condition. This may be due
to a number of external criteria, such as the pleasant layout of the advertisements, which
contained both graphics and an easy-to-read type size and font. However, given the
sample-college aged students of "the most mediated generation"-the subjects may
have felt more positive attitudes toward a corporation spending the time and money to put
together an advertisement to speak directly to an audience. Conversely, the advertising
may have "associative" credibility, since it has already been included in a newspaper-
that is, students may have assumed that the advertisement was "true,"
simply because a
newspaper had chosen to run it.
The main implications of the study are threefold. First, culture affects the perceived
credibility of an organization. Multinational corporations-or more specifically the
different cultural settings. Appropriate assessment of baseline credibility in each cultural
setting in which the corporation has a significant presence appears to be a crucial public
relations function. Any adverse event is likely to diminish hard-won credibility.
This study indicates that the level of corporate credibility at the end of a corporate
crisis is directly related to its level going into the crisis. Therefore, proactive public
relations aimed at improving corporate credibility before an adverse incident, especially
in cultural settings where corporate credibility is known to be low, may serve as a form of
"damage control" in the wake of a corporate crisis. Hofstede's cultural values may
ultimately serve as one indication of likely baseline corporate credibility, but far more
research is necessary. At present, improving environmental monitoring capabilities in
countries with significant, relevant publics appears to be the only recourse of
multinational corporations wishing to better understand their position in host countries.
Second, some excuses offered in response to an adverse incident may benefit
credibility more than others. Judgments of expertise were lower when a corporation
offered a prescription clarity or organizational obligation excuse than when it offered
organizational control excuse. For corporations'
whose profitability rests on their
perceived expertise, such as those in the pharmaceutical, high-tech, or medical industries,
the effects of an excuse based on a lack of prescription clarity may prove especially
costly. Although it remains to be tested, there may be times when offering an apology
may serve a corporation better than trying to excuse the incident. At the very least,
corporations that have built their reputations upon being "expert"
in their fields should
4~~~ 4.4 'S 4 4 U 4 6 aC
Third, advertising may be a more effective delivery format for advocacy messages
than generally believed.
While public relations professionals may shy away from using
advertisements because of they believe they lack credibility, this study did not find that to
be the case. In fact, there may even be some advantages to using advertising, since the
study showed that subjects exposed to excuses in advertising judged MultiChem more
likeable and more active than those exposed to the news story condition. In particular,
corporations facing charges of "inactivity" or a "lack of community involvement" might
consider using advertising to give them a more active and likeable image.
While it may be suggested that the results in the advertising condition benefit from
the overall credibility of the news medium in which the ad was supposedly placed, the
implication for corporations is the same. At the very least, with no measurable difference
between the two formats, it may be advantageous to use advertising in situations where
having complete control of the message is essential to advancing a corporation's
Limitations and Future Research
As in all studies, this one has a number of limitations, some of them quite important
to note. Additionally, it serves as a starting point for a number of other inquiries.
The first and most important limitation of the study stems from the methodology of
the study itself: experimentation.
While testing conditions may consistently produce
specific results, the external validity of the results is not proven. This experiment was
specifically designed to test the impact of only a few variables on corporate credibility,
hand, hundreds of additional external factors would impact perceived credibility in the
wake of an adverse incident.
Some external agents have been shown to impact corporate credibility. As Siomkos
and Shrivastava (1993) note
Consumers place much value on how observers and agencies external to the firm
react to a company's handling of a crisis. External agents, such as the media,
regulatory agencies, local and state officials, and civic leaders, provide information
and neutral comment to the public. Consumers are more likely to view a troubled
company favourably, if these social agents respond positively. (p. 72)
Although information about external agents was minimized by design in this study, it is
difficult to state with certainty a measure of generalizeability for this study.
As Arpan and Pompper (2003) point out, a major problem with experimental
research in the crisis communication area is that in a real world scenario, previous history
with the corporation would impact credibility. For example, a lifetime Craftsman tool and
Kenmore appliance user who has been happy for years with Sears'
service and guarantees
is likely to react differently to a job dispute in a local store than someone who has had a
bad experience with an employee at the same store. For one customer, the negative event
is an aberration; for the other, it may be confirmatory.
While the use of a fictitious
corporation afforded an opportunity to minimize extraneous variables, it also reduces the
generalizeability of the current study.
In the case of this study, history may prove a threat to generalizeability as well.
Tragically, the Atocha bombings occurred the day after the data collection for this
dissertation was completed in Madrid, Spain. The September 11, 2001 attacks on the
lAlTrl 't1 ('nnrar 0A l-h0 lonnn iann+^y .onnl+auI :n :nnunnnnrl nan naA n+4r n44 nnan nfl Ar n
world in general, and foreign organizations specifically, since the data were collected.
However, how that change would impact the current findings is unknown.
The MultiChem scenario itself is a limitation of the study. First, because it involves
a corporation from only one industry-wholesale chemicals-the results may not be
generalizeable to other industries, such as foods and household products. In fact,
knowledge of past incidents in the chemical industry may have reduced the credibility of
the corporation, especially in Spain, where baseline credibility was much lower than in
Additionally, the number of students to which the researcher had access in Spain
limited the scope of the research. The number of students did not allow for a cell testing
the effects of information about the incident alone, without an excuse. Although much
research in the area of social psychology has indicated that excuses minimize the impact
of a predicament, the current project does not indicate how much they help, only the
relative credibility of the corporation in their wake.
The small amount of time that the students were available to meet with the
researcher also limited the scope of the research. Although a number of professors both in
Spain and the United States generously allowed the researcher access to their classes, the
relatively short time that could be spared by the professors precluded a more extensive
demographic questionnaire, including information about pre-existing attitudes toward
corporations in general, and the chemical industry specifically.
The stimulus materials may also have impacted results. Subjects saw
a,,,^-4n- KltT nni nS-t 1 4lr a V Vr
n. .~ i-C .' n.y 'fl-n-fl- n.i4 nl an n A~ rn.+~ non', an+n ~ fba nnrnarofa 1170h no a~ in a
Advertisements appeared alone on a page, while in a normal print environment they
would have been surrounded by editorial copy. This may have impacted credibility
scores. Additionally, without the direct contrast of surrounding editorial, the
advertisement may have had greater prominence than it would have in the context of an
actual newspaper. Conversely, subjects could not explore the Web site as they normally
would; they simply saw an image of the homepage, and had to guess at the content they
would encounter if they followed the links. Its impact may have been decreased by the
nature of the stimulus materials.
Another limitation of the scenario is that it tested only one rhetorical posture:
explanative. According to Ware and Linkugel (1963), apologiae derive their persuasive
power from one primary posture. However, it is possible a different posture, combining
differentiation with denial instead of bolstering for example, may have changed the
results. In addition to excuses, the impact of apologies and justifications, the two other
major responses to predicaments (Schlenker, 1980), remain untested. It is possible that
the same information presented in combination with one of these strategies, as is often
the case in real world "accounting," would have had a different effect on the subjects.
As with any cross-cultural research, the inability to perfectly match samples must
be seen as a limitation of the study. Hofstede (2002) notes that matching of samples is
critical to successful cross-cultural research. However, perfectly matching a long set of
criteria, including but not limited to age, gender, education level, occupation, and socio-
economic status, can prove daunting in a cross-cultural context, especially when
international researchers are likely to have limited access to local nonulations. and
is hoped that by using students of similar age, educational level and socio-economic
status, such limitations are minimized in this study. However, the question of student
generalizeability to their cultural populations at large remains.
Finally, in its current form it is impossible to see whether Hofstede's
variables are impacting results, and if they are affecting results, which are most
important. For example, while prior research has shown a link between MAS and
advertising and tolerance of wrongdoing, a two-culture sample cannot by itself prove a
proposition. Hofstede's variables, while correlated, are thought to work independently.
To adequately test the four variables in Hofstede's
Value Survey Module, assigning their
mean scores to members of each culture, at least four culture groups are required for
"quadrangulation." At best, this study begins to document the impact of culture on
corporate credibility after an adverse incident. However, the root causes of observed
differences cannot be explained with certainty.
Future research should expand the cultural groups measured with the scale created
for this dissertation, to better evaluate the instrument itself, as well as the impact of
Hofstede's cultural variables on corporate credibility in the wake of an adverse event.
A research design including more than one type of organization or excuse posture
would increase the generalizeability of future studies.
While including additional
treatments involves recruiting additional subjects, which can be a difficult task in the
international arena, the increased information drawn from such studies would be of great
c1QP tn the 22rlAdlmir 2nd nrmfccinnal rnmmrnntiep alilP Tt wnnlrl heH e.neni1llv helnfill tn
collected from subjects, including information about pre-existing attitudes toward
Another avenue of research might test the relative effectiveness of various
corporate voices offering excuses on trustworthiness and expertise in different cultural
settings: local versus national or international spokespeople, high-level vs. low-level
corporate executives. This research would be useful to transnationals attempting to
choose the most appropriate spokesperson in a given country in the wake of an adverse
The experimental design could also be used to test the "country of origin"
determine whether and to what extent some countries are judged more harshly than others
in the wake of an adverse incident. Different levels of variables such as physical and
cultural proximity, past positive or negative history and extent of trade relations could be
tested, providing important information for transnational corporations.
Finally, future research should allow subjects to encounter stimulus materials in a
more natural way. Allowing subjects to use a computer to navigate a Web site for
information about a corporation, and then use the information they gathered to answer
questions about corporate credibility would better simulate the real world decision
processes that go into assessing responsibility for an adverse incident.
This dissertation used the theoretical framework of impression management,
specifically the triangle model of responsibility (Schlenker et al., 1994), to study the
impact of culture, excuses and media format on corporate credibility. In order to
created, translated and backtranslated. The experiment was conducted in two different
cultural settings. Finally, the subject data were analyzed.
The stated purpose of this dissertation was to extend the body of work already
completed in the social psychology field of impression management in two areas where
there has been relatively little study thus far: mass communication, specifically corporate
crisis communication, and cross-cultural comparative study. The data indicate that the
MultiChem scenario has accomplished that goal. First, the study further extends
impression management study into the mass communication field, indicating that
contrary to popular belief, impressions of corporate credibility are not substantially
reduced by presenting advocacy messages in advertising. Second, the study contributes to
the body of literature in the cross-cultural impression management field, confirming
empirically what has often been posited: that impressions of the same event vary based
the cultural values of the individuals interpreting the event.