AuditquAl:
dimensions of Audit quAlity
Angus Duff
University of Paisley
Published by
The Institute of Chartered Accountants of Scotland
CA House, 21 Haymarket Yards,
Edinburgh EH12 5BH
First Published 2004
The Institute of Chartered Accountants of Scotland
© 2004
ISBN 1 904574 02 5
This book is published for the Research Committee of
The Institute of Chartered Accountants of Scotland.
The views expressed in this report are those of the author
and do not necessarily represent the views of
the Council of the Institute or the Research Committee.
No responsibility for loss occasioned to any person acting
or refraining from action as a result of any material
in this publication can be accepted by the author or publisher.
All rights reserved. No part of this publication may be
reproduced, stored in a retrieval system, or transmitted, in
any form or by any means, electronic, mechanical, photocopy,
recording or otherwise, without prior permission of the publisher.
Printed and bound in Great Britain
by Antony Rowe Ltd
ReseaRch RepoRts
RefeReeing pRocess
The Research Committee applies a rigorous refereeing process to all
stages of its research reports. The refereeing process operates by sending
the initial research proposal to two independent referees (one academic
and one practitioner). The academic referee will either be a member
of the Panel listed below or an ad hoc referee. All proposals are also
reviewed by the Director of Research who remains in close contact
with the project. The two referees are consulted on the academic and
technical standard of the draft research report. In particular, they are
asked to comment on:
•
•
•
•
theacademicrigourofthedocument;
thevalidityoftheapproachtakeninthereport;
whetherthepresentationofthereportidentiiesthekeyissues
andbringsthesetotheattentionoftheintendedreader;and
whether the document will add to the knowledge and
understanding of the interested reader.
Professor J Bebbington
Professor V Beattie
Professor J Broadbent
Professor J Haslam
Professor J Holland
Professor W M McInnes
Professor C Mallin
Professor H Mellett
Professor M J Page
Professor C Roberts
Professor M J Sherer
ProfessorRTafler
Professor P Weetman
Professor R M S Wilson
University of Aberdeen
University of Glasgow
Royal Holloway, University of London
HeriotWatt University
University of Glasgow
University of Stirling
University of Birmingham
Cardiff University
University of Portsmouth
University of Aberdeen
University of Essex
CranieldUniversity
University of Strathclyde
University of Loughborough
The Research Committee is grateful to all those who participate in
the refereeing process.
Contents
Foreword .............................................................................. i
Acknowledgements ................................................................. iii
Executive Summary ................................................................ v
1. APPROACH TO THE PROJECT .......................................... 1
Aims of the research project .................................................. 5
2. THE
NATURE OF AUDIT QUALITY AND RELATED
LITERATURE ........................................................... 9
Audit environment ............................................................... 9
The demand for auditing services ..........................................13
Audit quality ........................................................................ 15
Audit expectations gap ......................................................... 24
Service quality ......................................................................26
Auditor attributes ................................................................. 37
Key points ............................................................................ 41
3. DEVELOPING
AN ALTERNATIVE APPROACH TO AUDIT QUALITY
USING OTHER LITERATURES ....................................... 43
Introduction ........................................................................ 43
Understanding client expectations ........................................ 44
Audit expectations gap ......................................................... 55
Key points ............................................................................ 58
4. ANALYSIS OF AUDITORS’ EXTERNAL COMMUNICATIONS ........
The importance of external communications to service
quality ..............................................................................
Annual reports and marketing literature ...............................
Key points ...........................................................................
61
61
62
70
CONTENTS
5. DEVELOPMENT OF AUDITQUAL .................................71
Stage 1: Item generation ....................................................... 72
Stage 2: Scale development .................................................. 73
Scale validation .................................................................... 74
Stage 3: Scale evaluation ....................................................... 80
Key points ............................................................................ 81
6. CONDUCT AND ANALYSIS OF SURVEY .............................. 83
Conduct of survey and demographics of respondents by
sample ............................................................................... 83
Responses to survey by sample group ................................. 90
Differences across the three sampled groups ....................... 105
Key points ......................................................................... 107
7. DISCUSSION AND IMPLICATIONS ................................... 109
Introduction ....................................................................... 109
Implications arising from the analysis of AUDITQUAL .......109
Implications for educators and those responsible for the
professional development of accountants ...........................111
Recommendations ............................................................. 112
Avenues for future research ................................................. 114
Conclusion .........................................................................115
REFERENCES ............................................................. 117
APPENDIX ONE ........................................................... 131
APPENDIX TWO ........................................................... 135
APPENDIX THREE ........................................................ 141
foreword
Following on from the scandals over Enron and WorldCom in the US
two years ago, and the more recent, developing scandal over Parmalat
in Europe, interest in audit quality from the world of business, the
auditing profession and indeed the general public has never been
higher. The audit profession has for many years now been subject
to increasing competition and a stagnating market for audit services
resulting in downward pressure on audit fees and margins. At the same
time, auditirmshavebeenaffectedbythetrendsofglobalisationin
business and the commercialisation of public practice. This is forcing
auditors to concentrate on improving the quality of service they offer
while maintaining or increasing the technical effectiveness of their
workinordertorestoreproitability.
This research report considers service quality and technical quality
as components of overall audit quality,and explores their importance to
audit irms as a means to help attract and retain high calibre staff as well
as to generate suficient income. It considers the audit expectations gap
and service quality and from this develops a model of audit quality based
on the existing research literature on the topic. Finally, it examines
the external communication and brochures provided for clients by the
country’s largest audit irms. The research report relects a combination
of documentary research and analysis together with interviews of a
sampleofauditors,inancedirectorsandfundmanagers.
Thereportidentiiesanumberofissuesthatwillbeofinterest
to auditors interested in service and technical quality issues. It
concludes by making a number of recommendations which could be
adoptedbyauditirmswishingtoimprovetheirlevelofauditquality.
Therecommendationscoversuchmattersasirm’sreputationsand
capabilities, learningfromclient’sperceptionsoftheirm’sstrengths
ii
FOREWORD
and weaknesses and implications for the recruitment, development
and retention of staff.
The Research Committee of The Institute of Chartered
Accountants of Scotland has been happy to sponsor this project and
is pleased that the research is becoming available at a time when the
subject matter is so topical. As such, the Research Committee hopes
that this project will be seen as a valuable contribution to current
thinking about audit quality.
Nigel Macdonald
Convener
Research Committee
April 2004
ACknowledgements
Thanks are due to the helpful comments of ICAS’s Director of Research,
Vivien Beattie,along with the comments of two anonymous reviewers.
The support of the ICAS research team is also acknowledged.
exeCutive summAry
Interest in audit quality is at an alltime high. The sudden collapse of
Enron in the United States in 2001, followed by a number of other
high-proileUScompaniesreportinginancialdificulties, createda
crisisofpublicconidenceconcerningthecorporategovernanceand
auditing of publiclyquoted companies in the US. The deep concern
about the quality of reported earnings led to the SarbanesOxley Act,
imposing new requirements on directors and audit committees and
effectively ending the selfregulation of the auditing profession in
theUS. Althoughthereexistsigniicantdifferencesintheregulatory
framework,corporate governance regime,and accounting and auditing
standardsbetweentheUKandtheUS, theinancialregulationofthe
publiclyquoted companies in the UK has come under scrutiny.
Within the UK, signiicant competition and slow growth in
the market for audit services over the past two decades have created
downward pressure on audit fees. At the same time the audit profession
has been affected by the globalisation of business, commercialisation of
practice and stakeholder dissatisfaction with the level of audit quality.
These issues are focusing auditors’ minds on improving the quality
of service they offer to their clients, while maintaining the technical
effectiveness of the audit. Despite a widespread recognition within
the audit industry of the need to maintain effective relationships with
clients,whilst recognising who the ‘real’customer is (ie the stakeholder),
relatively little academic research considers the relationship between
service quality and technical issues. The corporate strategy literature
indicates service quality can be a source of competitive advantage,
ie a sustainable means of providing clients (and stakeholders) with
what they want or need, better and more effectively. A premise of
this monograph is that service quality and technical quality are both
vi
EXECUTIVE SUMMARY
necessarycomponentsofauditquality, ifauditirmsaretogenerate
suficientfeeincometoattractandretainhighcalibrestaff.
This investigation was conducted in the period 2001 to 2002.
The report develops a broadbased model of audit quality on the
basis of prior literature concerning the audit expectations gap and
service quality. A second element of the investigation consisted of an
analysisoftheexternalcommunicationsofthe20largestauditirms
in the UK. External communications refers to the publicly available
annual reports, annual reviews, marketing literature and websites
of the accounting irms. Finally, a questionnaire, largely based on
prior literature, was developed. The questionnaire was administered
to three groups: partners in the 16 largest audit irms in the UK
(representing auditors); inance directors of UK listed companies
(representinginancialstatementpreparers); andfundmanagersinthe
UK (representing external users).
The audit quality model
The service quality and audit quality literature is synthesised to
developamodelofauditquality. Themodelidentiiesdiscrepancies
between client and stakeholder expectations as the key to identifying
where audit quality failure may occur. Specifically eight gaps
(discrepancies)areidentiied–seeigure1.
vii
EXECUTIVE SUMMARY
Figure 1:Audit quality gaps 1 to 8
Client/stakeholder
expectations
Gap 5
Perceived quality of
audit service
Audit
quality
External
communications
Gap 4
Gap 6
Gap 1
Gap 8
Legal and professional
standards
Gap 7
Stakeholder
expectations of
‘reasonableness’
Gap 3
Auditirmquality
standards
Gap 2
Auditor perceptions
of client
expectations
Gaps 1 to 4 represent quality discrepancies which occur within the
auditirm. Gap5representsthedifferencebetweenclientexpectations
of the audit quality they expect and the audit quality they perceive
they receive. Gaps 6 to 8 represent socalled audit expectations gaps,
viii
EXECUTIVE SUMMARY
discrepancies between stakeholder expectations and perceptions of the
quality of the audit.
Gaps 1 to 4 are speciic to service quality factors within the
accounting irm. Gap 1 represents the difference between client
expectations and the audit irm’s perception of client expectations.
Gap 2 is the discrepancy between client expectations and the audit
irm’s quality standards. The difference between the audit irm’s quality
standards and the observed quality of the audit creates gap 3. Gap 4
is the difference between promises (both explicit and implicit) made
aboutauditqualityintheirm’sexternalcommunicationsliterature
and the quality of the audit.
Gap 5 is the difference between a client’s and stakeholders’
perceptions of the quality of the audit they expect and their perceptions
of the audit quality delivered. Furthermore, gap 5 results in the four
gapscreatedwithintheaccountingirm.
Gaps 6 to 8 collectively represent the audit expectations gap.
The difference between what stakeholders expect the auditor to
accomplish and what they can reasonably be expected to accomplish
is labelled gap 6. Gap 7 represents a discrepancy between what an
auditor can reasonably be expected to attain and the responsibilities
the auditor has as laid down by law and professional promulgations.
The difference between auditors’ responsibilities enshrined in law, and
other regulations and actual auditor performance is labelled gap 8.
Gaps 7 and 8 represent ‘performance’ gaps, between what an auditor
can reasonably be expected to achieve and what they are perceived to
accomplish. By contrast, gap 6 is a reasonableness gap representing a
lack of stakeholder awareness of what an auditor can be expected to
achieve.
EXECUTIVE SUMMARY
ix
AUDITQUAL survey indings
The AUDITQUAL instrument consisted of 56 questions
(items) relating to audit irm factors, engagement partner factors
and audit team factors. These 56 items were reduced to create nine
distinct dimensions which were labelled Reputation, Capability,
Responsiveness, Independence, Nonaudit services, Empathy, Client
service, Expertise, and Experience. The statistical analyses indicated
these nine dimensions could be reduced to create two distinct factors
relating to technical quality and service quality. However,it is important
to note that important correlations exist between the nine dimensions,
meaning they cannot simply be interpreted as independent variables.
The AUDITQUAL model is shown in Figure 2.
x
EXECUTIVE SUMMARY
Figure 2: The AUDITQUAL model
Reputation
Status
Capability
Technical
Quality
Independence
Expertise
Knowledge
Experience
Responsiveness
Nonaudit
services
Service
Quality
Empathy
Understanding
Client service
The questionnaire also elicited information about the background
oftherespondents, includingtheirage, andgender, andsomespeciic
questionsrelatingtotheiremploymentasanauditor, inancedirector
or fund manager.
All three groups rated the technical quality dimensions of
reputation and capability the highest. The dimension considered the
least important by all three groups in the ability to provide nonaudit
services. Accountingirmpartners’ andinancedirectors’ responses
EXECUTIVE SUMMARY
xi
were broadly similar, with inance directors tending to rate service
qualityattributeshigherthanaccountingirmpartners. Accounting
irmpartnersgenerallyratedtechnicalqualitydimensionshigherthan
inancedirectors. Asexpected, fundmanagers, asstakeholders, rated
technical quality issues as being more important than accounting
irmpartnersorinancedirectors. Fundmanagersalsoratedservice
quality issues as being less important than the other two groups. The
latter inding probably relects the fact that stakeholder groups are
not involved in the audit process, and are unable to directly observe
service quality issues.
Recommendations
The research identiied a number of issues that are relevant to audit
practitioners interested in service and technical quality issues and the
potential dichotomy between serving stakeholder and client needs.
Basedontheindingsoftheempiricalstudyandthedevelopmentof
the theoretical model of audit quality several recommendations are
proposed for those wishing to improve the level of audit quality.
The nature of audit quality
•
Participants in the audit market should recognise audit quality is a multidimensional construct. Audit quality is made up of both service
quality issues and the need to deliver technical quality. Service
quality in the present investigation is described by four factors,
Responsiveness, Provision of Nonaudit services, Empathy and
Client Service. Technical quality, usually conceptualised as the
competence and the objectivity of the auditor, is described by
ive factors, Reputation, Capability, Independence, Expertise
and Experience. Although technical quality and service quality
xii
EXECUTIVE SUMMARY
are distinct constructs, important relationships exist between the
factors which describe the two facets of audit quality.
Client and stakeholder perceptions of audit quality
•
Auditors, finance directors (clients) and fund managers (stakeholders)
rated the two technical quality dimensions of Reputation and Capability
the highest. Consequently, auditirmsandaccountingeducators
should be mindful of the importance the firm’s reputation
and the capability of its staff and partners have in client and
stakeholder perceptions of audit quality. The service quality
literatureemphasisesthesigniicanceofidentifyingthoseaspects
of quality most important to client and stakeholder groups. The
loss of reputation threatens the additional remuneration (fees) that
auditors with a good commercial reputation (ie largerirmsize)
can charge for their audit services.
Systems to monitor and improve audit quality
•
Audit firms can improve audit quality by monitoring clients’ perception of
audit quality, and identifying the causes of audit quality shortfalls. The
AUDITQUAL instrument developed in this research provides
a means of measuring and understanding clients’ perceptions of
audit quality. The conceptual (‘extended gaps’) model shown in
Figure1providesawayoflinkingclient-perceiveddeiciencies
inauditqualitytodeicienciesthatoccurwithintheirm.
Implications for accounting educators
•
Firms need to attract high quality individuals with the necessary
technical and interpersonal skills. The study has emphasised that
if irms are to provide audit of the requisite quality there is a
EXECUTIVE SUMMARY
xiii
needforirmstorecruitstaffofthehighestcalibre. Recruitment
of the ‘highlyers’ and ‘good quality people’ emphasised by
respondents in audit irms,and highlighted in the irms’marketing
literature needs to identify two aspects of performance. First,
their technical skills. Second, auditors and trainee auditors
should possess the social skills necessary to maintain and develop
longterm relationships with audit clients. Although it may be
desirable for auditors to see themselves as ‘relationship managers’
selling a relatively complex mix of professional services, there is
a danger that an overemphasis of clientcenteredness could lead
to accommodating behaviours. Educators and staff development
partners need to be mindful that audit trainees are fully aware of
whotherealclientis(theend-userofinancialstatements).
ChApter one
ApproACh to the projeCt
Audit quality is an important issue for an accounting profession
facing criticism from regulators and stakeholder groups. The quality
of auditing has been the topic of ongoing recent debate between
these groups and accountants. A higher quality audit should reduce
stakeholder groups’ uncertainty associated with inancial statements
preparedbymanagers(Wallace, 1980). However, noonedeinitionof
audit quality exists. Traditionally, academic accountants have thought
of audit quality as:
... the market assessed joint probability that a given auditor will both
(a) discover a breach in the client’s accounting system, and (b) report
the breach. (DeAngelo, 1981, p.186)
Usingthisdeinition, therearetwoaspectstoauditquality. First,
the ability of the accounting irm to either: discover a problem in
the client’s accounting system, or make a correct judgement while in
possession of the relevant knowledge (ie their competence). Second,
the willingness of the auditor to disclose the problem (ie their
independence).
The relationship between auditor independence and audit quality
has also been the subject of much debate. Factors that are said to drive
auditor dependence include auditors holding shares or having some
other inancial interest in audit clients,key audit client personnel having
a close relationship with the auditor, auditors forming alliances with
audit clients, and the provision by auditors of nonaudit services to
audit clients (Windmöller, 2000).
2
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
An increased interest in audit quality has evolved from indicators
intheUnitedStates(US)auditmarketrelectingalackofconidence
in current quality levels. These include reports that litigation against
auditors has increased both in the frequency and severity of claims
(Cloyd, Frederickson & Hill, 1996; Krishnan & Krishnan, 1997)
and a mismatch of society’s expectations of auditors and auditors’
performance (the expectations gap).
AdificultywiththeDeAngelodeinitionprovidedearlieristhat
it does not fully capture the potentially conlicting roles of the various
audit market participants. Audit market participants can be grouped
intothreecategories:irst,externalstatementusers,second,auditclients,
and third, auditors (Sutton, 1993). Another feature of the audit market
isasigniicantsourceofdemandforauditserviceswhichcomesfrom
external users, who pay only indirectly for the audit services. The
audit client is potentially a forced participant in the market, required
by law and regulation to engage the services of an auditor to obtain
anopiniononitsinancialstatements.
Moreover, the DeAngelo deinition limits itself to a technical
deinitionofquality–anabilitytoidentifyandreportproblemsin
a client’s accounting system. More recent research has considered
the nature of the auditorclient relationship. For example, Craswell,
Francis & Taylor (1995) argue the auditor has an incentive to provide
the client with superior quality of service in the irst few years of
the auditorclient relationship. Other research investigates auditors’
willingness to reduce fees to win new clients (introductory pricing or
‘lowballing’). However, evidence of introductory pricing is mixed
with studies in the US reporting evidence of lowballing (Francis
&Simon, 1987; Simon&Francis, 1988; Turpen, 1990), butastudy
inAustraliaindinginitialpriceswhichexceedcost(Francis, 1984).
Assuming introductory pricing is used to some extent within auditing
services, the combination of reduced fees and extra attention leads to
client perceptions in the US they are receiving high levels of value for
APPROACH
TO THE
PROJECT
3
money,increasing overall satisfaction levels (Behn,Carcello,Hermanson
& Hermanson, 1997).
Outsidetheieldofaccounting, asigniicantbodyofknowledge
has developed to evaluate service quality. Interest in evaluating service
quality has been motivated by recognition of the signiicance of
service quality in business success. High levels of service are seen as a
means for an organisation or irm to achieve competitive advantage by
positioning itself more effectively in the market. Achieving sustainable
competitive advantage lies at the core of strategy development (Lynch,
2000). Sustainable competitive advantage is as important for accounting
irmsasforcorporations, astheyneedtocompetetoattractandretain
clients. Research demonstrates that high levels of customer service
can lead to customer loyalty, attraction of new customers, positive
wordofmouth, employee satisfaction and commitment, enhanced
corporate image, reduced costs and increased business performance
(Berry, Bennett, & Brown, 1989). Consequently, a commitment to
servicequalityhasclearbeneitsforaccountingirms.
Despite an extensive service quality literature, the quality of
professional services provided to business clients has received little
research attention (Dart & Freeman, 1994). A number of studies have
considered the determinants of audit quality in North America (Behn
et al, 1997; Carcello, Hermanson&McGrath, 1992; Mock&Samet,
1982; Schroeder, Solomon&Vickery, 1986; Sutton&Lampe, 1990;
Sutton, 1993). However, onlylimitedwork-Moizer(1998), Beattie
&Fearnley(1995)-hasconsideredUKauditirmsortheirclients.
The distinction between technical quality and service quality is an
important one for the present investigation. Technical quality focuses
on the constructs of competence (eg technical skills, rigour, judgement
and integrity) and independence (an ability to be objective and express
opinions independently of the auditee). Service quality addresses
issues pertinent to audit clients such as responsiveness to client needs,
providing addedvalue services beyond the statutory audit, having
4
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
systems in place to provide high levels of client service, and having
the ability to empathise with challenges facing the client. Clearly, in
certain situations, technical and service quality might be at odds. For
example, it may be problematic for an auditor to empathise with an
audit client, in danger of breaching loan covenants, who engages in
aggressive earnings management.
Despite an academic focus on technical aspects of audit quality,
professional journals have recently published articles explaining the
beneitsnon-technicalservicequalityandrelatedissuessuchasclient
managementmayhaveforaccountingirms. Researchconsidering
audit quality focuses on matters of competence and independence:
issues of direct relevance to external users of accounting information.
An emphasis on (nontechnical) service quality aspects is likely to
beneittheauditor(throughincreasedfees, levelsofbusiness, anda
more proitable relationship) and the audit client (through better service
levels). However, thebeneitsofservicequalitytoexternalusersof
accounting information are much more intangible.
For example, Andersen (1999) identiied that few accounting
irmshavesuficientunderstandingofthemselves, ortheirclients, to
improve their quality of service. Walker (2001) describes a framework
for improving client management based on four factors. These factors
include client service review meetings, added value provided to clients,
service penetration and the number of key staff within the client’s
organisation with whom the auditor has a relationship. Gurton (1999
p.44) explained how a client relationship programme can achieve
“positive customer retention” and that implementing a customer
managementrelationshipsystem“shouldbeairstpriorityafterthe
accounting system”.
APPROACH
TO THE
PROJECT
5
Aims of the research project
The focus of the project is audit quality and the principal aims
of the project are to:
(i) identify the determinants of audit quality, using a broad deinition
of audit quality, embracing technical and service aspects of
quality;
(ii) determine the robustness of this audit quality model using samples
ofauditors,inancialdirectorsandexternalusergroups;and
(iii) identify differences in perceptions of the determinants of audit
quality between auditors, inancial directors and external user
groups, to determine if an ‘audit (service) quality expectations
gap’ exists.
The project is novel for two reasons. First, it assesses the feasibility
of a model of audit quality which includes both elements of technical
and service quality. Second, it provides empirical evidence of the
quality of auditing provided by UK auditors.
Theprojectexaminesauditqualityfromtwoperspectives: irst, a
desiretoprovidesuperiorlevelsofservicetoclientmanagement; and
second, the need to undertake a thorough examination of the client’s
accounts, detect possible anomalies (competence) and be willing to
provide an objective opinion in relation to them (independence).
Some commentators suggest auditors have emphasised growth and
the ability of individuals to generate fee income, at the possible
expenseofauditquality. Asauditqualityisdificulttoobserve, and
the shortterm consequences of poor quality may be limited, critics
of the audit profession may feel a tension exists between client service
and the willingness to challenge a management team employing
aggressive earnings management: a tension which compromises the
quality of audit work and the informational content of the accounts.
6
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Therefore, the present investigation is a topical one, seeking to identify
the elements of audit quality, their relationship and how different
participants in the audit market perceive audit quality.
The investigation was undertaken in the period 20012002 and
consisted of three stages. Stage one develops a theoretical model
of audit quality, encompassing service quality, technical quality and
issuesrelatingtotheso-calledauditexpectationsgap. Asasigniicant
proportion of the audit relationship relates to the interaction between
theauditpartnerandtheinancedirector, itispossiblesomeauditors
may not completely distinguish between technical audit quality
issues and nontechnical service issues. Recent research in the UK
considering the relationship between audit engagement partners and
inancial directors (Beattie,Fearnley,& Brandt,2001),has identiied the
behaviour of some audit partners to attempt to accommodate clients’
views of certain accounting practices. The second stage consisted of an
analysis of the external communications of the 20 largest accounting
irmsintheUK. Finally, stagethreewasapostalsurveytoasampleof
auditpartners, inancialdirectorsandamixedgroupofexternalusers
ofinancialinformation.
The project makes a distinction between technical audit quality and
nontechnical service quality. To keep the project within a manageable
size, itwasdecidedtorestricttheresearchtothedeterminantsofaudit
quality, rather than investigate related issues such as client satisfaction,
or to investigate the relationship between audit quality and client
retention.
Chapter two,describes and discusses the external audit environment,
along with a review of the relevant prior literature. It also attempts to
synthesise work that has considered technical audit quality with the
service quality literature in marketing and quality management. Chapter
three describes a theoretical model of audit quality encompassing the
extant service quality and audit expectations literatures. Chapter
fourconsidersaccountingirms’ externalcommunications. External
APPROACH
TO THE
PROJECT
7
communications such as annual reports and marketing literature
inluence client and stakeholder expectations of audit quality. In chapter
ive the development of AUDITQUAL – the research instrument
developed for the purposes of the study – is described. Chapter
sixdescribestheresponsesofauditors, inancialdirectorsandother
external users groups to the AUDITQUAL instrument. Finally,chapter
seven draws conclusions and makes recommendations for accounting
irmstoconsiderwhenassessingauditqualityanddevelopingwaysof
improving client service levels. Implications for educators and those
responsible for the continuing professional development of practising
accountants are also considered.
ChApter two
the nAture of Audit quAlity And relAted
literAture
This chapter considers the nature of the external audit environment and
reviews the work of other researchers considering audit quality, service
quality and related topics. The ideas and evidence from this body of
work is used to construct the AUDITQUAL model, the development
of which is described in chapter six. This chapter consists of six
sections. These consider the audit environment, demand for auditing
services, audit quality, the audit expectations gap, service quality and
the SERVQUAL model, and personal attributes of the auditor.
Audit environment
The audit environment is characterised by intense competition
(Martinov & Roebuck, 1998), pressure on fees and slow growth (Behn
et al, 1997). Although some commentators describe the audit market
as being characterised by oligopolistic competition, auditors claim the
profession has experienced increased competition as the industry has
shifted from a growth stage to a mature stage (Elliott & Pallais, 1997).
Furthermore, despite the 1989 Big 8 mergers, empirical evidence
indicates that audit fees remained lat in the 1990s (Menon &Williams,
1991). Auditing has also been affected by the globalisation of business
(Fogarty, Heian & Knutson, 1991), the commercialisation of practice
(Hanlon, 1994), improvements in audit technology making it less
labour intensive (Elliott, 1998) and stakeholder pressure expressing
dissatisfaction with the level of audit quality (Higson, 1997).
10
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
In the UK,auditors have typically enjoyed long periods of stability
(tenure) with their auditors, with the average rotation period estimated
atbetween30and40years(Ridyard&DeBolle, 1992). Thisinding
is also true of other European countries,Australia and the US. Despite
the longterm (or ‘cosy’) relationship between auditors and the client’s
inance director which has been sometimes been described in the
inancial press and by politicians,the audit environment has been subject
toanumberofsigniicantchangeswithinthepasttenyears.
A commonly used framework in strategic management to
recogniseenvironmentalchangesinvolvestheidentiicationoffour
categories of key environmental inluences, which are in effect drivers
of change. This is sometimes known as a PEST analysis, indicating the
importance of political, economic, social and technological inluences
on organisations (or professions).
Important political and legal inluences in the UK include changes
in the supervision of auditors, with the creation of the Auditing
Practices Board (APB) in 1991, being described as a committee “with
a measure of independence from the accounting bodies” (Hatherly,
1999 p.51). The APB has been supplanted with a new APB in 2002,
under the umbrella of the Accountancy Foundation (AF), which seeks
to maintain and enhance the standards of work and of conduct of
accountants working in the UK. The AF is funded by the professional
accountancy bodies in the UK and appoints the board members of the
APB. The new APB comprises membership of 40% audit practitioners
and60%accountantswhoarenotpartnersinanyauditirm. TheAPB
and its predecessor have published a number of reports, notably the
‘McFarlane Report’ (APB, 1992), the ‘Audit Agenda’ (APB, 1994), the
‘Audit Research Agenda’ (APB, 1996a) and the ‘Auditors Code’ (APB,
1996b). Each of these publications has a change agenda, querying: the
independence of auditors from directors and the need for improved
communication between auditors and primary stakeholder groups
(the McFarlane Report); audit (professional) judgement (theAudit
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
11
Agenda); drivers of audit quality (theAudit ResearchAgenda) and
auditor conduct, including ethics (the Auditors Code). Other political
inluences include the (very real) threat of litigation; deregulated
inancialmarkets, theprivatisationofkeyindustries; andinally, the
development and expansion of the European Community along with
the single European currency.
Economic influences include the globalisation of supplier,
consumer and capital markets. Global businesses require service
providers which offer global services to match the client’s global
coverage. An allied trend is change in the nature of suppliers of capital.
Increasingly, individuals invest directly or indirectly in stock markets
aspartoftheirpensionplanning, ratherthanrelyondeined-beneit,
inal-salaryschemesaspartofanoccupationalpension. Globalisation
and the changing proile of investors lead to developments in the
market for audit.
Sociocultural changes also inluence audit markets. The past
decade has seen a growing need for accountability, or the responsibility
owed and provided by one party to another regarding some past or
future action.As organisations form new ventures,there is an increasing
need for accountability among contracting parties. This creates an
‘audit explosion’ (Power, 1994), where tensions between auditors and
shareholders create demands for regulation of auditors, greater access
to litigation, and in turn, demands for more accountability.
Arguably,the greatest technological change to the market for audit
is the development of the Internet. The Internet has accelerated the
complexity of systems installed for running business processes in all sorts
of business. In related developments, the Internet creates unlimited
networking possibilities between organisations. Such ecommerce
developments have created demand for new audit services considering
the security and controls on such systems (APB, 2001). A summary of
thesechangedriversforauditirmsisshowninFigure2.1.
12
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Figure 2.1 Environmental influences on providers of audit services
Economic factors
• Capitalsuppliers
• Globalisationof
business
Political/legal
• Deregulation
• Developmentof
Euroland
• Privatisation
• Threatoflitigation
Audit firms
Socio-cultural factors
• Accountability
• Ageingpopulation
• Attitudestowork
• Temporary
employment
Technological factors
• Internet
• Changingbusiness
processes
Client companies’ and audit firms’ responses
In response, auditors have introduced new audit methodologies,
focusing on a riskbased strategic audit. Such approaches and
methodologieshavebeenclassiiedasastrategicorinvestigativeaudit
(Lemon,Tatum & Turley, 2001). Auditing itself has been rebranded
asan‘assuranceservice’. TheAICPAdeineassuranceservicesas:
“independent professional services that improve the quality or context
of information for decision makers.”
(AICPA, 2001)
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
13
Assurance services are a broad set of services, which include
the traditional audit, designed to improve the quality of information
(AICPA, 1997). In the US, the changing nature of audit services has
been the subject of healthy debate. The Elliott Committee (AICPA,
1997) suggested that some of the ways that providers of assurance
services (‘assurers’) change, will change the ways irms do business
with their clients.These include:
•
•
•
•
•
the need for irms to adopt a customer (user) focus as assurance
services are intended to beneit decision makers by improving the
information used in their decision processes;
the need for irms to produce higher value-added information,
providing greater value to clients and users;
the need for irms to embrace IT, understand how IT is changing
business, how to use new IT developments in delivering
assurance services and understand the risks involved in electronic
commerce;
the ability of irms to respond to an environment of rapid change
and increasing complexity by maintaining up-to-date knowledge
and skills; and
the need for competition within the audit market, ensuring growth
in assurance services will depend less on regulation and more on
market forces, while providers of assurance services will need to
develop their marketing skills.
The demand for auditing services
Theoretically, the demand for audit services originates from a
need to facilitate contractual relations between the audit client and
stakeholder groups, eg shareholders, employees and creditors. The
preparation of accounts is typically controlled by the board of directors
14
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
of the company who are separate from the company’s stakeholders.
This separation of ownership and control creates a contractual conlict
between the parties leading to socalled agency costs. Agency theory
suggests the provision of auditing services reduces agency costs and
consequently contractual conlict.
Moreover, by buying auditing services, the informational value of
the accounts will be increased. Having purchased auditing services,
the accounts of the audit client signal credibility and demonstrate
willingness to perform to the various stakeholder groups (Jensen &
Meckling, 1976). Beyond this explanation of agency theory, four other
reasons have been suggested as creating demand for audit services:
(i) An auditor’s review of the accounts, as an independent exercise,
enhances stakeholders’perceptions of the reliability of the accounts.
The selection of credible auditors signals management’s quality
andintegrity(Dopuch&Simunic,1980;1982).
(ii) Auditing also provides an insurance dimension, whereby
shareholdersandcreditorsareindemniiedagainstinancialloss
by the auditor’s professional liability (Wallace, 1980).
(iii) For listed companies, the audit provides a form of comfort, where
“the auditor can provide board members with the assurances they
need to sleep soundly at night” (Beattie & Fearnley, 1998, p.44).
(iv) Auditirmscanprovideresourcesoverandbeyondthecompany
audit,such as technical services (accounting,tax,corporate inance
and due diligence) and advice on inancial reporting,and proactive
involvement relating to the future development of the company
(Beattie & Fearnley, 1998).
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
15
Audit quality
DeAngelo’s (1981) seminal work considers the quality of an
audit to be dependent on two factors. First, the auditor’s ability to
examine the accounts and identify errors or anomalies,ie their technical
competence, and second, their objectivity, ie their independence.
Auditing quality is the combined probability that the auditor will
detect and report on defects in accounts (Watts & Zimmerman, 1986
p.8). DeAngelo’sdeinitionofauditqualityisusefulforanalysisas
competence and independence are quite distinct constructs. However,
these two dimensions are not completely separate: for example, the
auditirmcoulddecidenottomakeanefforttouncoverproblems
(competence) which they have no desire to report on (independence).
Technical competence is relatively easy to conceptualise. However,
independenceismoreproblematic, being“dificulttoproveandeasy
to challenge” (Mednick, 1990 p.6). The concept of independence
described by DeAngelo (an auditor’s willingness to report on defects
in audited inancial statements) can be thought of as independence
in fact, which in itself is not directly observable “the state of mind
which has regard to all considerations relevant to the task in hand
but no other” (Federation of European Accounting Experts, FEE,
1996, p.24). Independence in appearance, by contrast is deined
by signals or other directly observable indications. Regulators pay
considerable importance to this dual deinition of independence.
When independence is mentioned in laws or professional rules, it is
‘independence in appearance’ that is being referred to.
The economics of audit quality
Auditing can be thought of as an economic exchange between a
supplier, the auditor, and direct users, the audit client who purchases
the services, and indirect users, inancial statement users, for whom the
16
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
auditedaccountsaredestined. Inanyeconomicexchange, suficient
resources need to be employed to prevent the parties from maximising
their own interests at the expense of the other parties to the exchange.
This is achieved essentially by a market mechanism. If the auditor is
deemed not to have complied with the contract of the exchange by
behaving opportunistically,they will attract criticism from others in the
marketplace. Put more simply, the auditor will suffer damage to their
commercial reputation. An auditor with a good commercial reputation
can charge higher fees than an auditor who contracts without such
a reputation. This additional remuneration available is termed by
economists as a ‘quasirent’. A quasi-rentisdeinedas“thedifference
between the remuneration for any productive resource in its current
use and the maximum remuneration which would be received for its
alternative use” (Arruñada, 1999a, p.21).
Audit quality is protected by both explicit and implicit contracting
safeguards. In an explicit contract, if the auditor has undertaken a
defective audit, this may result in litigation against the auditor. If
the work is found defective by third parties, then the auditor may be
required to pay compensation to those injured parties.
In an implicit contract, the auditor may be punished by existing
or potential clients withdrawing their trust in the auditor. This loss of
reputation will result in existing or potential clients either withdrawing
their business, or demanding more onerous terms to the engagement.
ThecollapseofthemajorirmAndersen, couldbeattributedtothe
breach of an implicit contract safeguard.
For the reputation (quasirent) incentive mechanism to operate, a
number of conditions have to exist (Arruñada, 1999a). First, the sale
price (audit fee) must exceed the marginal or opportunity cost, so the
quasirent is created. Second,the expected value of the quasirents must
besuficienttodiscouragenon-compliance. Third, customers(audit
clients and users) must be aware of the supplier’s (Auditor’s) incentives
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
17
so they may trust them. In summary, quasirents, manifested as the
irm’sreputation,playanimportantroleinprotectingauditquality.
Marketing investments and the expense created by the development
of commercial brands play a role in safeguarding commercial
relationships (Arruñada, 1999a p.32). In the UK (as in most EU
countries) the unsolicited offering of audit services is prohibited, and
marketing activities are restricted. The creation of a brand provides a
guarantee. The marketing investment provides an incentive to provide
a service of suficient quality which meets expectations. Consequently
the marketing investment provides an intangible asset whose value
rapidly erodes if the irm delivers defective work. Therefore,marketing
activity has the ability to act as a quality safeguard. Only limited work
has considered the relationship between irms’ commercial activity and
audit quality. Jeter & Erickson Shaw (1995) report that defective audits
are less common in US states which permit the unsolicited offering of
auditservices. Thepresentstudyexaminesirms’ marketingliterature
within this monograph in chapter four,in part to assess the expectations
irmscreateofthequalityoftheworktheydeliver.
Empirical work considering audit quality
Research considering audit quality largely ignores the extant
service quality literature. In the area of accounting research, the
quality of audit services has been viewed from different perspectives.
Schroeder et al, (1986) note:
Audit quality was not defined in the questionnaire nor was it
identified in any other component of the research materials. Since
there is no general understanding of what constitutes audit quality,
the provisions of such a definition could have serious demand effects.
(p.89)
18
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Sutton(1993)identiiestheabsenceofasingledeinitionofaudit
quality; thisisnotsurprisinggiventheoftenconlictingrolesofthe
various market participants. Audit market participants can be grouped
into three categories: external users, including current and potential
investors and creditors; the client, including senior management,
accountingstaffandtheauditcommittee; andtheauditors, including
the engagement audit team and audit irm management (Sutton,
1993).
Accounting researchers have examined the issue of audit quality
fromthreeapproaches. Theirstgroupofresearchershaveexamined
a number of issues related to audit quality by reference to pricing
differentials (Francis & Simon, 1987; Palmrose, 1986, 1989; Simon,
1985;Simon&Francis,1988;Simunic,1980;Turpen,1990).
A second group of scholars have considered audit differences
betweendifferenttypesofirmsandbetweenindividualirmsusing
various surrogate measures of quality performance. These include
variables suchas: litigation against audit irms (eg St Pierre & Anderson,
1984; Palmrose, 1987; Stice, 1991; Carcello & Palmrose, 1994), the
nature of audit opinions (eg DeAngelo, 1981; Hopwood, McKeown
&Mutchler, 1994; Carcello, Hermanson&Huss, 1995), andauditor
selection, auditorchangesandirmsize(eg Beattie&Fearnley, 1995;
Menon&Williams, 1991; Nichols&Smith, 1983; Simunic&Stein,
1996).
The third group of researchers have examined the issue of audit
quality from a behavioural perspective, (Behn et al, 1997; Carcelloet
al, 1992; Mock&Samet, 1982; Moizer, 1998; Schroederet al, 1986;
Sutton, 1993; Sutton&Lampe, 1990)typicallyidentifyingattributes
thatareperceivedbyinancialstatementpreparers, auditorsandusers
that are related to audit quality. The present investigation can be
categorised into this third behavioural perspective. An overview of
related (behavioural) research is shown in Table 2.1.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
19
Table 2.1: Results of behavioural audit quality research
Authors
Method
Results
Mock & Samet
(1982)
Survey of auditors in
US to develop 32item
questionnaire to evaluate
audit quality
Identiiediveauditquality
dimensions: planning,
administration, procedures,
evaluation and conduct.
Schroeder et al,
(1986)
Survey of audit
committee chairs and
auditors in US to
determine effect of 15
factors on audit quality
Audit team factors more
importantthanauditirm
factors
Sutton &
Lampe (1990)
Group exercise with
practising auditors in the
US to develop an audit
quality evaluation model
Model used 19 attributes
thatwereclassiied
into three categories:
planning,ieldworkand
administration
Carcello et al,
(1992)
Surveyofinancial
statement preparers,
auditors and users in
US to reduce 41 audit
quality items from
literature to 12 audit
quality factors
Most important factors
identiiedwere:experience
with the client, industry
expertise, responsiveness to
client needs, and adherence
to general standards of
GAAS.
Sutton (1993)
Nominal group
technique on
experienced auditors
to develop and validate
a set of audit quality
factors and measures
Identiied19quality
factors which could be
categorised into three
groups:planning,ieldwork
and administration
Beattie &
Fearnley
(1995)
Surveyofinance
directors of 210 listed
UK companies to
identify the importance
of 29 desirable auditor
characteristics
Identiiedivekey
characteristics focusing
on the importance of the
audit engagement partner’s
technical skills and people
skills
20
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Authors
Method
Results
Behn et al,
(1997)
Survey of controllers in
US to evaluate existing
auditor using 12item
questionnaire derived
from Carcello et al’s
(1992) work to identify
determinants of audit
client satisfaction
Responsiveness to
client needs, executive
involvement, effective and
ongoing interaction with
audit committee, conduct
ofieldwork,industry
expertise, and prior
experience of client all
positively associated with
client satisfaction
Moizer(1998)
Surveys of company
directors in 1987 and
1996 of the then Big 8
and Big 6 accounting
irms.Multivariate
regression analysis with
composite measure
of audit performance
used as dependent
variable, 11 independent
variables
Both 1987 and 1996
samples“wellsatisied
with the quality of service
provided by main audit
irm”. Olderinancial
directorsmoresatisied
withtheirauditirmthan
their younger colleagues
Warming
Rasmussen &
Jensen (2001)
Survey of shareholders,
inancialjournalists,
auditors and managing
directors perceptions of
audit quality
External users tend to
perceive audit quality
attributes as attributes that
inspireconidenceinthe
auditor;sixmainquality
dimensionsidentiied
(moralandethicalaspects);
four groups rate quality
dimensions differently
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
21
Table 2.1 lists and summarises those empirical studies that examine
attributes of audit quality. Mock & Samet (1982) drafted a list of
110 factors derived from Statements on Auditing Standards, Statements
on Quality Control Standards, peer review manuals, and irm quality
control standards. The list of factors was reviewed by a small group
of auditors, a process which reduced the 110 factors to a 32item
questionnaire, scored on a dichotomous scale (ie yes/no). Mock &
Samet’sinvestigationwastheirstattemptbyscholarsattemptingto
measure audit quality.
Schroeder, Solomon & Vickery (1986) extended Mock & Samet’s
(1982) investigation by focusing on the perceptions of various audit
quality factors by sampling Fortune 500 audit committee chairpersons
and(thethen)BigEightaccountingirmpartners. TheSchroederet
al, (1986) investigation required respondents to rate the importance
of each of 15 factors to overall audit quality. They report that audit
committee chairpersons rated audit team factors as more important
thanauditirmfactors(eg level of partner/manager attention to the
auditwasratedmorehighlythantheoverallreputationoftheirm).
Sutton & Lampe (1990) used a group exercise with practising
auditors to develop an evaluative model of audit quality. Their model
included 19 attributes of audit quality which they classiied into
threecategories: planning, ieldwork, andadministration. Eachofthe
19 factors included one measure (or item).
Carcello et al, (1992) created a questionnaire based on 41 audit
quality attributes based on a literature review. This questionnaire was
administered to the three groups of inancial statement preparers,
inancialstatementusersandauditpartnersintheUS. Exploratory
factor analysis (EFA) a statistical data reduction technique explained in
greaterdetailinchapterive–reducedthese41auditqualityattributes
to12auditqualityfactors. Theivemostimportantqualityfactorswere:
(i) team experience with client;(ii) industry expertise;(iii) an audit team
thatoperatestohighethicalstandards; (iv)apartnerknowledgeable
22
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
abouttheclient’sindustry; and(v)frequentcommunicationbetween
auditors and management. Overall, team characteristics were rated as
moreimportantthanirmcharacteristics.
Extending Sutton & Lampe’s work,Sutton (1993) applied nominal
group techniques to establish the key factors that relate to audit quality.
Nominal group techniques are based on the premise that people
involved in the daytoday operation of a process (ie auditing) can
provide insights into the weaknesses and problems associated with
that area. Using groups of experienced auditors from two accounting
irms, 66 factors were identiied by Sutton (1993) categorised into
planning, ieldwork and administration, which were reduced to
19factors. Sutton’s(1993)keyindingwastheparticularfocusofthe
client environment rather than accounting issues (eg client competence,
client preparedness, client rapport and client reaction).
The irst UK study to examine audit quality attributes was
undertakenbyBeattieandFearnley(1995). Theysurveyedinance
directors of 210 listed UK companies to identify the importance of
29 ‘desirable’ auditor characteristics. An exploratory factor analysis
– a statistical data reduction technique described in greater detail
withinchapterive–identiiedivemainfactors: (i)integrityofthe
irm; (ii)thetechnicalcompetenceoftheirm; (iii)thequalityofthe
workingrelationshipwithauditpartner; (iv)thereputationoftheirm;
and (v) the technical competence of the audit partner.
As a development of Carcello et al’s (1992) investigation, Behn et
al, (1997) considered the relationship between audit quality attributes
and client satisfaction. Controllers of Fortune 1000 companies in the
US were asked to rate their current auditor on the 12 audit quality
attributesidentiiedbyCarcelloet al, (1992). The marketing literature
identiiesthatproductqualityandclientsatisfactionarerelatedbut
uniquecharacteristics(Cronin&Taylor, 1994;Taylor&Baker, 1994),
therefore audit quality and client satisfaction are likely to be separate
constructs that share a close relationship (Behn et al, 1997 p.8). Six (of
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
23
the 12) audit quality variables had a positive relationship with client
satisfaction with the audit team: (i) responsiveness to client needs;
(ii)auditirmexecutivesactivelyinvolvedintheaudit; (iii)effective
interaction with the audit committee; (iv) appropriate conduct of audit
ieldwork; (v)industryexpertise; and(vi)teamandirmexperience
with client. Only one of the audit quality variables was negatively
associated with client satisfaction (“the audit team members maintained
a skeptical attitude throughout the audit engagement”,Behn et al,1997
p.23). However, thisindingisperhapsindicativeofthebalancingact
anauditorhastoperformbetweensatisfyingtheclientandfulilling
professional and stakeholder expectations. Suggestions provided by
respondents to improve client satisfaction included the auditor being
more proactively involved in the client’s business, providing service
beyond statutory compliance and making valueadded suggestions.
Behn et al, (1997) also report that client satisfaction is higher during
theirstfewyearsafterachangeinauditorsandsomeevidencethat
satisfaction is higher when the controller has previous work experience
with the auditor.
Moizer(1998)surveyedUKcompanydirectorsincludedwithin
the Times 1000 in 1987 and again in 1996, to assess respondents view
of the performance of their auditors. The sample consisted largely of
organisations audited by the (then) 1987 Big Eight (79%) and 1996
BigSix(92%), relectingthemarketconcentration oflargeauditirms.
AnanalysisoftheBigEightandBigSixirmsaloneineachofthetwo
periods showed that respondents were equally satisied with the quality
of service provided by their audit irms. The variables which were rated
by respondents as being most inluential in their view of the audit irm
in1987and1996were: (i)personalcontactswiththeauditpartner;
(ii) personal contacts with the audit manager: (iii) personal contacts
with the advisors from the irm (eg tax, management consultancy);
and (iv) the attitude and appearance of the audit team. Interestingly,
a large age effect was observed, with older inance directors being
24
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
morelikelytobesatisiedwiththeperformanceoftheirauditorthan
theiryoungercolleagues. Thisindingsuggestsacohorteffectexists,
relecting the period when an individual was trained.
WarmingRasmussen & Jensen (2001) studying perceptions
of audit quality in Denmark, sampled two groups of external users
(shareholdersandinancialjournalists), managingdirectors(preparers)
and public accountants (auditors), to create an audit quality scale.
Their research was unusual in the sense it sampled external users’
perceptions of audit quality, rather than the simple auditor/auditee
samplesusedbypreviousresearchers. Theinvestigationidentiiedsix
audit quality dimensions, focusing largely on moral and ethical issues,
usingEFA. Thesetheylabelled: (i)personalcredibility; (ii)theauditor’s
independenceofcompanymanagement; (iii)opennessinthereport
aboutmattersofinteresttocreditorsandshareholders; (iv)knowledge
oftheindustry; (v)loyaltyaboutminorityshareholders; (vi)auditors
sceptical attitude to the auditee. The two groups of external users
tended to rate all six of the audit quality dimensions higher than the
auditors and managing directors.
Audit expectations gap
DeAngelo’s classic dual deinition of technical quality (ie of
competence and independence) is necessarily a subjective one. As the
demand for audit services comes from a variety of sources (ie insurance,
credibility, comfort for board members, access to other professional
services) so do expectations of what auditors can reasonably be asked to
perform. Majorinancialscandalshavefuelleddoubtsanduncertainties
about the role of auditing and what auditors can reasonably be expected
to do. Such concerns are not new. Chandler & Edwards (1996, p.12)
identify in the late nineteenth century that the accounting profession
had to deal with similar levels of public concern about the value of
audits as exist today.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
25
Much criticism of the audit profession is as a consequence of wellpublicised corporate failures (Porter, 1993). This mismatch between
stakeholder expectations of auditors and auditors’ performance is
referred to as the“audit expectations gap”,a term irst coined by Liggio
(1974). Theauditexpectationsgaphasbeendeinedasthedifference
between expected levels of performance as viewed by the independent
accountantandtheuserofinancialstatements(Liggio, 1974p.27).
In the US, The AICPA’s Commission on Auditors’ Responsibilities
(Cohen Commission, 1978) was established to consider whether a gap
existed between what the public expects or needs and what auditors can
reasonably be expected to accomplish. A more sophisticated deinition
of the audit expectations gap was attempted by Porter (1993, p.50)
being: “the gap that exists between society’s expectations of auditors
and auditors’ performance, as perceived by society. This gap is made
up of two components:
1.
The‘reasonablenessgap’ –thegapbetweenwhatsocietyexpects
auditors to achieve and what they can reasonably be expected
toaccomplish(similartothedeinitionprovidedbytheCohen
Commission).
2.
The ‘performance gap’ – the gap between what society can
reasonably expect auditors to accomplish and what they
are perceived to achieve. This dimension consists of two
components:
•
a‘deicient standards gap’– the gap between the responsibilities
that can be reasonably expected of auditors and auditors’
existingresponsibilitiesasdeinedbythelaw, regulationsand
professionalguidelines;and
•
a‘deicientperformancegap’ –thegapbetweentheexpected
standard of performance of auditors’ existing responsibilities
and auditor’s performance as expected and perceived by
society”.
26
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Identifying performance gaps is commonplace in the marketing
and service quality literature, and is seen as a means of understanding
perceptions and expectations of quality. The next part of this chapter
considers the development of the service quality literature and its
application to audit services.
Service quality
Interest in evaluating service quality has been motivated by
recognitionofthesigniicanceofservicequalityinbusinesssuccess.
Highlevelsofserviceareseenasameansforanorganisationorirm
to achieve a competitive advantage and position itself more effectively
in the market place (Lewis, 1993). Research demonstrates that high
levels of service quality can lead to: customer loyalty, attraction of
new customers, positive wordofmouth, employee satisfaction and
commitment, enhanced corporate image, reduced costs and increased
business performance (Berry, Bennett & Brown, 1989).
The concept of client satisfaction is less well-rooted in the accounting
literature. Walker (2001) suggests the effectiveness of maintaining good
relationships with clients is often overlooked by auditors. Other writers,
primarily in the US, have encouraged accounting irms to cultivate a
marketing culture (Ahmed & Hopson, 1990), to consider developing a
marketing positioning strategy (Ellis & Mosher, 1995), and to extend
their portfolio of services (Diamantopoulos, O’Donahue, & Petersen,
1995). Andersen (1999) claims few accounting irms have suficient
understanding of themselves, or their clients, to improve the quality of
service they provide to those clients. In conclusion, little is known of
the determinants of service quality provided by UK accounting irms.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
27
SERVQUAL Model
The seminal work of Zeithaml, Parasuraman and Berry (1990)
identiies three fundamental ways in which services differ from
manufactured goods in terms of the way they are produced, consumed
and evaluated. These differences relate to:
1.
Intangibility, relating to performances and experiences, rather
thanobjects. Assuchprecisespeciicationscanrarelybeset. The
nature of accounting relies to a degree on subjective judgements.
Although audit irms adopt procedures such as peer review
and have developed internal quality control procedures, these
quality tools relate more to technical audit quality rather than
considerations of client satisfaction.
2.
Heterogeneity. Serviceperformance–especiallythoseservices
withahighlabourcontent–variesfromdaytoday,fromproducer
to producer. Importantly the interaction between the service
provider and customer can rarely be standardised. Considering
the audit function, Beattie, Fearnley & Brandt (2001) identify that
the integrity of the audit process is critically dependent on the
outcomes of interactions between the audit engagement partner
andcompanyinancedirector.
3.
The inseparability of production and consumption of many
products and services. Service customers are often in the service
factory, observing and evaluating the production process as it
takesplace. Likeotherinancialservices, thecompanyauditis
not consumed in the process. Also, the audit process takes place
largely on company premises and audit staff can create considerable
disruption (Beattie & Fearnley, 1998, p.23).
The most widely known and discussed scale for measuring service
quality is SERVQUAL (Parasuraman et al, 1988; 1991). Since the scale
28
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
was developed, various researchers have applied it across such different
ields as securities brokerage, banks, utility companies, retail stores,
healthcare, and repair and maintenance shops.
Service quality is an exclusive and abstract concept relecting
its “intangibility” as well as the “inseparability of production and
consumption” (Parasuraman, Zeithaml & Berry, 1985). Different
approacheshavebeenproposedtodeineandmeasureservicequality.
Theservicesmarketingliteraturetypicallydeinesservicequalityin
terms of what service recipients receive in their interaction with service
providers (functional, interactive, or outcome quality) (Berry, Zeithaml
& Parasuraman, 1985), and how this technical quality is provided to the
recipients (functional,interactive,or process quality). Parasuraman et al,
(1985) assert that customers perceive service quality in terms of the gap
between received service and expected service. In their development of
SERVQUALtheyidentiiedtendimensionsofservicequality: access,
communication, competence, courtesy, security, tangibles, reliability,
responsiveness, credibility, and understanding. These ten dimensions
werethenclassiiedintothreecategories: searchproperties(credibility,
and tangibles dimensions customers can evaluate before purchase),
experience properties (reliability, responsiveness, accessibility, courtesy,
communication, and understanding dimensions that can be judged
during or after consumption of the service), and credence properties
(competence,and security - dimensions customers ind hard to evaluate
even after the service has been delivered). The ten dimensions discussed
in the 1985 study were reduced to ive in SERVQUAL following
empirical testing. SERVQUAL’s ive dimensions are: assurance,
empathy, reliability, responsiveness, and tangibles (Zeithaml et al, 1990
p.26). Theseivegenericservicequalitydimensionsareoutlinedin
Table 2.2.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
29
Table 2.2:The five generic service quality dimensions
Dimension
Description
Assurance
Knowledge and courtesy of employees and their
abilitytoconveytrustandconidence
Empathy
Caring, individualised attention
Reliability
Ability to perform the promised service
dependably and accurately
Responsiveness
Willingness to help customers and provide
prompt service
Tangibles
Appearance of physical facilities, equipment,
personnel and communication materials
The SERVQUAL instrument consists of two 22item inventories,
alongwithiveitemsthatrespondentsarerequiredtorank. Theirst
inventory requires the respondent to consider their expectations of the
service performance of a type of company (eg accountingirms). The
second inventory requires respondents to consider their perceptions of
theservicequalityofaspeciicservicecompany(eg XYZ & Partners
–CharteredAccountants). Finally, respondentsarerequiredtorank
descriptionsoftheivegenericservicequalitydimensions.
The SERVQUAL model deines service quality as“the extent
of discrepancy between customers’ expectations or desires and
their perceptions” (Zeithaml et al, 1990 p.19). That is, quality is
attained or exceeded when perceptions of service quality meet or
exceed expectations. Conversely, quality is deemed unsatisfactory
when perceptions of performance fail to meet expectations. The
30
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
SERVQUAL deinition of quality assumes consumers (or clients)
expecthighqualityservices. Apotentiallawinthedeinitionoccurs
in that if consumers/clients expect poor quality services and receive
them, then no discrepancy will arise. Zeithaml et al’s exploratory
research suggested four factors shaped customer expectations:wordofmouth communications, personal needs, past experience and external
communications. These factors are described in Table 2.3.
Table 2.3: Factors shaping customer expectations
Factor
Description
Wordofmouth
communications
What customers hear from other customers
Personal needs
Expectations vary depending on respondents’
individual characteristics and circumstances
Past experiences
Previous dealings with similar types of
organisations can increase or decrease
expectationsinspeciicqualitydimensions
External
communications
Direct and indirect messages conveyed by
the service provider to customers (including
price)
Extension to accounting irms
Extendingthisworktoaccountingirmssuggestsexpectationsof
audit quality could be driven by:
•
informal discussions between inance directors of their relationship
with and the service provided by their current auditor,
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
31
•
buyers of audit services require different things from the company
audit. Beattie & Fearnley (1998) identify four audit buyer types.
Theirstgroup(‘thegrudger’)resentsandseeslittlevalueinthe
company audit. The second group (‘the statusseeker’) seeks to
enhancetheirstandingintheinancialmarketbyemployingan
auditor of high repute. A third group,labelled ‘the comfortseeker’,
sees the audit exercise as an assurance exercise for the beneit of the
directors and the company itself. ‘Resourceseekers’ by contrast,
perceiveandvaluetheauditirmasasourceofexpertiseona
rangeofaccounting,inancialandbusinessmatters,
•
clients’previous experiences of dealing with auditors is conceivably
likely to lower their expectations regarding behavioural attributes
such as empathy, but be more demanding with respect to auditors’
technical competence and reliability,
•
externalcommunicationsincludeavarietyofdirectandindirect
messagesconveyedbytheauditirmtocustomers. Thesemight
include: annual reviews or reports published by irms, sponsorship
of charitable events implying social responsibility,an audit manager
promising technical advice by a certain time, or marketing
brochures that suggest a superior level of client service. Zeithaml
et al, (1990) also include price as being under the inluence of
externalcommunications. Signiicantpressureonauditfeesand
related issues such as low balling and auctions for the provision of
audit services have been part of the audit environment for over a
decade, asidentiiedinchapterone. Externalcommunicationhas
become more sophisticated for accounting irms over the past ive
tosixyearswithKPMGbeingtheirstauditirmtopublishan
annualreportwithmanyotherirmsfollowingsuitinsubsequent
years. Firms’ marketing communications are considered in the
present study in chapter four.
32
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Previous studies have indicated that SERVQUAL must be
modiied for each unique service sector (Babakus & Boller, 1992;
Carman, 1990). For example in the tourism industry, Frochot &
Hughes (2000) developed HISTOQUAL to understand the elements
of service delivery of historic houses. Similarly, heritage attractions,
hotel and motel services have received attention with the development
of LODGQUAL (Getty & Thompson, 1994) and LODGSERV (Suh,
Lee, Park, & Shin, 1997) respectively, and the holiday industry with
HOLSAT (Tribe & Snaith, 1998). A similarly developed literature
exists across different service sectors. Intriguingly, despite the extensive
service quality literature,the quality of professional services provided to
business clients has received little research attention (Dart & Freeman,
1994). A literature search reveals only one investigation to utilise the
SERVQUAL model in considering clients’ evaluation of their audit
irm (De Ruyter & Wetzels, 1999), who measured perceptions of
servicequalityusing213clientsofalargeauditirmintheNetherlands
using eight items adapted from the reliability, responsiveness and
assurancedimensionsofSERVQUAL. DeRuyterandWetzels’ work
examines commitment in auditor client relationships, conceptualising
commitment in two forms. These two aspects of commitment are
calculative commitment (an instrumental reasoning that evaluates
thecostsandbeneitsofdevelopingandmaintainingarelationship,
eg Kumar, Scheer, & Steenkamp, 1995) and affective commitment (an
attitudinal dimension representing the affective orientation towards,and
value congruence with, the business partner, eg Gundlach, Achrol, &
Mentzer, 1995). Perceivedservicequalitywasshowntobepositively
related to affective commitment, in line with previous indings
(Geyskens & Steenkamp, 1995; Kumar et al, 1995). Unexpectedly
however, a negative relationship between perceived service quality and
calculative commitment was reported. De Ruyter andWetzels suggest
thisindingislikelytobeduetorespondentsperceptionthereislittle
variationinthequalityofservicedeliveredbyauditirmsingeneral.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
33
The extended gaps model
The SERVQUAL approach is based on identifying differences
between client expectations and perceptions for the service they
receive. Thesedifferencesarelabelled‘gaps’. Fivegapsareidentiied
by the SERVQUAL model. Gaps 1 to 4 denote shortfalls within
the service provider’s organisation and gap 5 as the servicequality
shortfall perceived by clients. Zeithaml et al, (1990) have identiied the
antecedentsofeachoftheseivegapsandstrategiesto‘close’ insome
considerabledetail. TheivegapsaredescribedinTable2.4.
34
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Table 2.4: The five gaps – potential causes of service quality
shortfalls
Gap
Label
Description
Gap 1
Customers’
expectations
–Management
perceptions gap
Discrepancies between
customers expectations and
managers’ understanding of
those expectations
Gap 2
Management’s
perceptions–Servicequalityspeciications
gap
Discrepancies between
service quality standards and
management perceptions of
customer expectations
Gap 3
Servicequality
speciications–
Servicedelivery gap
Discrepancies between service
quality standards and actual
service delivery
Gap 4
Service delivery
–External
communications gap
Discrepancies between
the promises made by the
service provider via external
communications (promised
service) and actual service
Gap 5
Expected service
–Perceivedservice
Discrepancies between the
service expected by the
customer and the service
they have perceived to have
received
Figure 2.2 presents a conceptual model of service quality. The
modelisagraphicalrepresentationoftheivegapsofservicequality.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
35
Figure 2.2: Conceptual model of service quality
Personal
needs
Wordofmouth
communications
Past
experience
Expected service
Gap 5
CUSTOMER
Perceived service
Gap 4
PROVIDER
Service
delivery
Gap 1
External
communications
to customers
Gap 3
Service quality
speciications
Gap 2
Management perceptions
of customer expectations
Gap1identiiesdifferencesbetweencustomers’ expectationsof
quality service and executives’ understanding of those expectations.
36
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Forexample, partnersofsmalleraccountingirmscouldviewtheir
service quality as exemplary, when compared to larger international
irms,because they believe their irm to be more reliable and responsive
toclients’ concerns. Bycontrast, partnersofthoselargerirmsmight
believe their irm size and international network signals strength
from a quality standpoint. This view is supported by some academic
researchers who have used irm size as a proxy for audit quality
(eg Chan, Ezzamel, & Gwilliam, 1993; O’Sullivan, 2000; Palmrose,
1986). Theargumentthatirmsizeisindicativeofauditqualityrelies
onaneconomicargumentthatlargerirmshavefewerincentivesto
reduceauditqualitytoretainanyoneclient, asalargerirmsinancial
dependenceonanindividualclientislowerthanthatofasmallerirm.
The existence of gap 1 suggests such proxies provide a rather narrow
technicaldeinitionofauditquality.
Quality management theory suggests a prerequisite for providing
high quality service is the presence of performance standards which
mirror managements’perceptions of customers’expectations. However,
as Zeithaml et al, (1990) identify, the task of translating managements’
understandinginto service-qualityspeciicationsisnotstraightforward.
Gap 2 identiies differences between customers’ expectations (in
terms of expected service) and management perceptions of customer
expectations, often expressed formally by performance standards. In
an accounting context, an audit irm may have set procedures for
how the audit is conducted to minimise the disruption to the client.
However, theirm’sperceptionofdisruptionisquitedifferenttothat
of the client.
Gap 3 identifies differences between the servicequality
speciicationsandtheservicedeliveredtothecustomer. Suchgaps
are usually created by an unwillingness or inability of the individual
delivering the service to meet the service performance standards
(Zeithaml et al, 1990 p.42).
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
37
An important determinant of customer expectations are the
irm’s external communications (ie marketing literature, website,
sponsorship arrangements, promises made by audit partners). These
communicationssetstandardsagainstwhichtheirmisjudged. Gap4,
being the difference between the actual service and promised service,
will have an adverse affect on customers’ perceptions of service quality.
The quality message signalled by accounting irms’marketing literature
is examined within this monograph in chapter four.
The servicequality shortfall perceived by customers is represented
by gap 5. Gaps 1 to 4 are shortfalls which occur within the service
provider’s organisation. Gap 5 is simply the difference between the
service a customer expects to receive and the service they perceive thay
have received. Gap 5 can be closed by keeping gaps 1 to 4 closed.
Auditor attributes
A number of researchers have considered the inluence of the
auditor’s personal attributes (ie their personality and their moral
reasoning). Other scholars have emphasised the importance of the
relationship between the audit engagement partner and the client’s
inancedirector(egBeattie&Fearnley,1995;Beattieet al, 2001).
Personality proiles of accountants
Previousinvestigationshaveconsideredthepersonalityproiles(ie a
description of their behaviour) of practising accountants.The majority
of these studies have utilised the MyersBriggs Type Indicator (MBTI).
The MBTI assesses four bipolar dimensions of personality:introversionextroversion; perceiving-judging; sensing-intuition; thinking-feeling.
According to Jung, individuals may be categorised at the extremes
of these bipolar dimensions. This gives 16 possible personality types,
which are referred to a fourletter designation (eg ISTJ or ENFP).
38
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
However, the strength of preference on each dimension may vary
withineachpersonalitytype. Thestereotypicalproileforpractising
accountants is conceived as an ISTJ (introvertedsensingthinkingjudging) type,which is supported by empirical investigation in Australia
(Booth&Winzar, 1993), theUS(Descouzis, 1989; Jacoby, 1981;Wolk
& Nikolai, 1994) and UK (Shackleton, 1980). The characteristics of
anISTJproileareshowninTable2.5. Table2.5framesdescriptions
ofthetypicalaccountingproileintermsoftheBigFivefactormodel,
a competing personality model, widely accepted by personality and
individual difference scholars. MBTI by contrast is more widely used
by practitioners, although the nature of its measurement model makes
itdificultforthetypesofcorrelationalstudiesemployedbyacademic
researchers. Both however, are conceptually and empirically similar,
McCrae&Costa(1989)reportcorrelationcoeficientsrangingfrom
.46 to .69 when the four of the Big Five dimensions are correlated
with the four bipolar scales of MBTI.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
39
Table 2.5: MBTI and Big Five Personality domains and representative
traits
MBTI
Corresponding
Big Five label
Typical
Accounting
profile
Accounting
profile in terms
of Big Five
Description
of typical
accounting
profile
EI
Extraversion
I
Extraversion
Introverted,
reserved and
serious. Prefers
to be alone
or with a few
close friends.
SN
Openness
S
Openness
Down to
earth, practical,
traditional and
set in their
ways
TF
Agreeableness
T
Agreeableness
Hardheaded,
practical
sceptical,
proud and
competitive.
Tend to express
anger directly.
JP
Conscientiousness
J
+
Conscientiousness
Conscientious
and well
organised,
have high
standards and
always strive to
achieve goals.
Neuroticism
Unknown
Unknown
40
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
An individual with an ISTJ proile is likely to develop the necessary
technical skills, (based on their high score on the conscientiousness
dimension and low score on agreeableness). However, they are also
likely to be perceived by others as introverted, reserved, passive and
cold. Such an individual may lack the personal qualities necessary to
establish rapport and maintain a longterm relationship with a client.
However, a more ‘agreeable’ individual might be seen as compassionate
and good natured, but their eagerness to cooperate and tendency
to avoid conlict could lead to accommodating behaviours, clearly
undesirable in an audit engagement partner.
Ethical reasoning of auditors
Other researchers have considered the ethical reasoning of
accountants. Ponemon & Gabhart (1990) use Kohlberg’s stage model
of moral development and ethical cognition to examine an auditor’s
reasoning in a hypothetical audit independence conlict scenario.
Sampling 119 audit partners and managers report a systematic
relationship exists between auditors’ ethical cognition and their
resolution of an independence conlict. This work was developed
further by Windsor & Ashkansay (1995) to identify three styles of
auditor decision making: autonomous, responsive to personal beliefs
andlikelytoresistclientpressure; accommodating, wheretheauditor
isresponsivetobothpersonalbeliefsandclientmanagementpower;
and pragmatic, where auditors were responsive to client management
power, regardless of their own personal beliefs. Beattie, Fearnley &
Brandt (2001) developed six case studies describing the interaction
betweentheinancedirectorandtheauditengagementpartnersto
establish the factors that inluence the nature and outcome of key audit
interactionsinvolvingsigniicantaccountingissues. Usinggrounded
theory, a process of building theory inductively by means of the
qualitative analysis of interview data, they develop a hierarchy of audit
engagementpartner(AEP)types,showninigure2.3.
THE NATURE
OF
AUDIT QUALITY
AND
RELATED LITERATURE
41
Figure 2.3: A hierarchy of AEP (seller) types and their characteristics
Type
AEP Quality
Characteristics
Crusader
Extremely high professional and
personalintegrity;preparedto
escalate
Safe hands
Highprofessionalintegrity;
identiiescloselywithclient;
prepared to escalate
Accommodator
Moderateprofessionalintegrity;
will knowingly bend the rules
under pressure
Truster
Moderateprofessionalintegrity;
may unknowingly permit rules
to be bent
(Source: Beattie et al, 2001, p.276)
Key points
This chapter has reviewed the literature considering the conceptual
framework that surrounds audit quality. Audit quality is generally
conceived as a function of the auditor’s ability to identify a breach in
theclient’sinancialreportingsystem(ie their competence) and their
willingness to disclose such an error (ie their independence).
A number of studies have attempted to identify auditor
characteristics which contribute to audit quality. Such studies have
typically sampled the views of auditors,company directors and inancial
statement users.
A distinct yet related literature considers the audit expectations
gap: differences between expected levels of performance as viewed
between the independent accountant and the user of financial
42
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
statements. This gap has been subcategorised into a performance gap
and a reasonableness gap.
The concept of service quality and the related SERVQUAL
modelhavebeenidentiiedbymarketingandoperationsmanagement
scholars. SERVQUAL identiies ive potential ‘gaps’ between the
expectations and perceptions of the service provider and consumer.
The SERVQUAL model is driven by discrepancies between customer
expectations of service quality and perceptions of the service they have
received (gap 5). This gap is driven by four discrete service quality
gaps that exist within the customers’ business.
No auditing research appears to have considered the applicability
of the SERVQUAL model. This is perhaps surprising given societal
interest in the audit expectations gap, and auditor interest in client
satisfaction in a competitive audit environment, characterised by low
growth and downward pressure on fees (eg Elliott & Jacobson, 1995).
A related literature considers the attributes of auditors including
their personality, ethical reasoning, and attitudes and beliefs. A range of
personality studies applying the MBTI conclude a typical accounting
proileexiststhatlendsitselftothedevelopmentofhightechnicalskills,
but rather less to the development of social skills desirable in relationship
marketing. A hierarchy of AEP types has been developed by Beattie
et al, (2001) on the basis of the personal and professional integrity of
the auditor and their ability to take their professional responsibilities
to and beyond the level of their statutory duties where necessary.
ChApter three
developing An AlternAtive ApproACh to
Audit quAlity using other literAtures
Audit quality is not a unitary concept. It should be divided into
1. quality of service: ie factors which affect the client’s experience of
the audit process
2. quality of opinion: ie factors which contribute to a process which
is likely to reach the right answer
The two aspects can sometimes be in conflict, but the key to success
is delivering the highest “quality of service” without compromising
on the “quality of opinion”.
(Big 4/5 Partner)
Introduction
The purpose of this chapter is to propose a broad model of audit
quality, which provides a conceptual foundation for understanding
quality from the perspective of auditor, audit client and other
stakeholders. The conceptual framework integrates the service quality
andauditexpectationsliteratures. Eightgapsareidentiied. Gaps1
to4areassociatedwithshortfallsoccurringwithintheauditor’sirm.
Gap 5 represents the audit quality shortfall perceived by clients and
stakeholders. Gaps 6 to 8 denote the shortfalls perceived by stakeholder
groups, theauditexpectationsgap. Thischapteridentiiespotential
antecedents of each of these hypothesised eight gaps, and how they
might be closed.
44
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Understanding client expectations
Service and technical quality gaps
Figure 3.1 describes a conceptual model illustrating differences
between an individual client’s expectations of audit quality and the
quality of audit they actually receive. Four factors (or‘gaps’) contribute
to this problem. This analysis of these four gaps is adapted from
Zeithaml et al’s (1990) seminal work and modiied for accounting
irmsandtheirclients.
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
Figure 3.1 Gaps 1 to 4 - between client expectations and audit
quality
Client
expectations
Gap 1
Auditirm
perceptions of
client expectations
Gap 2
Auditirmquality
standards
Gap 3
External
communications
Audit quality
Gap 4
Technical
quality
Service
quality
A gap 1 problem exists where the firm has an insufficient
understanding of its client’s needs. Zeithaml et al, (1990) identify three
factors which are likely to create gap 1 problems:
45
46
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
1.
A lack of marketing research, preventing an understanding
of clients’ needs and expectations, using formal and informal
informationgatheringexercises;
2.
Poor upward communication from employees at lower levels to
seniormanagement;
3.
Too many levels of management, separating senior management
from those dealing with clients on a daytoday basis.
Gap1problemsarequitelikelyinaccountingirmsasrelatively
few irms undertake marketing research (Andersen, 1999) and the
structureofaccountingirms, withpartnersbeingquitedistinctand
elevated people within the organisation (see for example Johnson,
2002 p.1014). The key to closing gap 1 marketing research should
focus on service quality issues that are most important to clients and
whatclientsthinktheirmcan(andshould)dowhenproblemsoccur
(Zeithaml et al, 1990).
Gap 2 exists where differences occur between management
perceptionsofclientexpectationsandthequalityspeciicationsofthe
auditirm(ie theformalprocedurestheirmadoptsforthecompany
audit). Awidegap2couldexistwhentheirmlackscommitment
to audit service quality, for example where it places emphasis on
cost reduction or short-term proit. Four factors underlie gap 2
problems:
1.
A lack of management commitment to service quality, where
service quality is not seen as a strategic goal. For example,does the
irmplacegreateremphasisonfeegrowthratherthanimproving
service quality or developing staff to improve the quality of the
auditprocess?Anaccountingirmwhichbelievesitiscommitted
to quality may be committed to technical quality (competence
andindependence, inthetraditionaldeinition)orautomation
and productivity activities which compare equally (or favourably)
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
47
withotherauditirmsbuthavefewsystemsinplacetomanage
service quality.
2.
The extent to which managers believe it is feasible to provide
high quality service, the extent to which client expectations
can be met. Encouraging innovation is said to be the key to
developing perceptions of feasibility. For example, Ernst &
Young’s Online service, allowing clients access to their knowledge
management database, could be described as an example of such
an innovation.
3.
The degree to which automation, and formal and informal
operating procedures are standardised to ensure consistent levels
ofservice. Auditautomationprovidesanopportunityforirms
to standardise tasks. Changes in work processes, for example
establishing a culture of openness and proactive share of knowledge
also contribute to the development of service quality.
4.
The level to which service quality goals are based on clients’
standardsandexpectationsratherthanirmstandards. Adapting
Zeithaml et al’s (1990) analysis to accounting irms, suggests
effectiveservice-qualitygoalswillhaveivefeatures:
•
•
They are designed to meet client expectations
•
They are accepted by audit staff
•
•
Theyarespeciic
They identify the most important job dimensions
They are measured and reviewed with appropriate feedback
Gap3isthedifferencebetweentheauditservicespeciications
(asdrawnupbytheirm)andtheactualqualityoftheauditservice
delivered: a serviceperformance gap. The causes of a service
performancegapareidentiiedas:
48
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
1.
Role ambiguity, relecting uncertainties audit staff may have about
whattheirmortheauditengagementpartnerexpectsfromthem.
As professional service organisations,accounting irms traditionally
spend a high proportion of their fee income on training. It is likely
audit staff are wellequipped to deal with the technical aspects of
their job and cope with changes that affect their work. However,
communications and feedback between partners and staff are
notseenasstronginaccountingirms(eg KPMG’s Leadership
Project described in Johnson (2002 and Accountancy Age, February
1995).
2.
Role conlict, the extent to which an employee may perceive that
they cannot satisfy all the demands of all individuals (both clients,
groupsinternaltotheauditirmandstakeholdergroups). Role
conlict is a central issue in the study of audit quality. For example,
being clientcentred may be seen as desirable to service quality,but
as potentially accommodating behaviour from an independence
standpoint. ‘Accomodators’ (a term coined by Beattie et al, 2001
p.276) are described as having“moderate professional integrity,but
will knowingly bend the rules under pressure”, and possessing a
desire to be helpful to the company where possible. Role conlict
can be reduced by deining job roles in terms of client and societal
expectations. Employee-jobit, istheitbetweentheauditor’s
skillsandtheirjob. Accountingirmsinvestconsiderablesums
in recruiting and selecting appropriate people, a process which is
supported by professional training, appraisal of auditing trainees
andarelativelylargeattritionrateoftrainees. DeRuyter&Wetzels
(1999)suggestwhenhiringpersonnelauditirmsshouldscreen
for the social abilities that facilitate establishing and maintaining
longterm relationships based on affective commitment.
3.
Technology-jobit, thatis, theappropriatenessofthetoolsand
technology that auditors use to perform their work. Manson,
McCartney & Sherer (1997) found irms making use of IT
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
49
in planning, controlling and recording of audit work (audit
automation) and reported a perceived improvement in the quality
of audit work. Although no evidence of deskilling was evident
from the audit automation, it had supplanted some of the menial
work previously undertaken by clerical and secretarial workers,
requiring the redeployment of these workers.
4.
Supervisory control systems, how the audit irm evaluates and
rewards employees. For example,KPMG’s human resource strategy
identiiessixbehaviouralcharacteristicswhichareimportantforits
future success (client responsiveness, business skills, management,
personal effectiveness, social skills and thinking skills). Promotion
is said to be on the basis of performance in terms of these
competencies.
5.
Perceived control, being the extent to which employees are able
to act lexibly in dificult situations. Accounting irms have a
tradition of valuing individualism (eg Thornbury, 1999) where
mavericks are “lionised” (eg Johnson, 2002 p.1014). Increasing
audit staffs’ perceived control (ie theirabilitytoinluencedificult
situations and their ability to choose outcomes) will improve
service quality.
6.
Teamwork, the degree to which partners and employees function
as a team. Prior investigations into the dimensions of audit quality
have reported audit team factors as being more important than
auditirmfactors(Behnet al, 1997; Carcelloet al, 1992; Schroeder
et al,1986). When employees identify with the norms and values of
an organisation, service quality is likely to be improved (Schneider
& Bowen, 1985).
Differencesbetweenwhatairmpromisesaboutauditservices
and what it delivers is gap 4 of the SERVQUAL model. The factors
that contribute to such a gap are shown in Figure 3.2.
50
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Figure 3.2: Key features contributing to gap 4 (adapted from
Zeithaml et al, 1990, p.116)
Service
Delivery
Key contributing factors:
1. Inadequatehorizontal
communication
inadequate communication
between marketing and audit staff
differences in policies and
proceduresacrossoficesand
locations
2. Propensity to overpromise
External
communication
to customer
Discrepancies between ser vice deliver y and exter nal
communications, in the form of exaggerated promises, or an absence
of information about service quality can powerfully inluence client
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
51
expectations of service quality (Zeithaml et al,1990 p.116). Accounting
irms’ externalcommunicationsareexaminedinchapterfourofthis
monograph.
Zeithaml et al, (1990) suggest the key to developing appropriate
and effective communications about service quality is to:
1.
Emphasise primary quality determinants. That is, identify the
features of audit quality that are most important to clients and
stakeholders.
2.
Managing clients’ expectations. A client’s expectation is likely to
beaproductofexperiencewithotherauditirms(possiblyasa
result of working as an external auditor). Also, experience with
providers of other professional services (eg lawyers, bankers) are
likelytoaffecttheirperceptionofanaccountingirm’squalityof
service.
3.
Manage controllable sources of clients’ and stakeholders’
expectations such as external communications literature, and
price. A key question might be how an accounting irm can
lower clients’ and stakeholders’ expectations when competing
irmsmaybepromotinginlatingpromises. Zeithamlet al, (1990
p.126) identify this problem is particularly acute when an industry
is suffering from a poor image. Recent accounting scandals
have undoubtedly damaged the reputation of the accounting
profession and particularly its audit function. Zeithaml et al,
(1990 p.126) cite the example of American Airlines (AA) who
ran an advertising campaign informing airline customers of the
uncontrollable factors that make airlines unreliable (the dimension
of service quality airline customers rated most important). As
AA’s reliability was the highest of its peers, the advertisement
was credible. As clients and stakeholder groups rate Reputation
and Credibility as the most important factors inluencing audit
quality, anaccountingirmcouldattempttoidentifywithclient
52
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
and stakeholder groups’ frustrations (eg the reasonableness gap
between society’s expectations and what an audit can reasonably
provide), explaining how some matters are uncontrollable by the
accountingirm.
Gap 5 is the audit quality shortfall perceived by clients, as distinct
fromshortfallswhichexistwithintheaccountingirmitself. Gap5is
illustrated in Figure 3.3. Key determinants of the service expected by
clients include wordofmouth communications, personal needs, past
experience, andexternalcommunicationsfromtheaccountingirm
(Zeithaml et al, 1990). Gaps 1 to 4 are the four serviceprovider gaps
that contribute to gap 5.
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
53
Figure 3.3: Gap 5 - between client expectations and perceived audit
service
Word of mouth
communications
Client
needs
Past
experience
Expected audit
service
Gap 5
Client
Auditirm
Perceived
audit service
External
communications
The relationship between gaps 1 to 4 and gap 5 is illustrated in
Figure 3.4, which illustrates the various factors which relate to the
accountingirmandtheirrelationshiptoservice-providergaps(ie gaps
1 to 4).
54
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Figure 3.4: Extended gaps model of audit quality
Market research
Upward
communication
Gap 1
Levels of
management
Management
commitment to
quality
Goal setting
Task standardisation
Gap 2
Technical
quality
Perception of
feasibility
Teamwork
Gap 5
(Audit service
quality)
Employee-jobit
Technology-jobit
Perceived control
Gap 3
Supervision
Role conlict
Role ambiguity
Communication
Overpromising
Gap 4
Service
quality
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
55
Audit expectations gap
Figure 3.5 illustrates the components of the audit expectations
gap, gaps 6 to 8 and gap 4. The audit expectations gap represents the
audit quality shortfall perceived by stakeholders. Gaps 6 to 8 are an
adaptation of Porter’s (1993) conceptualisation of the audit expectations
gap’s structure and composition. Porter’s work is in turn an extension
of Liggio (1974) and the Cohen Commision (Commission on Auditors’
Responsibilities, 1978), described in chapter two.
56
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Figure 3.5 Gaps 4 and 6 to 8 - between stakeholder expectations
and audit quality
Societal
expectations
Gap 6
Reasonable
expectations
Gap 7
Legal & professional
expectations
Gap 8
External
communications
Gap 4
Technical
quality
Audit
quality
Service
quality
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
57
Gap 6 is labelled the ‘reasonableness gap’, representing the gap
between what society expects auditors to accomplish and what
they can reasonably be expected to accomplish (Porter, 1993). The
reasonablenessgapisanalogoustogap1, whereanaccountingirmhas
insuficientawarenessofitsclient’sneeds. Gap6bycontrastrepresents
a lack of awareness on the part of stakeholders of what an auditor can
reasonablybeexpectedtoachieve. Porter(1991, 2000)identiiesthe
antecedent of these unreasonable expectations (ie gap 6) is a lack of
education of non-auditor interest groups. An accounting irm’s external
communications will also inluence societal expectations of what can
reasonably be expected of an auditor. Although many irms produce an
annualreview, asidentiiedinchaptersix, noirmattemptstoaddress
within these documents what an audit can reasonably be expected to
achieve (or not achieve). If auditors are to close gap 6, then the audit
profession might consider ways in which direct (eg inancialstatement
users) and indirect (eg non-inancial community) stakeholders are
informed about the audit process.
Gaps 7 and 8 represent performance gaps, between what society
can reasonably expect auditors to achieve and what they are perceived
toaccomplish. Speciically, gap7islabelledadeicientstandardsgap,
a gap between those responsibilities an auditor can reasonably be
expected to perform, and those responsibilities as enshrined in law
and professional regulations. Corporate governance debates (eg the
Cadbury Report, 1992; the Greenbury Report, 1995; the Hampel
Report1998; theTurnbullReport, ICAEW, 1999; andICAEW, 2000)
create an expectation of what an auditor should be doing leading to
adeicientstandardsgap.
Gap 8 is a‘deicient performance gap’ (Porter, 1993), the gap
betweenauditors’ responsibilitiesasdeinedbylawsandprofessional
regulations and actual auditor performance (or audit quality). The
antecedents of gap 8 are, irst, a lack of knowledge on the part of
auditors under statute and case law, quasigovernmental regulations
58
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
and professional promulgations (Porter, 2000), second, a simple lack
of competence, and third, the lack of quality controls procedures
withintheirm. Porter(2000)suggeststhekeytoclosingthisgapis
further education for practising as well as trainee accountants. The
development of further quality control procedures eg second partner
review, aninternalmonitoringmechanismintroducedbyauditirms
to improve auditor independence (AICPA, 1986) or subjecting an
ongoing audit to a random hot review would also contribute to
closing gap 8.
Key points
The chapter has developed an alternative model of audit quality,
illustrated in Figure 3.6. Audit quality consists of both technical and
service quality aspects.
DEVELOPING
AN
ALTERNATIVE APPROACH TO AUDIT QUALITY USING OTHER
LITERATURES
59
Figure 3.6: Audit quality gaps 1 to 8
Client/stakeholder
expectations
Gap 5
Perceived quality of
audit service
Audit
quality
Gap 6
Gap 8
Legal & professional
standards
Gap 7
Stakeholder
expectations of
‘reasonableness’
Gap 4
External
communications
Gap 1
Gap 3
Auditirmquality
standards
Gap 2
Auditor perceptions of
client expectations
Themodelidentiieseightgaps. Gaps1to4relatetoservice
qualitydeiciencieswithintheaccountingirm.
60
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Gap 5 is the difference between client’s and stakeholders’
expectations of the quality of the audit and their perceptions of the
qualityoftheauditserviceprovidedbytheauditirm. Thefourgaps
createdbyqualitydeiciencieswithintheaccountingirmmakeup
gap 5. Gaps 6 to 8 represent the audit expectations gap, representing
a reasonableness gap and two performance gaps.
Each gap has a number of potential antecedents. Identifying
these antecedents provide the key to understanding how the gap may
be closed.
ChApter four
AnAlysis of Auditors’ externAl
CommuniCAtions
The importance of external communications to service
quality
Chapter two introduced the SERVQUAL model, developed to assist
executives identify differences between customer expectations and
perceptions of service quality performance. This chapter will examine
the external communications provided by audit irms. Differences
betweenwhatairmpromises–implicitlyorexplicitly-aboutthe
quality of audit services it provides, and what it is perceived to deliver
is gap 4 discrepancy, described in chapter three. The three major
contributing factors to gap 4 are: irst, inadequate communication
betweenmarketingandauditstaff; second, differencesinpoliciesand
proceduresacrossoficesandlocations; andthird, asimplepropensity
to overpromise. The objective of this chapter is therefore to identify
howirmspromotetheservicesauditorsdeliver.
As described in chapter two,the audit environment is characterised
by increasing competition and downward pressures on fees. These
environmental pressures mean auditors feel the need to acquire new
business and to meet or beat the competition. The greater the extent
to which a service organisation feels pressured to generate new clients,
and perceives the industry norm is to overpromise, the greater is the
irm’spropensitytoover-promise(Zeithamlet al, 1990 p.123). The
creation of the audit tender (a ‘beauty parade’ where the client chooses
the most attractive tender often on the basis of price), and ‘lowballing’
62
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
(a low introductory price,in the hope of costs decreasing in subsequent
years, or the expectation of fees rising, or gaining additional nonaudit
business in the future) have also become characteristics of the audit
environment in the past decade. Price is said to be an indicator of
service quality (Berry, Bennett, & Brown, 1989 p.1223). Price also
sets expectations for the quality of service, particularly when other
indicators are not available. That is,when service outcomes are dificult
to assess (eg a company audit), clients may use price as a surrogate
for quality. A similar approach has been adopted by a number of
investigationsintoinancialreportingandearningsquality, wherethe
auditfeeservesasaproxyforauditquality–seechaptertwo.
The next part of this chapter examines external communications
literatureofthe20largestaccountingirmsintheUK(asreportedby
Accountancy Age, April2002). Eachofthe20irmswascontactedby
telephone in April 2002 to request the following: their annual report (if
one was produced), any marketing literature which promotes auditing
and assurance services. Each of the 20 irms has a website which
promotestheirmanditsservices. Thedatasetincludesannualreports,
irm’swebsitesandothermarketingliterature. Usingthisdatasetthe
following issues were examined:
•
•
•
How audit services are described within the literature
Explicit (and implicit) references to audit quality (both technical
and service aspects)
Howirmsconveyanimageofqualitytopotentialandexisting
clients and stakeholders
Annual reports and marketing literature
Theconceptofaccountingirmsproducinganannualreportis
relativelyrecent, withtheirstbeingproducedbyKPMGin1996. At
the time KPMG, claimed its decision to open its books was to present
ANALYSIS
OF
AUDITORS’ EXTERNAL COMMUNICATIONS
63
an image of a transparent business to clients and stakeholders. Other
irmsquicklyfollowedsuit. InApril2002, thelargest20accounting
irmsintheUK(asreportedbyAccountancy Age) were contacted to
establish whether they produced an annual report or other literature
whichprovidedinformationonthescopeoftheirm’sservices. This
information is summarised in Table 4.1:
64
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Table 4.1: Top 20 firms (as reported in Accountancy Age 18 April
2002) external communications.
Firm
Pricewaterhouse
Coopers
2. KPMG
3. Deloitte & Touche
4. Ernst & Young
5. BDO Stoy
Hayward
6. Grant Thornton
7. PKF
8. Baker Tilly
9. Howarth Clark
Whitehill
10. HLB Kidsons
(Numerica)
11. Smith &
Williamson
12. Moore Stephens
13.MazarsNeville
Russell
14.Tenon
15. RSM Robson
Rhodes
16. BKR Haines
Watts
17. Levy Gee (now
Numerica)
18. Saffery
Champness
19. MacIntyre
Hudson
20. Chantrey
Vellacott DFK
Annual
Review
Marketing
Literature
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
√
Remarks
Employee annual
review
No annual review
No annual review
No annual review
No annual review
Noinancial
information
Website
No annual review
√
√
No annual review
√
No annual review
√
No annual review
√
Abbreviated annual
review
No annual review
√
√
No annual review
ANALYSIS
OF
AUDITORS’ EXTERNAL COMMUNICATIONS
65
Ofthe20irmscontacted, nineproducedsomeformofannual
reviewusuallywithsomesummaryinancialinformation. Allirms
produced marketing literature whether in the form of hardcopy
material or via their website. Website details accessed are shown in
Table4.2. Nineteenofthe20irmsundertooksomeformofaudit
and assurance activity, the exception being Tenon. The four largest
irmsmadelittlereferencetoauditservices, labellingtheseactivities
as assurance services. For example, KPMG describe the development
of their assurance service:
Evolution of our risk-based approach to audit attracts more and
more adherents. By helping organisations and individuals achieve
their objectives and succeed in the new economy, through measuring
performance, managing risks and leveraging knowledge, we increased
Assurance revenue last year by 21% to £413 million.
66
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Table 4.2 Top 20 UK firms website details
Firm name
Website address
1. Pricewaterhouse Coopers
www.pwcglobal.com/uk/eng/
main/home/index.html
2. KPMG
www.deloitte.co.uk
3. Deloitte & Touche
4. Ernst & Young
5. BDO Stoy Hayward
www.kpmg.co.uk/uk/index.
www.ey.com/uk
www.bdo.co.uk
6. Grant Thornton
www.grantthornton.co.uk
7. PKF
8. Baker Tilly
9. Howath Clark Whitehill
10. HLB Kidsons (Numerica)
11. Smith & Williamson
12. Moore Stephens
13.MazarsNevilleRussell
14.Tenon
15. RSM Robson Rhodes
16. BKR Haines Watts
17. Levy Gee (now Numerica)
18. Saffery Champness
19. MacIntyre Hudson
www.pkf.co.uk
www.bakertilly.co.uk
www.howathcw.co.uk
www.numerica.com
www.smith.williamson.co.uk
www.moorestephens.co.uk
www.mazars.co.uk
www.tenongroup.com/home
www.rsmi.co.uk
www.hwca.com
www.levygee.couk
www.saffery.com
www.hughesallen.com
20. Chantrey Vellacott DFK
www.cvdfk.com
Eachoftheremaining15irmsmadereferencetoauditservices,
perhaps relecting conservatism in audit methodology or how these
services are marketed to clients.
Larger irms emphasised their global networks and their provision
of some form of professional service to FTSE 100 companies. Medium
sized irms typically placed greater emphasis on particular sectors
ANALYSIS
OF
AUDITORS’ EXTERNAL COMMUNICATIONS
67
(eg PKF emphasise charities, BDO Stoy Hayward the growing business
andMazarsNevilleRussellinsurance).
Themajorityofirmsmadeeitherimplicitorexplicitreference
to aspects of technical and service audit quality within their literature.
Without exception, each attempts to differentiate itself from its
competitors to achieve competitive advantage. Firms try to differentiate
themselves by emphasising:
•
•
•
•
•
The uniqueness of their irm (eg Ernst & Young produces a
series of images to promote the idea that their employees are
fromdiversebackgroundsandhavediverseinterests; eachofthe
largestfourirmstendtophotographpartnerscasuallydressed, far
removed from a stereotypical image of the Chartered Accountant,
eg Bougen, 1994),
the uniqueness of their services (eg Ernst&YoungOnline–an
interactive online connection with clients, described as “putting
large amounts of our knowledge at our clients’ disposal in formats
designed to be useable and searchable, but also allows us to share
privileged information, to collaborate online and in a secure
environment, and to supply rapid answers to clients’ questions”
claiming 80% marketpenetration amongst FTSE100 companies.
By comparison, PwC offer a service labelled MindLinkTM),
the scope of their services (eg Smith & Williamson’s assertion they
offeranunparalleledrangeofservicesforairmoftheirsize),
the technologies they use to undertake the audit (eg Deloitte’s
AuditSystem/2TM),
the processes associated with the activities (eg service quality
processes, measured by feedback from clients).
Two themes are apparent in irms’ literature: the notion of
“adding value” to the client and the importance of the calibre of
theirm’semployees/partners(usuallycollectivelyrefereedtoasthe
68
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
“irm’speople”). Allirmsidentifyinsomewaythesuccessoftheir
organisation is dependent on attracting and retain the best staff. How
irmsachievethisisdoneinanumberofdifferentways. Forexample,
in a headline article on their website, labelled “The Promise of Quality:
doing the best work requires the best people”, Deloitte Touche state:
Deloitte’s business strategy is built on principles of quality, integrity
and efficiency of delivery. Our commitment to servicing business is
made even stronger by the enhanced strength of our professionals who
deliver quality service every day
(John P. Connolly, CEO, Deloitte Touche, 2002)
Perhaps referred directly to the expectations/perceptions
dichotomy that characterises the concept of service quality that
underlies the SERVQUAL model, Ernst & Young (2001) state:
Expect a lot from Ernst & Young
KPMG emphasise communication between staff, including open
meetings, feedback via the irm’s intranet site, and an independent
employeesurveytoestablishhowtheirm’svaluesaffecthowstaffgo
about their work. Ernst & Young highlight the “skills, diversity and
passion” (Ernst & Young, 2001 p.5) of their employees, of providing
value added.
Twomedium-sizedirmsmakespeciicreferencetoservicequality,
and implicit reference to the concepts underlying the SERVQUAL
model:
We are prepared to define our standards of service and adhere to
them, recognising that each client has a unique set of requirements.
We accept that each client, uniquely, is a judge of the standards we
have achieved in our service. At least once a year a senior partner of
the firm, usually a member of the management board unconnected
with service delivery, will carry out a service and quality review to
ANALYSIS
OF
AUDITORS’ EXTERNAL COMMUNICATIONS
69
establish whether we are meeting your expectations and to agree on
any areas where our service can be improved.
(RSM Robson Rhodes)
Today’s businesses demand value for money – and are sophisticated
judges of good service. It is reassuring that our clients value the
responsiveness and quality of our service. For them, our enthusiasm
and innovative flair or experience and commercial expertise make
us stand out from the crowd. Our client care programme and service
reviews ensure our clients continue receiving our very best, professional
service. If a client is ever less than satisfied, we will remedy the
situation immediately. We identify, prioritise and regularly exceed
their expectations, delivering that bit extra in an entrepreneurial
way.
(Baker Tilly)
Other irms communicate ideas of service quality less directly.
Saffery Champness identify simply they “are committed to a proactive
and personal approach that is responsive to your needs”. However, the
sections of their report bear labels such as personal approach, expertise,
independence, integrity, approachability, responsiveness, knowledge
and dependability. Each of these labels are the key characteristics
of quality and form part of the AUDITQUAL model developed in
chapterive.
Ofthefourlargestirms, KPMGistheonlyonewhichdescribes
how the clients’ views on the services provided and the quality of
relationships developed. KPMG use a system of independent client
reviews to “ensure that client comments can be both candid and
open” (KPMG, 2001 p.18) and a Client Service Board reporting to
the KPMG main board. The three key performance indicators used
in client feedback are: quality of overall relationship, proactivity, and
value for money.
70
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Rather less emphasis is given within irms’ annual reviews to
societal expectations of their performance. Arguably, annual reviews
may have more a client focus than a stakeholder focus, perhaps being
produced for existing and potential clients, rather than inancial
statement users, analysts, investors and employees.When stakeholders
are mentioned it is usually in the context of the irm’s charitable
activities,sponsorship activities are a form of marketing communication.
For example:
We are also committed to supporting our communities through
volunteering and partnership, especially to help those who are
disadvantaged. Our stakeholder report illustrates the breadth of
these activities.
(Mike Rake, UK Senior Partner, KPMG, 2001)
Key points
Airm’sexternalcommunicationshasabearingonclient’sand
stakeholders’ expectations of audit quality (both technical and service
dimensions). Poor horizontal communications between the irm’s
marketing function and operational staff and any tendency to overpromise will create a gap (SERVQUAL gap 4) leading to a negative
evaluationoftheirm’sperformance.
Accounting irms are keen to project an image of providing a
high quality of service, of adding value to clients. Firms attempt to
achieve this by the employment of imaginative, highcalibre staff, the
development of new assurance methodologies and technologies, and
ofthedevelopmentofaglobalirm.
Quality is usually described in relation to service quality, with
some irms explicitly describing how they measure and monitor service
quality. Less emphasis is given within their literature to technical audit
quality attributes.
ChApter five
development of AuditquAl
The development of a questionnaire (measure) to be used
successfullyinacademicorpractitionerresearchisadificultandtimeconsuming process (Schmitt & Klimoski, 1991). Prudent scholarship
suggests an instrument should demonstrate satisfactory measurement
(ie psychometric) qualities before it is used in applied research, which
may inluence policy (see Duff, 2001 for a recent review).
As Schwab (1980) points out,measures are often used before adequate
data exist regarding their reliability and validity. Many researchers
have drawn seemingly significant conclusions from the application of
new measures,only to have subsequent studies contradict their findings
(Cook et al, 1981). Often scholars are left with an uncomfortable
and embarrassing realization that results are inconclusive and that
very little may be known about a particular topic.
(Hinkin, 1995 p.967)
It is therefore essential that measurement instruments
(ie questionnaires) are constructed using sound development
procedures. Without such evidence, a researcher’s questionnaire may
benomorevalidorreliablethanthoseappearinginpopularmagazines
such as Cosmopolitan. The development of a new measure can be
broken down into three stages (Hinkin, 1995; Schwab, 1980). The irst
stage is item development, where the individual questions (items) are
created or generated. The second stage is scale development, or the
manner in which items are combined to form new scales. The third
stage is scale evaluation, where the new measure undergoes systematic
72
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
psychometric examination. This chapter will be presented in the
order of these stages and further broken down into steps of the scale
development process.
Stage 1: Item generation
In item generation the primary concern is content validity
(Schreisheim, Powers, Scandura, Gardiner, & Lankau, 1993). For a
measure to possess content validity the measure should be capable of
capturing the domain of interest, yet exclude extraneous items. That is,
the measure should contain a range of items which are representative
of the behaviour the test was designed to sample (Cohen, Swerdlik &
Phillips, 1996). For example, an auditing examination would possess
content validity if the proportion and type of auditing questions in
the examination were similar to the proportion and type of problems
taught during the course.
Item development in the present study was undertaken in two
stages. First, using the literature review of chapter two, the constructs
of audit quality were assessed. A construct can be thought of as “an
informed, scientiic idea developed or‘constructed’ to describe or
explain behavior”(Cohen et al,1996,p.193). Second,having identiied
the constructs to be examined, items were developed on the basis of
the literature review,other researcher’s efforts to study audit quality and
service quality, and the author’s understanding of the audit profession.
The origin of items is shown within Appendix 1.
The constructs to be examined were as follows:
Constructs relating to prior work considering audit quality
•
•
•
Auditorreputation
Capabilityoftheauditor
Auditorindependence
DEVELOPMENT
•
•
•
OF
AUDITQUAL
73
Auditorexperiencewiththeclient
Auditorexpertise
Provisionofnon-auditservices
Constructs relating to prior work considering service quality
•
•
•
Empathy
Assurance (later relabelled Client Service)
Responsiveness
As well as making use of the prior literature described in chapter
two of this monograph, the questionnaire also utilises the analysis of
auditors’ external communications, reported in chapter four. The
research reported in chapter four is included in two ways. First, the
themeoftheirms’ employees/partners“addingvalue”, formspartof
theclientservicedimension. Second, thequalityoftheirms’ staff/
partners within the proposed dimensions of Reputation, Capability,
Expertise and Experience.
Stage 2: Scale development
Design of the developmental study
The instrument consisted of 56 items,with a number of additional
questions relating to background variables such as the respondents age,
gender and experience of auditing (see appendix). Participants were
asked to respond on the basis of their general views about auditing
(rather than their experiences of a single audit irm). Items were scored
onaive-pointLikertscale. Anitemrated“5” wasregardedas“very
important for audit quality”, while an item rated “1” was rated “not
importantforauditquality” –seeAppendix1forthedirectionsgiven
to respondents. To make the questionnaire easier for respondents to
74
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
complete, items were categorised and numerically ordered into three
types: auditirmfactors, engagementpartnerfactorsandauditteam
factors.
The questionnaire consisted of three sections:
Section A: The AUDITQUAL scale, subdivided into four categories
of items:
•
Auditirmfactors
•
Engagementpartnerfactors
•
Auditteamfactors
•
Otherfactors
Section B: An openended question inviting respondents to comment
on any aspect of the questionnaire or add comments pertaining to
audit quality
Section C: Background details, including the respondent’s job title,
age group, gender and work experience. See Appendix for further
details.
Scale validation
To assess the psychometric qualities of the scales, an analytic
technique called exploratory factor analysis (EFA) was employed.
EFA is essentially a data reduction technique which seeks to represent
a set of interrelated observed variables (ie in the present study the
56 items) in terms of a smaller number of latent variables (or factors).
This makes the task of interpreting the results more straightforward. In
the context of the present study, EFA can be thought of as a statistical
technique which groups together characteristics of audit quality and
service quality where there are similar patterns of response.
DEVELOPMENT
OF
AUDITQUAL
75
Researchers frequently use EFA methods when they have no clear
idea of how many underlying factors may exist. There are dangers
however, withsuchanapproach, asEFAwillsimplyindthebestlinear
combination of variables, based on their statistical interrelationship
(ie theircorrelations). Therefore, thefactorsidentiiedbyEFAmay
bear no reference to any underlying theory. The results of the EFA
are reported in Table A2 in Appendix three.
Another approach is to compare the factor solutions to the
hypothesised internal structure of the instrument (ie the nine constructs
ordimensionsofauditqualityandservicequalityidentiiedinthe
literature)andusetheresultsoftheEFAtoreinetheconstructs. That
is,theorydevelopsthehypothesisedconstructs;empiricaltestingreines
thesetheories. This‘reinement’ approachwastheoneemployedin
the present study.
Inthepresentstudy, EFAidentiiedsixfactors, removingthree
itemsfromtheoverallanalysisthatfailedtocorrelateasexpected–see
Table A2. Three of the six factors each consisted of related dimensions
(reputationandcapability; empathyandclientservice; expertiseand
experience). In conclusion, the EFA results provide some evidence
that the internal structure of the instrument (ie its construct validity)
is as hypothesised.
To test the appropriateness of the six audit quality/service quality
constructs that did not appear as single unidimensional constructs on
thebasisoftheEFA, internalconsistencyreliabilitycoeficientswere
calculated. Internal consistency reliability refers to the homogeneity of
items within the scale (ie the extent to which item responses correlate
withthetotaltestscore). Coeficientalphaisthemostcommonly
used measure of internal consistency reliability, with a minimum
value of 0.70 recommended for scales which are suitable for applied
research(Nunnally,1978). Alphacoeficientsforeachoftheninescales
ranged from 0.70 to 0.80, indicating each dimension yields scores of
76
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
satisfactoryinternalconsistencyreliability–seetableA.1inappendix
three for further details.
Asidentiiedearlierinthischapter, EFAmethodsareessentially
‘atheoretical’, and create a factor structure based on participants’
responses to individual items. An important concept in scientiic
researchisPopper’s(1959)conceptoffalsiication. Theconceptof
falsiicationisthatatheorycannotberegardedascredibleuntilithas
survived serious disconirmation efforts. In EFA,an expected structure
either emerges or does not (Thompson & Daniel, 1996 p.204). An
alternativetoEFAisconirmatoryfactoranalysis(CFA). CFAallows
theresearchertotestplausiblerivalmodelsandtoquantifytheitof
each model to the data:“testing rival models is usually essential because
multiplemodelsmayitthesamedata” (Thompson&Daniel, 1996
p.204). In the present study, the need for testing multiple models is
apparent. For example, are the nine hypothesised audit quality/service
quality dimensions unidimensional (ie independent of each other) or are
they better represented by only six factors (as suggested by the EFA)?
Furthermore, CFAprovidesthemeanstoassessmodelitacrossthe
threesamplesofauditors, companiesandexternalinancialstatement
users, and assess differences in the patterns of response across the three
samples. Appendix three provides further details of the multiple models
testedandthechoiceofgoodness-of-itstatistics. Theresultsofthe
competing CFA tests are shown in Table A.3 within Appendix three.
TheinalmodelisshowninFigure5.1:
DEVELOPMENT
OF
77
AUDITQUAL
Figure 5.1: The AUDITQUAL model
Reputation
Status
Capability
Independence
Technical
quality
Expertise
Knowledge
Experience
Responsiveness
Empathy
Client
service
Nonaudit
services
Understanding
Service
quality
78
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
The two factors labelled “Technical Quality”and “Service Quality”
are the two socalled‘higherorder’factors that determine audit quality
asdeinedbytheAUDITQUALmodel. TechnicalQualityisdeined
by ive lower-order factors (Reputation, Capability, Independence,
Expertise and Experience). Service Quality is described by four lower
order factors (Responsiveness, Empathy and Client Service, Non
audit Services). The AUDITQUAL model is a nested model, ie audit
quality is most adequately described by a hierarchical structure. The
composition of the AUDITQUAL model is described in Table 5.1.
Table 5.1: Nine first-order factors of AUDITQUAL scale
First-order factor
No.
items
Typical item
1. Reputation
8
“Theauditirmisobjective”
2. Capability
7
“The audit team staff are highly
competent”
3. Responsiveness
9
“Theauditirmiswillingtoprovide
detailed cost information”
4. Independence
5
“The audit fee controlled by the
engagement partner does not represent
more than 10% of the total fees
controlled by the engagement partner”
5. Nonaudit
services
4
6. Empathy
4
7. Client service
7
“Theauditirmconductsclientservice
reviews”
8. Expertise
6
“Theauditirmhasclientsinthesame
industry”
9. Experience
3
“Themanageroftheauditirmhas
been performing the audit for at least
two years”
“Theauditirmisabletosupplytax
services”
“The engagement partner is proactive
and contributory”
DEVELOPMENT
OF
AUDITQUAL
79
Table5.2reportsthecorrelationcoeficientsbetweenthenine
AUDITQUAL dimensions.. Although the results of the CFA support
the existence of two groupings of Technical Quality and Service
Quality, notable relationships exist between the nine AUDITQUAL
dimensions that make up these scales. Within the Technical Quality
factor, the dimensions of Reputation, Capability and Expertise are
all found to be highly related to each other (r ranging from .38 to
.63). However, Reputation, Capability and Expertise are all positively
associated with the Client Service dimension from the Service Quality
grouping. Similarly, Responsiveness and Empathy from the Service
Quality grouping are both positively related to the Provision of Non
audit Services. Therefore, although Technical Quality and Service
Quality emerge as distinct constructs, it is important to understand
signiicantrelationshipsexistbetweenthosedimensionsthatmakeup
these constructs.
80
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Table 5.2: Correlation coefficients between nine AUDITQUAL
dimensions
I
Technical Quality
I.
Reputation
II. Capability
III. Independence
IV. Expertise
V. Experience
Service Quality
VI. Nonaudit
services
VII. Responsiveness
VIII. Empathy
IX. Client service
II
III
IV
V
VI
.63
.13
.38
.23
.05
.41
.19
.06
.26
.07
.08
.08
.03
.05
.07
.12
.15
.44
.04
.20
.54
.16
.02
.07
.14
.26
.58
.14
.19
.31
.33
.29
.15
VII
VIII
IX
-.51
.32
.61
Figures in italics,statisticallysigniicantatp<0.01
Stage 3: Scale evaluation
Theirsttwostagesofdevelopinganewmeasurefocusesonthe
reliability (internal consistency) and validity (content and construct) of
scoresproducedbythenewmeasure. Modelitstatisticsarepresented
in appendix three, table A.3.
Construct validity can be further demonstrated by relating scores
on the new measure to some other existing measure with known
psychometric qualities. This is said to provide evidence of criterion
validity and is demonstrated by correlating scores on the two measures.
A second method is assessing the patterns of responses to the instrument
across two or more groups who could be expected to differ on the
measure (eg auditors, inance directors, and external inancial statement
users). The present study uses the second approach, which is described
in the following chapter.
DEVELOPMENT
OF
AUDITQUAL
81
Key points
This chapter has described the methods (internal consistency
reliability analysis, EFA and CFA) used to construct and validate the
AUDITQUAL instrument. The importance of scale development
methods has been emphasised. Before the results of the administration
of any questionnaire can be taken seriously the instrument must yield
scores of adequate measurement properties. The resulting measure
–showninFigure5.1-producesscoreswithsatisfactorypsychometric
qualitieswhenappliedtosamplesofauditors, inancedirectorsand
external inancial statement users. As such,it is possible to be conident
that the AUDITQUAL instrument is suitable for applied research and
correlational studies that may inluence policy development within
theaccountingprofession. TheinalAUDITQUALscaleidentiies
two constructs that describe this conception of audit quality: technical
qualityandservicequality. Technicalqualityisdeinedbyirst-order
factors of Reputation, Capability, Independence, Nonaudit Services,
ExpertiseandExperience. Servicequalityisdescribedbytheirstorder factors of Responsiveness, Empathy and Client Service.
ChApter six
ConduCt And AnAlysis of survey
The purpose of this chapter is to irst,describe the administration of the
survey; second, describethedemographicsofrespondentsbysample;
andthird, reporthowthethreegroupsofauditors, inancedirectors
and fund managers responded to the AUDITQUAL instrument. Each
of these two elements is examined in turn.
Conduct of survey and demographics of respondents by
sample
Accounting irm partners
The questionnaire was mailed to 500 partners from the 20 largest
irmsintheUK, asdeinedbytheAccountancy Age 2001 survey. The
partners were drawn at random from the Institute of Chartered
Accountants in England andWales Member Firms’Directory 2001. All
questionnaires were completed anonymously. 109 useable responses
were returned, resulting in a response rate of 21.8%. Five individuals
indicated they did not wish to participate due to a lack of management
time,and two said the addressee had retired. Although the response rate
may appear relatively low,when the questionnaire length,the somewhat
unorthodoxnatureofthesubjectwithintheaccountingieldandthe
fact that marketingrelated studies usually achieve a response rate of
between 10 and 30% (Green,Tull, & Albaum, 1988), the response rate
is acceptable for a postal survey of this nature. 59 respondents were
employedinBig4/5irms,44respondentswereemployedinmedium-
84
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
sizedirms, sixdeclinedtorespond. AsummaryoftheAUDITQUAL
responses by age and gender is set out in Table 6.1:
Table 6.1: Demographics of AEP sample
Age group
Female
Male
Under 35 years
35–45years
1
4
3
37
4
41
(3.7%)
(37.6%)
Over 45 years
1
56
57
(52.3%)
7
(6.4%)
Not known
Total
Total
7
6
96
109 (100.0%)
Table 6.2 profiles the work experience of the participants.
Participants have signiicant amounts of work experience (postqualiication experience mean 21.8 years). The majority of this
experiencehasbeengainedwithinthesameirm.
Table 6.2: Work experience of accounting firm sample
Mean
(years)
Standard
deviation
Servicewithpresentirm
20.2
10.0
No.yearspost-qualiicationexperience
21.8
8.42
Totestforresponsebiasthesamplewasdividedintoearly(irst36)
and late (last 36) respondents. This assumes late respondents are similar
to nonrespondents (Oppenheim, 1966). Early and later respondents
were compared in their response to the nine AUDITQUAL dimensions
using the MannWhitney nonparametric test. No statistically
signiicantdifferenceswerefoundacrosstheninedimensionsbetween
the two groups (α = .05).
CONDUCT
AND
ANALYSIS
OF
SURVEY
85
Finance directors
The questionnaire was mailed to 500 inance directors (FDs)
chosen at random from 1831 listed UK companies, as reported in the
Thompson Financial Global Access company database (December
2001). All questionnaires were completed anonymously. 75 useable
responses were returned, resulting in a response rate of 15.0%. Seven
companies indicated it was company policy not to respond to surveys,
seven said the addressee had left the company.
Totestforresponsebiasthesamplewasdividedintoearly(irst
25) and late (last 25) respondents. Early and later respondents were
compared in their response to the nine AUDITQUAL dimensions using
theMann-Whitneynon-parametrictest. Nostatisticallysigniicant
differences across the nine dimensions were found between the two
groups (α = .05).
Twenty-iverespondentswereemployedinlargecompanies(assets
>£100 million), 19 respondents were employed in medium-sized
companies (assets £30 million £100 million), 30 respondents were
employed in small companies (assets <£30 million), one declined
to respond. A summary of the AUDITQUAL responses by age and
gender is set out in Table 6.3:
Table 6.3: Demographics of finance director sample
Age group
Female
Male
Total
Under 35 years
1
12
13
(17.3%)
35–45years
2
24
26
(34.7%)
Over 45 years
2
34
36
(48.0%)
Total
5
70
75
(100.0%)
86
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Similartothesampleofaccountingirmpartners, theFDsample
is largely male. An analysis of the FD sample suggests they are rather
younger than their colleagues in accounting irms with 17.3% of
the sample being under 35 years of age (compared to 3.7% of the
accountingirmsample). Themajorityofrespondents(84.0%)had
previously worked as external auditor.
Other descriptive information concerning the companies of the
FD sample is shown in Table 6.4. It can be seen the sample comes
from a range of industries, and an even mix of company size, as assessed
byassetsize. Alargemajorityofcompanies(92.0%)sampledhadan
audit committee. Furthermore, 41.3% of respondents indicated their
company held a recognised quality assurance standard (eg ISO9000,
EFQM). Company’s which hold quality assurance standards are likely
to see quality as a source of competitive advantage. Therefore, FDs
of companies holding quality assurance standards may have different
perceptions of the dimensions of audit quality to those in companies
not holding such awards.
CONDUCT
AND
ANALYSIS
OF
SURVEY
87
Table 6.4: Information relating to the company
Industry
Capital goods
N= (%)
4
(5.3%)
Consumer goods
14
(18.7%)
Financial
15
(20.0%)
Other
42
(56.0%)
Total
75 (100.0%)
Company size
N= (%)
Small(assets<£30million)
30
(40.0%)
Medium (assets £30£100 million)
19
(25.3%)
Large (assets >£100 million)
26
(34.7%)
Total
75 (100.0%)
Yes
No
Total
Existence of an audit
committee?
69
(92.0%)
6
(8%)
75
(100.0%)
Possession of a recognised
quality assurance standard?
31
(41.3%)
44
(58.7%)
75
(100.0%)
Table 6.5 reports information concerning the characteristics of
the auditor’s employed by the sample of company FDs. Big 4/5 audit
irmsrepresentthemajorityofauditirmsemployedbythesample
(81.3%), relectingthemarketconcentrationofBig4/5auditirms
amongst listed companies. Interestingly, a large number of companies
have employed their current auditors for a period of less than ive
years (37.3%). Only 25.3% of respondents’ companies had employed
88
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
their current auditors for a period in excess of ten years. Finally, a
signiicantproportionofthesamplereportmakinguseofnon-audit
services from their current auditor with 37.4% reporting nonaudit
services accounted for over 50% of the total fees paid to the irm
during the year.
Table 6.5: Information relating to the auditors employed by the
company
The size of current auditor
Big 4/5
NonBig 4/5
Total
61
14
75
(81.3%)
(18.7%)
(100.0%)
Auditor tenure
<5years
5–10years
28
26
(37.3%)
(34.7%)
> 10 years
19
(25.3%)
Don’t know
Total
2
75
(2.7%)
(100.0%)
0%
<25%
3
22
(4.0%)
(29.3%)
2550%
22
(29.3%)
> 50%
28
(37.4%)
Total
75
(100.0%)
Proportion of total audit fees paid to
audit firm in respect of non-audit services
CONDUCT
AND
ANALYSIS
OF
SURVEY
89
UK fund managers
The questionnaire was mailed to 500 fund managers (FMs) chosen
at random from 1550 fund managers based in the UK, as reported
in the Caritas Data European Fund Managers 2002 database. All
questionnaires were completed anonymously. 76 useable responses
were returned, resulting in a response rate of 15.2%. Two fund
managers indicated it was policy not to respond to surveys. To test
forresponsebiasthesamplewasdividedintoearly(irst25)andlate
(last 25) respondents. Early and later respondents were compared in
their response to the nine AUDITQUAL dimensions using the Mann
Whitneynon-parametrictest. Nostatisticallysigniicantdifferences
across any of the nine dimensions were found across the two groups
(α = .05). A summary of the AUDITQUAL responses by age and
gender is set out in Table 6.6.
Table 6.6: Demographics of Fund Manager sample
Age group
Under 35 years
35–45years
Over 45 years
Total
Missing
Female
6
0
0
6
Male
8
26
30
64
Total
14
26
30
70
(18.4%)
(34.2%)
(39.5%)
(92.1%)
6
(7.9%)
Similar to the sample of accounting irm partners and inance
directors, the sample is largely male. 18.4% of the FM sample are
aged under 35 years of age, as opposed to 17.3% of the FD sample,
and only 3.7% of the accounting irm sample. The majority of
respondents(51.0%)reportedtheyhadsigniicantexperienceofthe
audit profession, having previously worked as an auditor, with a further
37.0% considering they possessed “an understanding of the audit
process”. Only 12.0% reported having minimal or no experience of
90
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
the audit profession. The majority of FMs spend little time on audit
activities. 39.5% of the FM sample reported spending no time on
auditrelated activities, with a further 47.4% reporting spending under
25% of their time on auditrelated activities. In summary, the FM
samplecanbeconsideredasrelativelysophisticatedusersofinancial
information, possessing either experience of working as an auditor
or an understanding of the audit process. However, relatively little of
their work time is spent on audit activities suggesting their current
daytoday contact with the profession is slight.
Responses to survey by sample group
Accounting irm partners
Responses to the AUDITQUAL section of the questionnaire from
thesampleemployedinaccountingirmsareshowninTable6.7. The
highest mean scores are for the two factors representing Reputation
and Capability, which are combined within the AUDITQUAL
model shown in Figure 5.1 to create a higherorder factor labelled
‘Status’. There is a high level of consensus of the relative importance
of both Reputation and Capability characteristics. The four
remaining technical quality dimensions (Independence, Nonaudit
services, Expertise and Experience) were regarded as less important
than Reputation and Capability and at a similar level to the three
service quality dimensions. The relatively low mean score and lack
of consensus for the Independence dimension is perhaps surprising,
as independence represents one test of audit quality (the other being
competence) from positive accounting theory. Much of this result
can be explained by responses to items concerning the audit fee paid
bytheclient, withasizeableproportionofauditorsconsideringthe
proportion of fees controlled by an engagement partner as being
relatively unimportant.
CONDUCT
AND
ANALYSIS
OF
SURVEY
91
Table 6.7: Auditors’ response to AUDITQUAL section of
questionnaire
Non-Big 4/5
partners
Big 4/5 partners
Mean
score1
Level of
consensus2
Mean
score1
Level of
consensus2
Technical quality
1. Reputation
4.53
Medium
4.50
High
2. Capability
4.65
High
4.53
High
3. Independence
2.64
Low
3.25
Medium
4. Expertise
3.90
Medium
3.57
Medium
5. Experience
3.08
Low
3.15
Low
6. Nonaudit services
2.55
Low
2.67
Low
7. Responsiveness
2.95
Medium
3.23
Medium
8. Empathy
3.61
Medium
3.60
Medium
9. Client service
3.98
Medium
3.79
High
Service quality
Notes to table:
1. Response scale is 5 (very important), 4 (important), 3 (fairly important),
2 (little importance), 1 (not important at all).
2. Levelofconsensusisclassiiedbaseduponthestandarddeviation(SD)
of responses: SD < 0.50(high); 0.50<SD< 0.8(medium); SD>0.8
(low).
Somescholarshavearguedauditqualityisrelatedtoirmsize
(eg DeAngelo,1981) as outlined in chapters two and three. Consequently,
it is worth comparing scores on the nine AUDITQUAL dimensions
92
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
across the two groups of Big 4/5 partners and NonBig 4/5 partners.
Examiningthemeanscoresacrossthetwosamplesindsbothgroups
rate Reputation and Capability as the most important determinants of
audit quality. However a number of differences in mean scores can be
observed across the two groups. To determine whether the responses
of the two groups are different, the statistical technique of discriminant
analysis is employed. Discriminant analysis is a means of distinguishing
between two or more groups, based on a collection of variables (in this
case,AUDITQUAL dimensions). The concept underlying discriminant
analysis is straightforward: linear combinations of the independent
variables (the AUDITQUAL dimensions) are formed and serve as the
basis for classifying cases into one of the groups (ie Big4/5irmor
Non-Big4/5irm). Discriminantanalysisisamulti-variatestatistical
technique, which provides a simple means of differentiating between
the two groups. This is preferable to other (univariate) analytic
techniques which would test for differences between the two groups
across all the dimensions individually. Such multiple testing creates the
possibilityofstatisticallysigniicantdifferencesarisingbychancealone.
Table6.8displaysthediscriminantfunctioncoeficientsalongwith
the Wilks’ lambda statistic. Wilks’ lambda can take on values between
0 and 1. Large values of lambda (ie close to 1) indicate the groups
are not different, whilst small values indicate the group means do
appeartobedifferent. Wilks’ lambdaisameasureofeffectsize, ie the
magnitude of the difference between the two groups. Furthermore,
thestatisticalsigniicanceoftheresult(p) can also be computed, to
provide an indication of whether the difference calculated for the
samples, as compared to the populations from which it is drawn can be
attributed to mere chance. Thus the interpretation of these two results
necessarily goes hand in hand. That is, the research is concerned with
anassessmentofpracticalsigniicance, aswellasstatisticalsigniicance
(see for example,Kirk,1996). The Wilks’lambda value of 0.767 suggests
that the groups are different in some respects.
CONDUCT
Table 6.8:
AND
ANALYSIS
OF
SURVEY
93
Discriminant function coefficients – Big 4/5 versus
Non-Big 4/5 sample
Technical quality
Standardised discriminant function
coefficients
Reputation
.298
Capability
.043
Expertise
.611
Experience
.018
Independence
.530
Service quality
Nonaudit services
Responsiveness
.094
.651
Empathy
.176
Client Service
.313
Wilks’lambda=0.751; χ2=27.309;p<.001
Interpretingdiscriminantcoeficientsrequiressomeexplanation.
First, the coefficients are standardised so the magnitude of the
coeficientsarecomparable. Forexample,scoresoftheResponsiveness
dimension with nine items (and hence a maximum score of 45) are
standardised to make them comparable with scores on the Empathy
dimension with only four items (and hence a maximum score of only
20). However, the values of the standardised discriminant function
coeficientsarealsoaffectedbythecorrelationsofotherindependent
variables–aproblemknownascollinearity. Therefore, thecoeficients
do not in any absolute sense relect the importance of the various
independent variables.
94
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Ananalysisofthestandardiseddiscriminantfunctioncoeficients
suggests that responses to the AUDITQUAL dimensions of Expertise,
Responsiveness and Independence explain most of this difference.
Table5.2intheprecedingchapter,identiiesExpertise,Responsiveness
and Independence are relatively uncorrelated (r ranging from .06 to .16)
and are consequently less likely to suffer from the previously described
problems of collinearity. NonBig 4/5 partners rate Independence
and Responsiveness higher than Big 4/5 partners. Big 4/5 partners
rate Expertise higher than NonBig 4/5 partners. It is likely NonBig
4/5 partners are likely to perceive Responsiveness to clients’ needs as a
means of gaining competitive advantage. Big 4/5 partners, working in
irms with considerable technical resources are likely to value Expertise
higherthantheircolleaguesinNon-Big4/5irms. Thedifference
across the two groups in mean scores on the Independence dimension
is less easy to explain. However, the low level of consensus particularly
in the Big 4/5 partner group may explain the result. An analysis at
item level, reveals a wide distribution of responses to questions relating
toanappropriateupperlimitforaspeciicclient’sauditfeeinrelation
to the overall level of fees controlled by the audit engagement partner
responsible. As one respondent noted:
(The) audit fee relationship to firm’s income (say 10%) is more
important than individuals in the firm
(Big 4/5 audit engagement partner)
Another respondent perceived independence as rather less an
issue:
Independence is an issue as it relates to other services is overplayed.
There is not one instance in the UK where audit failure can be put
down to a lack of objectivity caused by the lack of independence as
opposed to rank bad work or negligence.
(Big 4/5 audit engagement partner)
CONDUCT
AND
ANALYSIS
OF
SURVEY
95
Accountingirmpartnersarenotaloneinstrugglingwiththe
concept of independence. Van Der Plaats (2000, p.625) suggests
“auditor independence is one of the most thorny and political issues for
a regulator to deal with”. When auditors’ responses are considered in
the light of the discussion concerning independence in fact in chapter
two, it is unsurprising there is marked variation amongst auditors on
the concept of independence.
Finance directors
The pattern of responses of the FD sample reported in Table 6.9
is broadly similar to that of the accounting irm partner sample.
Reputation and Capability receive the highest ratings and achieve
a high level of consensus amongst participants. Again, similar to the
accounting irm partner sample, three of the four service quality
dimensions receive a similar ratings with mean scores varying from
3.49 to 3.61, with the exception of Nonaudit services which is
regarded as less important (mean = 2.44) and achieves only a low level
of consensus.
96
AUDITQUAL: DIMENSIONS
Table 6.9:
OF
AUDIT QUALITY
Finance directors’ response to AUDITQUAL section of
questionnaire
Summary statistics from
AUDITQUAL analysis
Mean score1
Level of
consensus2
Technical quality
1. Reputation
4.39
High
2. Capability
4.35
High
3. Independence
2.92
Low
4. Expertise
3.53
Medium
5. Experience
3.05
Medium
6. Nonaudit services
2.44
Low
7. Responsiveness
3.49
Medium
8. Empathy
3.57
Medium
9. Client service
3.61
Medium
Service quality
Notes to table:
1. Response scale is 5 (very important), 4 (important), 3 (fairly important),
2 (little importance), 1 (not important at all).
2. Levelofconsensusisclassiiedbaseduponthestandarddeviation(SD)
of responses: SD < 0.50(high); 0.50<SD< 0.8(medium); SD>0.8
(low).
It is also of interest to examine the relationship between some
of the background and demographic variables of the FD respondents.
Table6.10reportsthecorrelationcoeficients(ie degree of association
orrelationship)betweenthevariablesofcompanysize, auditorsize,
auditor tenure, the length of service of the FD, and the age group of
CONDUCT
AND
ANALYSIS
OF
SURVEY
97
the FD. A number of large relationships are observed. Some large and
statisticallysigniicantrelationshipsareobservedandyettrivial(eg the
age of an FD is positively related to their length of service). Similarly,
auditor size is positively associated with company size. However,
two relationships are observed which are not intuitive. First, auditor
tenureispositivelyrelatedtocompanysize(r = .35, implying 12% of
auditortenurecanbeexplainedbycompanysize). Second, auditor
tenure is positively associated with the age group of the FD, implying
the relationship between the FD and the auditor is a personal one.
The importance of the FDAEP relationship has been stressed in a
number of previous investigations (eg Beattie&Fearnley, 1995; Beattie
et al, 1998, 2001; Fellingham&Newman, 1985; Fisher, Schatzberg&
Shapiro, 1996).
Table 6.10: Correlation coefficients between selected background
variables
Company
size
Auditor
size
Tenure
Auditorsize
.32 *
Tenure
.35 *
.10
FD length of
service
.39 *
.10
.47 *
Age of FD
.10
.01
.37 *
FD length
of service
.45 *
* p<0.01
Table 6.11 reports correlation coeficients calculated between
the nine AUDITQUAL dimensions and the background variables of
companysize, auditorsize, auditortenure, FDlengthofservicewith
company,age group of the FD and the existence of an audit committee.
Althoughthemagnitudeofthemajorityofcorrelationcoeficients
98
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
are small, indicating scores on AUDITQUAL are largely unrelated
to the background variables, a number of important relationships are
observed. First, companysizeisrelatedtoscoresontheReputation
dimension. This relects the concentration of Big Five irms within the
large company audit market and the association between company size
andauditorsize(r = .32) in the present investigation. The background
variable which reports the greatest relationship with AUDITQUAL
dimension scores is the age of the FD. Age is positively related to
scores on Reputation (r = .28) and Experience (r = .21),and negatively
associated with Expertise (r = .22). To establish whether practical
and statistically signiicant differences exist between different age
groups, discriminant analysis is performed. Because a relatively small
proportion of the FDs sampled in the study were aged under 35 years,
the under 35 years group were combined with the 35 to 45 years of
age group. This created two age groups of under 45 years and over
45years, eachofapproximatelysimilarsamplesize. Theresultsofthe
discriminant analysis are shown in Table 6.12.
CONDUCT
AND
ANALYSIS
OF
SURVEY
99
Table 6.11: Correlation coefficients:AUDITQUAL dimensions and
selected background variables – FD sample
FD
length
of
service
Company
size
Auditor
size
1. Reputation
.32**
.13
.02
.15
.28*
2. Capability
.16
.08
.05
.05
.14
3. Independence
4. Expertise
.03
.18
.01
.03
.09
.04
.20
.04
.18
.22
5. Experience
.05
.07
.05
.05
.21
.15
.04
.05
.04
.08
.02
.07
.01
.10
.04
8. Empathy
.03
.11
.10
.07
.08
9. Client service
.10
.05
.21
.07
.01
Tenure
Age
of FD
Technical quality
Service quality
6. Nonaudit
services
7. Responsiveness
* p<0.05
** p<0.01
100
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Table 6.12: Discriminant function coefficients – FDs, under
45 years versus over 45 years
Standardised
Discriminant function
coefficients
Technical quality
Reputation
.264
Capability
.100
Expertise
.833
Experience
.563
Independence
.484
Service quality
Nonaudit services
.342
Responsiveness
.056
Empathy
Client Service
.414
.264
Wilks’lambda=.754; χ2 = 19.10, p = .02.
The value of Wilks’ lambda calculated suggests the groups are
differentinsomerespects. Thetestofstatisticalsigniicanceofthe
observed Wilks’ lambda value indicates the means of the discriminant
functions are not equal across the two populations of FDs aged
under 45 years and aged over 45 years. Generally, older FDs rated
AUDITQUAL dimensions higher than their younger colleagues, the
exceptions being: Expertise, Nonaudit services and Client Service,
which were regarded as more important by those FDs aged under
45years. Thelargeststandardiseddiscriminantfunctioncoeficients
values occur for Experience, Expertise, Independence and Empathy.
CONDUCT
AND
ANALYSIS
OF
SURVEY
101
The indings of the present investigation are comparable to those
ofMoizer(1998)whoexaminedUKcompanydirectors’ viewson
theperformanceofthe(then)BigEightaccountingirmsin1987
and Big Six accounting irms in 1996. Moizer (1998) measured
perceptions of performance using a composite measure of performance,
including aspects of technical and service quality relating the aggregate
performance score to 11 independent variables in a multivariate
regression equation. The multivariate regression procedure is
mathematically similar to the discriminant analysis used in the present
study. ThemajorexplanatoryvariableinMoizer’s(1998)analysiswas
age, witholdercompanydirectorsbeingmorelikelytobesatisied
withtheperformanceoftheirolderirmthantheiryoungercolleagues.
Moizer’sanalysisdrawsontheworkofHunter&Sundel(1989)to
identify men in their forties are more likely to visualise the socalled
big picture, have enhanced judgement skills, and that men of middle
agearemostlikelyofalltheagegroupstobesatisiedwiththeirjobs
(Tamir, 1989).
UK fund managers
The pattern of responses of the FM sample is broadly similar to
thatofboththeaccountingirmpartnersampleandthesampleof
FDs. Reputation and capability receive the highest ratings and achieve
ahighlevelofconsensusamongstparticipants–seeTable6.13. The
importance of the provision of nonaudit services receives the lowest
rating (2.14), a result similar to the analysis of results for the accounting
irmpartnerandFDsamples.
102
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Table 6.13: Fund managers’ response to AUDITQUAL section of
questionnaire
Summary statistics from
AUDITQUAL analysis
Mean score1
Level of consensus2
Technical quality
1. Reputation
4.56
High
2. Capability
3. Independence
4.45
3.36
High
Medium
4. Expertise
3.82
Medium
5. Experience
3.21
Medium
Service quality
6. Nonaudit services
2.14
Low
7. Responsiveness
3.13
Medium
8. Empathy
3.35
Medium
9. Client service
3.75
High
Notes to table:
1. Response scale is 5 (very important), 4 (important), 3 (fairly important), 2
(little importance), 1 (not important at all).
2. Level of consensus is classiied based upon the standard deviation (SD)
of responses: SD < 0.50(high); 0.50<SD< 0.8(medium); SD>0.8
(low).
Table 6.14 examines the association between the background
variables of the proportion of time spent on audit activities, FMs
experience of the audit process, the length of service of the FM with
their current employer,the age group of the FM,and scores on the nine
CONDUCT
AND
ANALYSIS
OF
SURVEY
103
AUDITQUAL dimensions. Although the majority of the correlation
coeficientsareoflowmagnitude, anumberofinterestingrelationships
are found. Those fund managers who spend proportionately more
time on audit activities rate Expertise (r = .20), Empathy (r = .29) and
Client Service (r = .45) as more important than FMs who spend less
time on audit activities. This result may however be attributable to
FMs assessment of their own auditor, rather than auditors of companies
they consider for inclusion in their own portfolio.
Anotherunexpectedindingisthatexperienceoftheauditprocess
is negatively related to both Reputation (r = .29) and Independence
(r =-.21). ThisindingsuggeststhatthoseFMswhohaveworked
as auditors place less emphasis on the need for a company to be
auditedbyaBigFourirm, andhaveadifferentviewoftheconcept
of independence to those with little audit experience.
Finally, the age group of the FM is related to scores on two
AUDITQUAL dimensions: Expertise (r = .29) and Experience
(r = .40), suggesting older FMs either place less emphasis on these
dimensions or take them for granted.
104
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Table 6.14: Correlation coefficients - AUDITQUAL dimensions and
selected background variables
% time
spent
on audit
activities1
Experience
of audit
process2
FM
length of
service3
Age
group
of FM4
Technical quality
1. Reputation
.06
.29 *
2. Capability
.14
3. Independence
.12
.01
.02
.13
.11
.02
.21
.15
.01
.29
*
.00
.40
**
.03
.01
.05
4. Expertise
.20
.20
.29
5. Experience
.14
.05
6. Nonaudit
services
.04
7. Responsiveness
.09
*
Service quality
8. Empathy
9. Client service
Key:
.19
.02
.09
*
.05
.01
.01
**
.12
.09
.16
.29
.45
1
1=0% to 4=75100%.
1=Minimalto3=Signiicant.
3
Continuous.
4
1=<35yearsto3>45years.
* p<0.05
** p<0.01
2
Differences across the three sampled groups
An important research question is how do three sampled groups of
auditors, FDs and FMs differ in their attitudes to audit quality. Mean
scores across the three groups are shown in Table 6.15.
CONDUCT
AND
ANALYSIS
OF
SURVEY
105
Table 6.15: Fund managers’ response to AUDITQUAL section of
questionnaire
Auditors
FDs
FMs
1. Reputation
4.51
4.39
4.56
2. Capability
4.61
4.35
4.45
3. Independence
2.90
2.92
3.36
4. Expertise
3.76
3.53
3.82
5. Experience
3.12
3.05
3.21
6. Nonaudit services
2.51
2.44
2.14
7. Responsiveness
3.06
3.49
3.13
8. Empathy
3.61
3.57
3.35
9. Client service
3.90
3.61
3.75
Technical quality
Service quality
In general terms, auditor and FDs mean scores across the nine
dimensions are broadly similar, with auditors tending to rate technical
quality attributes higher than FDs, and FDs rating service quality
attributes higher than auditors.
However, the most marked difference is the users group of FMs
who rate technical qualities higher than service quality attributes,
thaneithertheFDorauditorsample. Thisisanexpectedinding, as
external users of accounting information are less likely to be interested
in service quality issues between the auditor and their client. This
inding is similar to the results ofWarming-Rasmussen & Jensen’s
(2001) study conducted in Denmark, where Danish external users
(shareholdersandinancialjournalists)ratedauditquality(ie technical
106
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
quality attributes in the present investigation) higher than auditors or
managing directors.
To test whether these differences are practically and statistically
signiicant across the three groups, discriminant analysis is used to
distinguish between members of the three groups. This technique is
similar to the twogroup analysis performed earlier, with the difference
that it computes two functions, which are uncorrelated with each
other. The results of the three group discriminant analysis are shown in
Table 6.16,and suggest the pattern of scores on the nine AUDITQUAL
dimensions across the three samples are both practically and statistically
signiicant.
CONDUCT
AND
ANALYSIS
OF
SURVEY
107
Table 6.16: Discriminant function coefficients – Auditors, FDs and
FMs
Standardised Discriminant function
coefficients
Technical quality
Reputation
Function 1
Function 2
.266
.481
.522
.006
.088
.047
.623
.400
.011
.586
.479
.441
1.015
.218
Empathy
.084
.369
Client service
.513
.230
Capability
Expertise
Experience
Independence
Service quality
Nonaudit services
Responsiveness
Function1through2:Wilks’lambda=.644; χ2 = 95.138, p<.0005
Function2: Wilks’lambda=.880; χ2 = 27.539, p<.001
Key points
All three groups rate the technical quality dimensions of
Reputation and Capability the highest. The dimension rated least
important is the service quality dimension considering the provision
of Nonaudit services. Auditors and FDs were broadly similar in
their pattern of response to the inventory, with FDs tending to rate
service quality attributes higher than the accounting irm partners, and
accountingirmpartnersratingtechnicalqualityissueshigherthan
FDs. FMs tending to rate technical issues higher than either auditors
andFDsandservicequalityissueslessthanaccountingirmpartnersor
108
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
FDs. Theseresultswereconirmedbyadiscriminantanalysisapplied
across the three samples.
Big 4/5 partners rate Expertise higher than nonBig 4/5 partners.
NonBig 4/5 partners rate Responsiveness and Independence as being
more important than their Big 4/5 colleagues.
Auditor tenure is positively associated with the age of the FD,
reconirmingtheimportanceofthepersonalrelationshipbetweenthe
FD and audit engagement partner (eg Beattie et al, 2001; DeRuyter
&Wetzels, 1999; Moizer, 1998). Ingeneralolder(ie over 45 years of
age) FDs tended to rate AUDITQUAL dimensions higher than their
younger counterparts.
Model it is also excellent for the three samples. That is, the
questionnaire is equally applicable across the three populations of
auditors,inancedirectorsandfundmanagers.
ChApter seven
disCussion And impliCAtions
Introduction
Assurance services guru Robert Elliott (Elliott, 1995 p.119) argues that
the real customers for audit services (eg stakeholder groups) have“quite
avarietyofunfulilledneedsinusinginformationfordecisionmaking”.
The AICPA’s Committee on Assurance Services (1997) recognised the
necessity of the customer (ie stakeholder)comesirstapproachand
the desirability of offering expanded services to decision makers. To
conclude this monograph, this chapter considers how audit quality
mightinluenceirms’strategicdecisionmakinginthefuture.
Earlierchaptershaveidentiiedauditquality(bothserviceand
technical aspects) can be a source of competitive advantage for
accountingirms. Toachievecompetitiveadvantageanorganisation
needs to pursue a strategy of differentiation or cost leadership. An
important conclusion is that it’s not just OK to be different,it is essential
to sustain competitive advantage. Doing the same things better achieves
operational effectiveness, but as competitors are also trying to be better,
competitive advantage does not occur (Bachmann, 2002 p.61).
Implications arising from the analysis of AUDITQUAL
Perceptions of audit quality (as measured by AUDITQUAL)
across the three sampled groups of auditors, auditees and external users
were relatively homogenous. This conclusion is also supported by the
indingthattheAUDITQUALmodelittedthedatawellforeachof
110
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
thethreesamples. However, accountingirmpartnerstendedtorate
technicalaspectsofauditqualityhigherthanthesampleofinance
directors, andtheinancedirectorsrateservicequalityhigherthanthe
accountingirmpartners. Whilethisindingmaybereassuringfor
external users of accounting information and regulators, auditors may
wishtoconsidertheimportanceinancedirectorsplaceonservice
aspects of the client relationship. Unsurprisingly,fund managers tended
to rate technical quality higher than service quality, presumably as
service quality represents an unobserved variable to external users of
accounting information.
Technical quality consists of ive scales: reputation, capability,
expertise, experience, and independence. Service quality consists of
the four scales labelled: nonaudit services, responsiveness, empathy,
and client service. Although the existence of these two higherorder
factorsisconirmedbythetestingofmultiplemodels, itisimportant
to note that the scales labelled independence and nonaudit services
are relatively distinct from both technical quality and service quality,
and from each other.
For each of the three samples, reputation and capability were
the scales, which received the highest ratings. That the reputation of
theirmandthecapabilitiesofitspartnersandstaffisunsurprising.
However, the scale rated as least important by all three groups was the
provisionofnon-auditservices. Thisindingisatoddswiththeanalysis
ofirms’ externalcommunications, whereirmstendedtoemphasise
the range of nonaudit services they could provide as means of
differentiatingthemselvesfromcompetitors. Furthermore, irmshave
promotedvalue-addingauditingtoassisttheriskmanagementproile
of the client. As auditors also promote the importance of providing
consultingservicesnecessarytocreatetheclientspeciicknowledge
(ie economies of scope) to audit complex international businesses, the
low rating of nonaudit services conlicts with both economic theory
and professional practice. Perhaps rather than simply eulogise over the
DISCUSSION
AND IMPLICATIONS
111
scope, uniquenessandtechnologiesoftheirnon-auditservices, irms
might promote the role nonaudit services play in audit quality.
Implications for educators and those responsible for the
professional development of accountants
Audit quality is very much dependent on the quality of people. If
auditors are not also permitted to provide non-audit services, the
career of an auditor will be such that good quality people do not join
the audit profession.
(Big 4/5 partner)
As the present study has stressed the role of both technical
and service quality in audit quality, accounting irms will need to
attract, retain and develop staff of the highest calibre. Firms’ external
communications described in chapter six, emphasises the importance
of staff in the provision of quality professional services. Perhaps then
themostobviousindingofthisworkforeducatorsisthatthesocalled‘goodqualitypeople’ thatirmswishtoattractandretainshould
have the necessary skills, both technical and interpersonal, to deliver
technical quality to stakeholders (and clients) whilst providing the best
possible service to clients.
Thepersonalityproileexhibitedbythemajorityofpractising
accountants (ie ISTJ using the MBTI framework) suggests they are
individuals who are likely to possess high technical skills, but less likely
to have the social skills necessary to establish and maintain longterm
relationships. In particular, ISTJ may be perceived as being reserved,
passive, cold, andevenselishindividuals. Giventheimportanceof
individualandteamskillsinworkingwithclients, accountingirms
may wish to consider how they screen individuals prior to employment
in their irm. De Ruyter &Wetzels (1999 p.72) suggest auditors
should view themselves as “salespeople representing complex bundles
112
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
of services and view themselves as relationship managers”. Such a
screening process, perhaps using psychometric testing and assessment
centres, might identify people who possess warmer and more outgoing
personalityproiles.
Afurtherconsiderationiswhetherairm’sdesiretomaximise
client satisfaction and encourage client centeredness (Thornbury,1999)
could encourage accommodating the client or doing what the client
wishes without question. Educators and staff partners in accounting
irmsinterestedindevelopingsystemstoimproveauditqualityshould
ensure trainees, and more senior staff are fully aware of who the ‘real’
client is (ie the stakeholder) at all times.
Recommendations
1. It is important that audit quality should be seen as a multi-dimensional
construct
Studies which examine audit quality purely from the perception
of technical competence and independence are missing other related
variables from their analysis. Although the present research sees
technical quality and service quality as distinct constructs, empirical
relationships exist between the scales of these twin factors. Therefore,
practitioners, researchers and educators should think of audit quality in
broader terms, which should encompass service attributes (ie empathy,
client service, responsiveness and provision of nonaudit services), as
well as technical ones (reputation, capability, independence, expertise
and experience).
2. Audit irms should emphasise those elements of audit quality most
important to clients and stakeholders when promoting their services
The present investigation found all three groups of auditors,clients
(inance directors) and stakeholders (fund managers) consistently rated
DISCUSSION
AND IMPLICATIONS
113
the two technical quality dimensions of Reputation and Capability
the highest. The service quality literature stresses the importance of
identifying the features of quality that are considered most important
by users. Therefore, when promoting assurance services, accounting
irmsshouldemphasisetheirreputationandthecapabilitiesoftheir
audit partners and staff. Clients and external users of accounting
informationarelikelytovaluetheirm’sreputationandpartner/staff
capability much more than for example, their ability to offer non
audit services.
3. Auditors should put in place systems to monitor and improve audit
quality
Audit irms trying to achieve a distinctive position and sustainable
competitive advantage by delivering the highest levels of audit quality
should put in place a continuous process to:
(i)
monitorclients’perceptionsofauditquality;
(ii) identifythecausesofauditqualityshortfalls;and
(iii) take appropriate action to improve audit quality.
Auditors should use the AUDITQUAL instrument described
inchapterivetoassessandunderstandclients’ perceptionsofaudit
quality. Applying the conceptual (‘extended gaps’) model described
in chapter three provides the means to link clientperceived quality
deicienciestowithin-irmdeiciencies(‘gaps’).
Avenues for future research
Part of the strength of any research project lies in the recognition
of its limitations. This will in part identify potential issues that
meritfuturework. First, thisstudyusessamplesofaccountingirm
114
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
partners to proxy for auditors, inance directors as representatives
of auditees and fund managers as external users (or customers) of
auditedinancialinformation. Futureworkthatrevealswhetherthe
results are generalisable across other groups of auditors, preparers of
inancialstatementsandexternaluserswouldbeinteresting. Secondly,
all attributes of quality were measured at one point in time, from a
static perspective. Recent corporate accounting scandals (eg Enron,
WorldCom) which have almost certainly damaged the accounting
profession may have altered the dynamics of audit quality.
Theextendedgapsmodelidentiieseightpotentialauditquality
gaps, their antecedents and means of closing these gaps. Work using
speciicauditirms, theirclientsandstakeholdershastheabilitytotest
these propositions empirically. Zeithaml et al, (1990) have developed
generic questionnaires, one to be administered to the service providers’
managers and customers (to measure gaps 1 and 2), the second to
employees (to measure gaps 3 and 4). Porter’s (1993, 2000) work uses
a questionnaire approach to identify what has been categorised within
the present investigation as gaps 6 to 8. The extended gaps model has
the capacity to inform all parties (auditors, clients and external users)
of the antecedents of audit quality and how these might be closed
This analysis of audit quality has drawn from the extant audit
quality literature (from the accounting research), service quality
literature (derived from marketing and operations management
research) and competitive rivalry and strategic positioning literature
(fromstrategicmanagementresearch). Givenirms’ desiretoextend
the boundaries of the traditional audit to develop riskbased assurance
services emphasising the advisory role (eg Elliott, 2002), auditors
and auditing researchers might ind some utility in the knowledge
management literature.
Finally, this research project has identiied the important role audit
staff and audit teams play in delivering audit quality. In particular, staff
with the necessary social skills who are capable of developing long-
DISCUSSION
AND IMPLICATIONS
115
term relationships with clients, whilst simultaneously possessing the
high levels of professional and personal integrity. Future work may
wishtoconceptualiseapersonalityproilethatmeetssuchapersonspeciicationandsubjectittoempiricaltesting. Thedevelopmentof
suchaproilemightassistauditirmstoscreenindividualsseeking
employment as an auditor using existing personality and individual
difference inventories.
Conclusion
Qualityisseenasonemeansofabusiness, oraccountingirmas
gainingcompetitiveadvantage. Althoughallaccountingirmswould
claimtosatisfyDeAngelo’s(1981)narrowdeinitionofauditquality
(competence and independence), larger irms might claim higher
levels of competence, based on the provision of specialist services,
ability to command higher fees and hence attract and retain higher
calibre staff. This research project has developed an empirical model
of the determinants of audit quality and developed an alternative
model of audit quality using the extended gaps model of Zeithaml
et al, (1990) and Porter (1993) to identify where discrepancies occur
between client and stakeholder perceptions and expectations of audit
quality. Chapterfourdescribedirms’ identiicationofthevaluethey
perceive to add to clients’ businesses by the provision of knowledge
other quality related aspects.
Audit quality bears some direct parallels to corporate strategy.
To quote Michael Porter in a recent interview (Argyres & McGahan,
2002 p.52).
Everything is important. Choices matter, leadership matters, values
matter, random events matter. It (corporate strategy) is an essentially
integrative topic that needs to allow for complexity.
116
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
It is hoped the audit quality model developed provides a
framework to assist practitioners and academics in understanding what
the dimensions of audit quality are, how discrepancies can arise and
how these might be managed.
referenCes
Accountancy Age (2002), Top 50, www.accountancyage.com/top50.
AICPA (1986), Division for CPA firms SEC Practice Section: SECPS
Manual, New York: American Institute of Certified Public
Accountants.
AICPA (1997), Assurance services: definition and interpretive commentary.
AICPA: Special Committee on Audit Services.
AICPA (1997), Recommendations to Improve, Expand CPAs Offerings,
http://www.aicpa.org/pubs/cpaltr/mar97/scasrr.htm.
AICPA (2001), ‘CPA Vision Project 2011 and Beyond’, American
InstituteofCertiiedPublicAccountants,NewYork.
Ahmed, Z and J Hopson (1990),‘A strategic plan for marketing audit
services’, CPA Journal,Vol.60, pp.5057.
Andersen, P (1999),‘Client or customer’, Accountancy, March, pp.7475.
APB (1992), The McFarlane Report, Auditing Practices Board,
London.
APB (1994), The Audit Agenda, Auditing Practices Board. London.
APB (1996a), The Audit Agenda - Next Steps, Auditing Practices Board,
London.
APB (1996b), Auditors Code, Auditing Practices Board, London.
APB (2001), E-business: Identifying Financial Statement Risks – Bulletin
2001/3,Auditing Practices Board, London.
Argyres, N and A M McGahan (2002), ‘An interview with Michael
Porter’, Academy of Management Executive,Vol.16(2), pp.4352.
Arbuckle, J L (1999),‘Amos Users’ Guide Version 4.0’, SPSS, Chicago,
IL.
118
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Arruňada, B (2000), ‘Audit quality: attributes, private safeguards and
the role of regulation’, The European Accounting Review, Vol.9(2),
pp.205224.
Arruňada, B. (1999a), The economics of audit quality: Private incentives
and the regulation of audit and non-audit services, Kluwer Academic
Publishers, Dordrecht, Netherlands.
Arruňada, B (1999b),‘The provision of nonaudit services by auditors:
let the market evolve and decide’, International Review of Law and
Economics,Vol.19, pp.513531.
Babakus, E and G W Boller (1992),‘Adapting the SERVQUAL scale
to hospital services: an empirical investigation’, Journal of Business
Research,Vol.24(3), pp.253268.
Bachmann, J W (2002),‘Competitive strategy: its OK to be different,
Academy of Management Executive,Vol.16(2), pp.6165.
Beattie, V and S Fearnley (1995), ‘The importance of audit irm
characteristics and the drivers of auditor change in UK listed
companies’, Accounting and Business Research,Vol.25, pp.227239.
Beattie,V and S Fearnley (1998), What companies want (and don’t want)
from their auditors, ICAEW Research Board, London.
Beattie,V, S Fearnley and R Brandt (2001), Behind closed doors: what
company audit is really about, Palgrave, London.
Behn,B K,JV Carcello,D R Hermanson and R H Hermanson (1997),
‘The determinants of audit client satisfaction among clients of Big 6
irms’, Accounting Horizons,Vol.11(1), pp.724.
Berry, L,V Zeithaml and A Parasuraman (1985), ‘Quality counts in
services, too’, Business Horizons,Vol.28 (May/June), pp.4452.
Berry,L L,D R Bennett and C W Brown (1989),Service Quality:A profit
strategy for financial institutions, Dow JonesIrwin, Homewood, IL.
Berry,L L,D R Bennett and C B Brown (1989),Service Quality - A Profit
Strategy for Financial institutions, DowJonesIrwin, Homewood, IL.
REFERENCES
119
Booth, P and H Winzar (1993), ‘Personality biases of accounting
students:some implications for learning style preferences’, Accounting
and Finance, November, 109120.
Bougen, P (1994), ‘Joking apart: The serious side to the accounting
stereotype’, Accounting, Organizations and Society,Vol.19(3), pp.319335.
Bowman, C and D Faulkner (1996), Competitive and Corporate Strategy,
Irwin, Homewood, IL.
Byrne, B M (1989), A Primer of LISREL: Basic applications and
programming for confirmatory factor analytic models, SpringerVerlag,
New York.
Cadbury,A (1992),Report of the Committee of FinancialAspects of Corporate
Governance, Gee & Company, London.
Carcello, JV, R H Hermanson and NT McGrath (1992),‘Audit quality
attributes: theperceptionsofauditpartners, preparers, andinancial
statement users’, Auditing: A Journal of Practice & Theory,Vol.11(1),
pp.115.
Carcello, J V, D R Hermanson and H F Huss(1995), ‘Audit quality
research in the United States’, Maanblad voor Accountancy en
Bedriffskunde,Vol.69 (June), pp.35964.
Carcello, JVandZPalmrose(1994),‘Auditorlitigationandmodiied
reporting on bankrupt clients’, Journal of Accounting Research,
Supplement, pp.130.
Carman, J M (1990), ‘Consumer perceptions of service quality: an
assessment of the SERVQUAL dimensions’, Journal of Retailing,
Vol.66(1), pp.3355.
Chan, P, MEzzamelandDGwilliam(1993), ‘Determinantsofaudit
fees for quoted UK companies’, Journal of Business, Finance and
Accounting,Vol.20, pp.765786.
Chandler, R and J R Edwards (1996), ‘Recurring issues in auditing:
back to the future?’ Accounting, Auditing & Accountability Journal,
Vol.9(2), pp.429.
120
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Cloyd, C, J Frederickson and J Hill (1996),‘Motivating factors against
independent auditors: experimental evidence on the importance of
causality’, Journal of Accounting and Public Policy,Vol.15(3), pp.185218.
Cohen Commission (Commission on Auditors Responsibilities)
(1978), Report conclusions, and recommendations,AICPA, New York.
Cohen, R J, M E Swerdlik and S M Phillips (1996), Psychological Testing
and Assessment (3rd Edition), MayieldPublishing, MountainView,
CA.
Cook, J D, S J Hepworth,T D Wall and P B Warr (1981), The experience
of work, Academic Press, San Diego, CA.
Craswell,A T, J R Francis and S L Taylor (1995),‘Auditor brand name
reputation and industry specialization’, Journal of Accounting and
Economics,Vol.20, pp.297322.
Cronin, J J and S A Taylor (1994), ‘SERVPERF versus SERVQUAL:
reconciling performancebased and perceptionsminusexpectations
measurement of service quality’, Journal of Marketing, Vol.58(1),
pp.125131.
Dart, J and K Freeman (1994),‘Dissatisfaction response styles among
clientsofprofessionalaccountingirms’, Journal of Business Research,
Vol.29(1), pp.7581.
De Ruyter, K and MWetzels (1999), ‘Commitment in auditorclient relationships: antecedents and consequences’, Accounting,
Organizations and Society,Vol.24, pp.5775.
DeAngelo, L E (1981), ‘Auditor size and audit quality’, Journal of
Accounting and Economics,Vol.3, pp.183199.
Deloitte Touche (2002), ‘www.deloitte.co.uk’, website accessed
September 2002.
Descouzis, D(1989)‘Psychologicaltypesoftaxpreparers’, Journal of
Psychological Type, 17, 3638.
REFERENCES
121
Diamantopoulos,A, S O’Donahue and N Petersen (1995),‘Marketing
priorities and practice within the audit profession:does formalization
make a difference?’,in W Winstone (Ed.),Marketing for CPAs,Auditors
and Tax Professionals, pp.5172, Hayworth Press, New York.
Dopuch, N and D Simunic (1980),‘The nature of competition in the
auditing profession: a descriptive and normative view’, in Buckley,
J W. & J F Weston (Eds.), Regulation and the Accounting Profession,
Lifetime Learning Publications.
Dopuch, N and D Simunic (1982), ‘Competition in auditing: An
assessment’, Fourth Symposium on Auditing Research, University of
Illinois.
Duff,A (2001) ‘Psychometric measurement in accounting education: a
review and some comments’, Accounting Education:An International
Journal, Vol.10(4), pp.383401.
Elliott, R K (1995),‘The future of assurance services: implications for
academia’, Accounting Horizons, Vol.9(4), pp.118127.
Elliott,R K (1998),‘Assurance services and the audit heritage’,Auditing:
A Journal of Practice and Theory,Vol.17 (Supplement), pp.17.
Elliott,RK(2002),‘Twenty-irstcenturyassurance’, Auditing:A Journal
of Practice and Theory, Vol.21(1), pp.139148.
Elliott, R K and P D Jacobson (1995), ‘AICPA assurance services
committee:What is the future of auditing?’ The Journal of Corporate
Accounting and Finance, (Winter), pp.8797.
Elliott, R K and D Pallais (1997),‘Are you ready for the new assurance
services?’, Journal of Accountancy, (June), pp.4751.
Ellis, B and J S Mosher (1995), ‘Six P’s for four characteristics: a
complete positioningstrategyfor theprofessional services irm(CPA
irmexample)’, inWWinstone(Ed.)Marketing for CPAs, Auditors
and Tax Professionals, pp.295312, Hayworth Press, New York.
Ernst & Young (2001), Annual Review, Ernst & Young, London.
122
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Federation des Experts Comptables Europeens (FEE) (1996), ‘The
Role,Position and Liability of the Statutory Auditor in the European
Union’, FEE, Brussels.
Fellingham, J C and D P Newman (1985),‘Strategic considerations in
auditing’, The Accounting Review, October, pp.634650.
Fisher, J, JWSchatzbergandBPShapiro(1996), ‘Atheoreticaland
experimental examination of strategic auditorclient interaction’,
Advances in Accounting,Vol.14, pp.135160.
Fogarty,T, J Heian and D Knutson (1991), ‘The rationality of doing
nothing: auditors’ responses to the legal liability in the institutional
environment’, Critical Perspectives on Accounting,Vol.2, pp.201226.
Francis, J(1984),‘Theeffectofauditirmsizeonauditprices:Astudy
of the Australian market’, Journal of Accounting and Economics,Vol.6,
pp.133151.
Francis, J and D Simon (1987), ‘A test of audit pricing in the small
client segment of the US audit market’, The Accounting Review,
Vol.3(1), pp.145157.
Frochot, I and H Hughes (2000),‘HISTOQUAL: the development of
a historic houses assessment scale’, Tourism Management, Vol.21(2),
pp.157167.
Getty,J and KThompson (1994),‘A procedure for scaling perceptions of
lodging quality’, Hospitality Research Journal,Vol.18(2), pp.7596.
Geyskens, I and J B Steenkamp (1995),‘An investigation into the joint
effects of trust and interdependence on relationship commitment,’
EMAC Proceedings, pp.351371.
Green,P,D S Tull and G Albaum (1988),Research for Marketing Decisions.
Englewood Cliffs, Prentice Hall International.
Greenbury, R (1995), Report of a Study Group on Directors’ Remuneration,
Gee Publishing Ltd, London.
Gundlach, GT, RSAchrolandJTMentzer(1995),‘Thestructureof
commitment in exchange’, Journal of Marketing,Vol.59, pp.7892.
REFERENCES
123
Gurton, A (1999),‘A bird in the hand’, Accountancy, December, pp.445.
Hampel, R (1998), Committee on Corporate Governance: Final Report,
Gee Publishing, London.
Hanlon, G (1994), The Commercialization of Accountancy: Flexible
Accumulation and the Transformation of the Service Class, St. Martins
Press, London.
Hatherly, D (1999), ‘The future of auditing: the debate in the UK’,
European Accounting Review,Vol.8(1), pp.5165.
Higson, A (1997), ‘Developments in audit approaches – from audit
eficiencytoauditeffectiveness?’. InMSherer&STurley(Eds.)
Current Issues in Auditing, 3rd Edition, pp.198215, Paul Chapman
Publishing, London.
Hinkin,T R (1995) ‘A review of scale development practices in the
study of organisations’, Journal of Management, Vol.21(5), pp.967988.
Hopwood,W, J C McKeown and J Mutchler (1994),‘A reexamination
of auditor versus model accuracy within the context of the going
concern opinion decision’, Contemporary Accounting Research,
Vol.10(2), pp.40931.
Hoyle, R H and A T Panter (1995),‘Writing about structural equation
models’. In R H Hoyle (Ed.) Structural equation modeling, pp.7699.
Sage,Thousand Oaks, CA.
Hu, L and P M Bentler (1998), ‘Cutoff criteria for it indexes in
covariance structure analysis: conventional criteria versus new
alternatives’, Structural Equation Modeling,Vol.6(1), pp.155.
Hu, L and P M Bentler (1999), Fit indices in covariance structure
modelling:sensitivitytounderparameterisedmodelmisspeciication’,
Psychological Methods,Vol.3(4), pp.423453.
Hunter, S and M Sundel (1989), ‘Introduction: An examination of
key issues concerning midlife’, In: Midlife Myths (Eds.) S Hunter &
M Sundel, pp.828, Sage, London.
124
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
ICAEW (1999) (Turnbull Report), Internal Control: Guide for Directors
on the Combined Code,Institute of Chartered Accountants in England
and Wales, London.
ICAEW (2000), The Role of Audit and Assurance Engagements, ICAEW,
London.
Jacoby, P F (1981), ‘Psychological types and career success in the
accounting profession’, Research in Psychological Type,Vol.4, pp.2437.
Jensen,M C andW H Meckling (1976),‘Theory of the irm:managerial
behavior, agency costs and ownership structure’, Journal of Financial
Economics,Vol.3(4), pp.30560.
Jeter, D C and P Erickson Shaw (1995), ‘Solicitation and auditor
reporting decisions’, The Accounting Review,Vol.70, pp.293315.
Johnson, G (2002), ‘KPMG (A): Strategic Change in the 1990s’,
Exploring Corporate Strategy: text and cases, pp.10121030, Prentice
Hall, London.
Johnson, G and K Scholes (1997), Exploring Corporate Strategy: text and
cases, Prentice Hall, London.
Jung, C G (1921), Psychological Types, Princeton Press, NJ.
Kirk, RE(1996), ‘Practicalsigniicance: Aconceptwhosetimehas
come’, Educational and Psychological Measurement, Vol.56, pp.746759.
KPMG (2001), KPMG UK Annual Report 2001, KPMG, London.
Krishnan J and J Krishnan (1997), ‘Litigation risk and auditor
resignations’, Accounting Review, Vol.72(4), pp.539-60.
Kumar, N, L K Scheer and J E M Steenkamp (1995), ‘The effects of
perceived interdependence on perceived dealer attitudes’, Journal of
Marketing Research,Vol.32, pp.348356.
Lemon,W M, K W Tatum and W S Turley (2001), Developments in the
Methodologies of large accounting firms,ABG Professional Information,
London.
REFERENCES
125
Lewis, B R (1993),‘Service quality measurement’,Marketing Intelligence
and Planning,Vol.11(4), pp.3644.
Liggio, C D (1974), ‘The expectation gap: the accountants legal
Waterloo’, Journal of Contemporary Business, Vol.3, pp.2744.
Lynch, R (2000), Corporate Strategy, Financial TimesPrentice Hall,
London.
MacCallum, R C and J T Austin (2000), ‘Applications of structural
equation modelling in psychological research’, Annual Review of
Psychology,Vol.51, pp.201226.
McCrae, R R and P T Costa (1989),‘Reinterpreting the MyersBriggs
Type Indicator from the perspective of the ive factor model of
personality’, Journal of Personality,Vol.57(1), pp.1737.
McDonald, R P and H W Marsh (1990), ‘Choosing a multivariate
model: Noncentrality and goodness-of-it’, Psychological Bulletin,
Vol.107, pp.247255.
Manson, S, S McCartney and M Sherer (1997), Audit Automation: the
use of information technology in the planning and recording of audit work,
ICAS, Edinburgh.
Marsh, HW, J R Balla and KT Hau (1996), ‘An evaluation of it
indices: A clariication of mathematical and empirical processes’.
In G A Marcoulides & R E Schumacker (Eds.), Advanced structural
equation modeling techniques, Vol.4, pp.5998, Lawrence Erlbaum,
Hillsdale, NJ.
Marsh, H W, J R Balla and K T Hau (1997), Is more ever too much:
the number of indicators per factor in confirmatory factor analysis, Paper
presented at the Annual Meeting of the American Educational
Research Association, Chicago.
Marsh, HW and K-L Hau (1999), ‘Conirmatory factor analysis:
strategiesforsmallsamplesizes’. InRHoyle(Ed.) Statistical Strategies
for Small Sample Research, pp.285284.
126
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Martinov, R and Roebuck, (1998), ‘The assessment and integration
of materiality and interest risk: an analysis of major irms’ audit
practices’, Accounting, Auditing & Accountability Journal, Vol.14(1),
pp.109130
Mednick, R (1990),‘Independence: Let’s get back to basics’, Journal of
Accountancy,Vol.169(1), pp.8693.
Menon, K and D Williams (1991),‘Auditor credibility and the initial
public offering’, Accounting Review,Vol.67(2), pp.313332.
Menon, K and D D Williams (2001),‘Longterm trends in audit fees’,
Auditing:A Journal of Practice and Theory,Vol.20(1), pp.11536.
Moizer, P(1998), ‘Companydirectors’ viewsontheperformanceof
the 1987 Big Eight and the 1996 Big Six’, Leeds University Business
School Working Papers.
Mock, T J and M Samet (1982), ‘A multiattribute model for audit
evaluation’, In Proceedings of the VI University of Kansas Audit
Symposium.
Nichols D and D Smith (1983), ‘Auditor credibility and auditor
changes’, Journal of Accounting Research,Autumn, pp.53444.
Nunnally, J C (1978), Psychometric Theory (2nd ed.), McGrawHill, New
York.
O’Sullivan,N (2000),‘The impact of board composition and ownership
on audit quality: evidence from large UK companies’, British
Accounting Review,Vol.32, pp.397414.
Oppenheim,A N (1966),Questionnaire Design and Attitude Measurement’
Heinemann, London.
Palmrose, Z(1986), ‘Auditorfeesandauditorsize: furtherevidence’,
Journal of Accounting Research, Spring, pp.97110.
Palmrose, Z (1987),‘Litigation and independent auditors:The role of
business failures and management fraud’,Auditing:A Journal of Practice
& Theory, Vol.6(2), Spring, pp.90103.
Palmrose, Z (1989),‘The relation of audit contract type to audit fees
and hours’, The Accounting Review,Vol.65(3), pp.488499.
REFERENCES
127
Parasuraman,A,V A Zeithaml and L Berry (1985),‘A conceptual model
of service quality and its implications for future research’, Journal of
Marketing,Vol.49 (Fall), pp.4150.
Parasuraman, A,V A Zeithaml and L Berry, L (1988), ‘SERVQUAL:
a multiitem scale for measuring consumer perceptions of service
quality’, Journal of Retailing,Vol. 64(1), pp.1240.
Parasuraman, A,VAZeithamlandLBerry(1991), ‘Reinementand
reassessment of the SERVQUAL scale’, Journal of Retailing,Vol.67(4),
pp.420449.
Ponemon, L A and D R L Gabhart (1990), ‘Auditor independence
judgements: a cognitive developmental model and experimental
evidence’, Contemporary Accounting Research,Vol.7(1), pp.227251.
Popper, K R (1959), The Logic of Scientific Discovery, Basic Books, New
York.
Porter, B A (1991), ‘Narrowing the expectationperformance gap: A
contemporary approach’, Pacific Accounting Review, Vol.3(1), pp.136.
Porter, B A (1993), ‘An empirical study of the audit expectations
performance gap’, Accounting and Business Research, Vol.24(93),
pp.4968.
Porter, B A (2000),‘An empirical study of the UK audit expectations
gap’,Working Paper, University of Warwick.
Power, M. (1994), The Audit Explosion, Demos, London:
PwC (2002),‘www.pricewaterhousecoopers.co.uk’, website accessed
September 2002.
Ridyard, D and J de Bolle (1992), Competition in European Accounting,
Lafferty Publications Ltd, Dublin.
Schmitt, N W and R J Klimoski (1991), Research methods in human
resource management, SouthWestern Publishing, Cincinnati, KY.
Schneider, B and D E Bowen (1985), ‘Employee and customer
perceptions of service in banks: Replication and extension, Journal
of Applied Psychology,Vol.70, pp.423433.
128
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Schriesheim, C A, K J Powers,T A Scandura, C C Gardiner and M J
Lankau (1993),‘Improving construct measurement in management
research: Comments and a quantitative approach for assessing the
theoretical content adequacy of paperandpencil surveytype
instruments’, Journal of Management,Vol.19, pp.385417.
Schroeder, M S, I Solomon and D W Vickery (1986), ‘Audit quality:
the perceptions of audit committee chairperson and audit partners’,
Auditing:A Journal of Practice & Theory,Vol.5(1), pp.8694.
Schwab, D P (1980),‘Construct Validity in Organisational Behavior’.
In B M Straw and L L Cumings (eds.), Research in Organizational
Behavior,Vol.2, JAI Press, Greenwich, CT.
Shackleton,V (1980), ‘The accountant stereotype: myth or reality?’,
Accountancy,Vol.113(11), November, pp.122123.
Simon, D (1985), ‘The audit services market: additional empirical
evidence’, Auditing:A Journal of Practice & Theory,Vol.4 (Fall), pp.7178.
Simon, D and J Francis (1988),‘The effects of auditor change on audit
fees: tests of price cutting and price recovery’, The Accounting Review,
Vol.63(2), pp.255269.
Simunic,D (1980),‘The pricing of audit services:theory and evidence’,
Journal of Accounting Research (Spring), pp.161190.
Simunic, D and M Stein (1996),‘The impact of litigation risk on audit
pricing: a review of the economics and the evidence’, Auditing: a
Journal of Practice & Theory,Vol.15 (Supplement), pp.11934.
St.Pierre,K and J Anderson (1984),‘An analysis of the factors associated
with lawsuits against public accountants’, The Accounting Review,
Vol.59(2), pp.24263.
Stice, J (1991), ‘Using inancial and market information to identify
preengagement factors associated with lawsuits against auditors’,
The Accounting Review,Vol.66(3), pp.51633.
REFERENCES
129
Suh, S H, YH Lee, Y Park and C C Shin (1997), ‘The impact of
consumer involvement on the consumers’ perception of service
quality–focusingontheKoreanhotelindustry’, Journal of Travel
and Tourism Marketing,Vol.6(2), pp.3352.
Sutton, S G (1993),‘Toward an understanding of the factors affecting
the quality of the audit process’, Decision Sciences, Vol.24, pp.88105.
Sutton,S G and J G Lampe (1990),‘Formulating a process measurement
system for audit quality’, in Proceedings of the 1990 University of
Southern California Audit Judgement Symposium.
Tamir, L (1989),‘Modern myths about men at midlife: an assessment’.
In S Hunter & M Sundel (Eds.) Midlife Myths, pp.157179, Sage,
London.
Taylor, S A and T T Baker (1994), ‘An assessment of the relationship
between service quality and customer satisfaction in the formation
of consumers’ purchase intentions, Journal of Retailing, Vol.70(2),
pp.163178.
Thompson, B and L G Daniel (1996), ‘Factor analytic evidence for
the construct validity of scores: a historical overview and some
guidelines’, Educational and Psychological Measurement, Vol.56(2),
pp.197208.
Thornbury, J (1999), ‘KPMG: revitalising culture through values’,
Business Strategy Review,Vol.10(4), pp.115.
Tribe, J andT Snaith (1998),‘From SERVQUAL to HOLSAT: holiday
satisfaction inVaradero,Cuba’, Tourism Management,Vol.19(1),pp.2534.
Tucker,L R and C Lewis (1973),‘A reliability coeficient for maximum
likelihood factor analysis’, Psychometrika, Vol.38, pp.110.
Turpen,R (1990),‘Differential pricing on auditors’initial engagements:
further evidence’, Auditing: A Journal of Practice & Theory, Vol.9(1),
pp.6076.
130
AUDITQUAL: DIMENSIONS
OF
AUDIT QUALITY
Van der Platts, E (2000),‘Regulating auditor independence’, European
Accounting Review,Vol.9(4), pp.625638.
Walker, K (2001), ‘A measured view of clients’, Accountancy, April,
pp.6263.
Wallace,W (1980), The Economic Role of the Audit in Free and Regulated
Markets, Touche Ross, London.
WarmingRasmussen, B and L Jensen (1998), ‘Quality dimensions
inexternalauditservices–anexternaluserperspective’, European
Accounting Review,Vol.7(1), pp.115.
Watts, R L and J L Zimmerman (1986), Positive Accounting Theory.
Prentice, Hall, Englewood Cliffs, NJ.
Windmöller,R.(2000),‘The auditor market and auditor independence’,
European Accounting Review,Vol.9(4), pp.639642.
Windsor, C A and N M Ashkansay (1995), ‘The effect of client
management bargaining power, moral reasoning, and belief in a just
world on auditor independence, Accounting, Organisations and Society,
Vol. 20(7/8), pp.701720.
Wolk, C and L A Nikolai (1994), ‘Personality types of accounting
students and faculty: Comparisons and Implications’, Journal of
Accounting Education,Vol.12(1), pp.117.
Zeithaml,V A, L L Parasuraman and L L Berry (1990), Delivering quality
service: balancing customer perceptions and expectations, The Free Press,
New York.
Appendix one
items used in AuditquAl questionnAire
And their origin
Directions: A list of statements is given below. On the basis of
your general views about auditing (rather than your experiences of
asingleauditirm), pleaseratetheimportanceofeachstatementto
audit quality by circling the number on the scale on the right. If you
feel such a factor is not important for high audit quality, please circle
the number 1. If you feel a feature is very important for high audit
quality, circle 5. If your feelings are less strong, please circle one of the
numbers in the middle.
Audit firm factors
1
Theauditirmishighlycompetent(Beattie&Fearnley,1995)
2
Theauditirmmakesextensiveuseofstatisticaltechniqueswhen
undertaking the audit (Carcello et al, 1992)
3
Theauditirmoperatestothehigheststandardsofintegrity
(Beattie & Fearnley, 1995)
4
Theauditirmisskilfulindevisingaccountingtreatmentsthat
generate results management wishes to obtain (Carcello et al, 1992)
5
Theauditirmconductingtheauditprovidesnonon-auditservices
totheirm(adaptedfromCarcelloet al, 1992)
6
Theauditirmregularlyconductsclientservicereviewmeetings
(new)
7
Theauditirmhasrarelybeenfoundnegligentinlitigationagainst
it–alleginginadequateauditperformance(Carcelloet al, 1992)
8
Theauditirmconductsathoroughstudyoftheclient’ssystemof
internal control
9
Theauditirmhasotherclientsinthesameindustry(Carcelloet al,
1992)
10 Theauditirmisobjective(Warming-Rasmussen&Jensen,1998)
11 Theauditirmisconscientious(Warming-Rasmussen&Jensen,
1998)
132
APPENDIX ONE
Audit firm factors (continued)
12
Theauditirmiscredibletothirdparties
13
14
Theauditirmundertakesresearchintotheclient’sindustry(new)
Theauditirmemploysindividualsindependentoftheauditirm
to conduct client service reviews (new)
Theauditirmisabletosupplyadditionaltaxservices(Beattie&
Fearnley, 1995)
Theauditirmisabletoprovideadditionalconsultancyservices
(Beattie & Fearnley, 1995)
15
16
18
Theauditirmisabletoprovideadditionalaccountingservices
(Beattie & Fearnley, 1995)
Theauditirmisabletoprovideinternalauditservices(new)
19
Theauditirmenjoysagoodreputation(Beattie&Fearnley,1995)
20
Theauditirmisindependentoftheboardofdirectors(DeAngelo,
1981)
21
Theauditirmiswillingtoprovidedetailedcostinformation
(Beattie & Fearnley, 1995)
17
Engagement partner factors
25
26
27
28
29
30
31
The engagement partner is proactive and contributory (eg suggests
potential acquisition targets)
The engagement partner arranges regular meetings with the client’s
key staff to identify issues of concern (new)
The engagement partner has been performing the audit for the past
three years (Carcello et al, 1992)
The engagement partner is subject to internal review during the
auditbyotherpartnersoftheauditirm(Beattie,Fearnley&Brandt,
2001)
The engagement partner and senior manager make frequent visits to
the audit site for technical review purposes (Carcello et al, 1992)
Theengagementpartnerprovidestheclient’sinancedirectorwith
individual attention (Parasuraman et al, 1991)
The engagement partner has the client’s best interests at heart
(Parasuraman et al, 1991)
APPENDIX ONE
133
Engagement partner factors (continued)
32
33
34
35
36
37
38
39
40
41
The engagement partner is highly competent (Parasuraman et al,
1991)
The engagement partner has high ethical standards (Parasuraman et
al, 1991)
Theengagementpartnerregularlyidentiiesexamplesofadded
value to the client (new)
The engagement partner is actively involved in the engagement
beginning with the initial planning and throughout the audit process
(Carcello et al, 1992)
Theengagementpartnerhasinancialstatementusers’bestinterests
at heart (Parasuraman et al, 1991)
The engagement partner is keen to understand what is happening
within the client’s organisation (new)
The engagement partner is very knowledgeable about the client’s
industry (Carcello et al, 1992)
The engagement partner is easily contactable (eg by phone) (new)
Thereisa‘goodit’betweenthepersonalityoftheengagement
partnerandtheinancedirector(new)
Therelationshipbetweentheengagementpartnerandinance
director is relatively informal (new)
Audit team factors
42
Audit team staff create the minimum of disruption so far as
practically possible (new)
43
The audit team are willing to provide guidance on accounting
principles Beattie & Fearnley, 1995)
44
The audit team staff are highly competent (Beattie & Fearnley,
1995)
45
The audit team staff operate to high ethical standards (Beattie &
Fearnley, 1995)
The audit team provides the client with personal attention
(Parasuraman et al, 1991)
46
47
The audit team develops stringent time budgets for each audit area
and expects people to meet them (Carcello et al, 1992)
134
APPENDIX ONE
Audit team factors (continued)
48
49
Themanageroftheauditirmhasbeenperformingtheauditfor
at least two years (Carcello et al, 1992)
The audit fee paid by the client does not represent more than
10% of the total audit fees controlled by the engagement partner
(Carcello et al, 1992)
50
The audit fee paid by the client does not represent more than
25% of the total audit fees controlled by the engagement partner
(Carcello et al, 1992)
51
The audit fee paid by the client does not represent more than
50% of the total audit fees controlled by the engagement partner
(Carcello et al, 1992)
There is frequent communication between the audit team and
the audit committee (Carcello et al, 1992)
52
53
Theseniormanageroftheauditirmhasbeenperformingthe
audit for at least two years (Carcello et al, 1992)
54
The client has a knowledgeable and active audit committee
(Carcello et al, 1992)
The senior manager and manager assigned to the audit are very
knowledgeable about the client’s industry (Carcello et al, 1992)
There is frequent communication between the audit team and
executive management (Carcello et al, 1992)
55
56
Appendix two
sCoring key
Each of the nine dimensions of the AUDITQUAL instrument are
shown in italics. Those higherorder factors (ie composite dimensions)
are shown in bold. The alpha coeficients calculated are for the
composite sample of accounting irm partners, inance directors,
and fund managers. Coeficient alpha is a measure of the internal
consistency reliability of the dimension.
Items used in survey excluded from the analysis
Reputation & capability (alpha = .867)
Reputation (alpha = .798)
1 Theauditirmishighlycompetent
3 Theauditirmoperatestothehigheststandardsofintegrity
7 Theauditirmhasrarelybeenfoundnegligentinlitigationagainstit
–alleginginadequateauditperformance
10 Theauditirmisobjective
11 Theauditirmisconscientious
12 Theauditirmiscredibletothirdparties
19 Theauditirmenjoysagoodreputation
20 Theauditirmisindependentoftheboardofdirectors
136
APPENDIX TWO
Capability (alpha = .774)
32 The engagement partner is highly competent
33 The engagement partner has high ethical standards
35 The engagement partner is actively involved in the engagement
beginning with the initial planning and throughout the audit process
36 Theengagementpartnerhasinancialstatementusers’bestinterestsat
heart
37 The engagement partner is keen to understand what is happening
within the client’s organisation
44 The audit team staff are highly competent
45 The audit team staff operate to high ethical standards
Responsiveness (alpha = .802)
4 Theauditirmisskilfulindevisingaccountingtreatmentsthat
generate results management wishes to obtain
21 Theauditirmiswillingtoprovidedetailedcostinformation
23 Theauditirmiswillingtobelexiblewhenschedulingthetimingof
audit visits
24 Theauditirm’soficesaregeographicallyclosetotheclient
39 The engagement partner is easily contactable (eg by phone)
40 Thereisa‘goodit’betweenthepersonalityoftheengagementpartner
andtheinancedirector
41 Therelationshipbetweentheengagementpartnerandinancedirector
is relatively informal
42 Audit team staff create the minimum of disruption so far as practically
possible
47 The audit team develop stringent time budgets for each audit area and
expects people to meet them
APPENDIX TWO
Independence (alpha = .768)
5 Theauditirmconductingtheauditprovidesnonon-auditservicesto
theirm
14 Theauditirmemploysindividualsindependentoftheauditirmto
conduct client service reviews
49 The audit fee paid by the client does not represent more than 10% of
the total audit fees controlled by the engagement partner
50 The audit fee paid by the client does not represent more than 25% of
the total audit fees controlled by the engagement partner
51 The audit fee paid by the client does not represent more than 50% of
the total audit fees controlled by the engagement partner
Non-audit services (alpha = .786)
15 Theauditirmisabletosupplyadditionaltaxservices
16 Theauditirmisabletoprovideadditionalconsultancyservices
17 Theauditirmisabletoprovideadditionalaccountingservices
18 Theauditirmisabletoprovideinternalauditservices
Empathy & client service (alpha = .824)
Empathy (alpha =.708)
25 The engagement partner is proactive and contributory
30 Theengagementpartnerprovidestheclient’sinancedirectorwith
individual attention
31 The engagement partner has the client’s best interests at heart
47 The audit team provides the client with personal attention
137
138
APPENDIX TWO
Client service (alpha = .749)
6
Theauditirmconductsclientservicereviews
26 The engagement partner arranges regular meetings with the client’s
key staff to identify issues of concern
34 Theengagementpartnerregularlyidentiiesexamplesofaddedvalue
to the client
29 The engagement partner and senior manager make regular visits to the
audit site for technical review
43 The audit team are willing to provide guidance on accounting
Principles
52 There is frequent communication between the audit team and audit
committee
56 There is frequent communication between the audit team and
executive management
Knowledge (alpha = .755)
Expertise (alpha = .70)
9
Theauditirmhasotherclientsinthesameindustry
13 Theauditirmundertakesresearchintotheclient’sindustry
28 The audit partner is subject to internal review during the audit by
otherpartnersoftheirm
38 The engagement partner is very knowledgeable about the client’s
industry
54 The client has a knowledgeable and active audit committee
55 The senior manager and manager assigned to the audit are very
knowledgeable about the client’s industry
Experience (alpha = .81)
27 The engagement partner has been performing the audit for the past
three years
48 Themanageroftheauditirmhasbeenperformingtheauditforat
least two years
53 The senior manager has been performing the audit for at least two
years
APPENDIX TWO
Items used in survey excluded from the analysis
8
20
27
Theauditirmconductsathoroughstudyoftheclient’ssystemof
internal control
Theauditirmisindependentoftheboardofdirectors
The engagement partner has been performing the audit for the
past three years
139
Appendix three
development of AuditquAl model
Methods and statistical analyses
EFAwasusedtoreinethehypothesisedstructureofAUDITQUAL.
The extraction method used was principal components analysis
followed by oblique rotations to allow for the correlations among
the scales. Principal components was chosen because this method
yieldscomponentscoresthathavethesamecorrelationcoeficientsas
the rotated factors and because component analysis does not unduly
capitalizeonsamplingerrorasthepriceforestimatingmeasuringerror
(Thompson&Daniel, 1996). Factorpatternmatrixcoeficientswith
values less than 0.3 are excluded from this analysis.
The hypothesised constructs developed from the literature review
identiiedninedimensionsofauditquality: Thosespeciictoextant
evidenceconsideringauditqualityincluded:(i)AuditorReputation;(ii)
CapabilityoftheAuditor; (iii)Independence; (iv)AuditorExperience
with the Client; (v)Auditor Expertise; and (vi) Provision of NonauditServices. Thosespeciictotheunderstandingofauditquality
included: (vii)Empathy; (viii)Assurance; and(ix)Responsiveness.The
internalreliabilityconsistencyestimates(alphacoeficients)areshown
in Table A.1.
142
APPENDIX THREE
Table A.1:Alpha coefficients for AUDITQUAL scales
No. Items
Alpha
coefficient
Status
Capability
Responsiveness
Independence
Nonaudit services
Empathy
Assurance
Expertise
Experience
8
7
9
5
4
4
7
6
3
.798
.774
.802
.768
.786
.708
.749
.700
.810
Auditor reputation
Relationship with
client
Knowledge of client
Audit quality
Service quality
15
11
.867
.824
9
33
20
.755
.819
.856
Scale
Lower-order factors
I.
II.
III.
IV.
V.
VI.
VII.
VIII.
IX.
Higher-order factors
I+II
VI+VII
VIII+IX
I+II+IV+V+VIII+IX
III+VI+VII
Goodness-of-it and testing of multiple models
Testing multiple models is usually undertaken to establish which
modelitsthedatabest. Sixmodelswereittedtothedata, theresults
areshowninTableA.2. Toestablishmodelit, goodness-of-itindices
are calculated using structural equation modelling (SEM) software, the
AMOS v4.0 program (Arbuckle, 1999). Although various “rulesofthumb”have developed over the past two decades of SEM development,
a general consensus now exits, that using two different types of it
indices in combination yields the most reliable results (see MacCallum
143
APPENDIX THREE
& Austin, 2000). As the present dataset is relatively small (N=260), Hu
&Bentler(1998, 1999)suggestusingtheStandardizedRootMean
Squared Residual (SRMR) in combination with the Comparative
Fit Index (CFI) will reject reasonable proportions of truepopulation
andmis-speciiedmodels. Hu&Bentler(1998, 1999)suggestaCFI
greater than 0.95 and an SRMR of close to 0.10 are indicative of
satisfactoryit.
Table A2: EFA of AUDITQUAL items
Item
1
10
11
20
45
3
44
32
12
19
33
37
36
35
40
23
24
41
48
39
42
4
21
51
50
52
5
I
.713
.708
.680
.678
.664
.663
.641
.579
.565
.563
.484
.480
.426
.382
II
III
.659
.646
.635
.628
.547
.518
.495
.403
.391
.882
.845
.769
.396
IV
V
VI
144
APPENDIX THREE
Item
14
2
7
16
17
15
18
30
47
29
25
31
26
53
34
6
43
57
54
9
49
56
27
13
28
38
55
I
II
III
.381
.334
.326
IV
V
VI
.738
.736
.702
.652
.719
.665
.584
.544
.521
.518
.482
.482
.473
.382
.300
.609
.597
.592
.541
.536
.452
.414
.408
.395
Factor pattern matrix coeficients shown greater than 0.3; after oblique
rotation;principalcomponentsanalysis.
Conirmatoryfactoranalyses(CFAs)wereconductedwiththe
SPSS version of Amos v3.6 (Arbuckle, 1999). When undertaking CFA,
there are “vague and sometimes contradictory guidelines about the
desirable amount of data” (Marsh & Hau, 1999 p.252). Marsh, Balla,
APPENDIX THREE
145
& Hau (1997) using a Monte Carlo study investigated the effect of
varying numbers of indicators (items) per factor (p/f ratio) on varying
samplesizes. Theirresultssupporta“moreisbetter” approachtoboth
samplesizeandp/fratio. Forap/fratioassmallassix, asamplesizeof
50wasadequate. ThereforethecombinedsamplesizeofN = 260, is
satisfactory for conducting CFAs at the item level of the LSQ, where
the number of items per factor (p/f) equals 20 (Marsh et al, 1997).
Hoyle & Panter (1995) recommend when using multiple indicators
ofoverallitshouldbeselectedfrom“absolute-itindexes” (suchas
χ2, and theAGFI) and“incremental it indexes,” which should be
selected from “type2” and “type3” indexes, such as the TLI (Tucker
& Lewis, 1973) and RNI (McDonald & Marsh, 1990). A type2
indexcomparesthelackofitofatargetmodeltothelackofitofa
baseline model, usually the independence model. Value estimates the
relative improvement per degrees of freedom of the target model over
a baseline model (Hoyle & Panter, 1995). A type3 index “indexes the
relativereductioninlackofitasestimatedbythenoncentral(2ofa
target model versus a baseline model” (Hoyle & Panter, 1995).
Therefore, evaluatinggoodness-of-it, presentsthe χ2 statistic, the
Relative Noncentrality Index (RNI), the TuckerLewis Index (TLI),
the ratio of the discrepancy, χ2, divided by the degrees of freedom (χ2
/df), the AdjustedGoodness of Fit Index (AGFI) and an evaluation of
parameter estimates to ensure the solution is proper (Marsh, Balla &
Hau, 1996; McDonald&Marsh, 1990).Althoughnoprecisestandards
existtoindicatewhatvalueofindicesareneededforasatisfactoryit,
typical guidelines are that the RNI should exceed .9. Various rules
ofthumb ranging from 2 to 5 have been suggested as cutoffs for
CMIN/df. The present study follows the recommendations of Byrne
(1989) that a χ2/dfratioofgreaterthan2.0representsaninadequateit.
Also computed is the expected cross validation index (ECVI), which
is useful when comparing models (MacCallum & Austin, 2001). No
146
APPENDIX THREE
absolute cutoff value exists for ECVI,values are for model comparison
purposes only.
TableA.3reportsthegoodness-of-itstatisticsforsixcompeting
models. The first is the simple ninefactor model of the nine
AUDITQUAL dimensions. The second is a sixfactor model,where six
of the original nine factors are combined into three new dimensions.
The third model is a nested model, ie a sixfactor model, with two
higherorder factors to represent technical quality and service quality.
The fourth model is a hybrid of models one and two, testing nine
irst-orderfactorsandthreesecondorderfactors(ie recognising the
existence of nine dimensions,whilst combining six of these factors into
threehigher-orderfactors). Theifthmodelisalsoanestedmodel.
Model ive consists of nine irst-order factors, three second-order
factors and two thirdorder factors (representing technical quality and
servicequality). Theinalmodeltestedisaone-factormodel, which
assumes there are no underlying factors in the data, and that all items
combinetocreateoneunitaryauditqualityfactor. Theinalmodel
is constructed for comparison purposes only as the literature review
suggests audit quality is made up of a number of different factors.
147
APPENDIX THREE
Table A3 Scale correlations and goodness-of-fit statistics
I
II
III
IV
V
VI
VII
VIII
Nine-factor model
χ2 (1239) = 2299.747; χ2/df = 1.856; CFI = .966; TLI = .962; ECVI =
14.816.
Six-factor model
I. Auditor
reputation
II. Responsiveness .072
III. Independence
.061 .247
IV. Nonaudit
.007 .306 .124
services
V. Relationship
.462 .431 .138 .238
with client
VI. Knowledge of
.468 .217 .174 .067 .514
client
χ2 (1362) = 3054.683; χ2/df = 2.269; CFI = .958; TLI = .954; ECVI =
14.215.
Six first-order factors, two higher-order factors
χ2 (1269) = 2993.046; χ2/df = 2.359; CFI = .954; TLI = .950, ECVI =
14.734.
Nine first-order factors, three second-order factors
χ2 (1254) = 2601.50; χ2/df = 1.889; CFI = .963; TLI = .960, ECVI =
15.103.
Nine first-order factors, three second order factors, two third-order factors
χ2 (1261) = 2407.51; χ2/df = 1.908; CFI = .963; TLI = .959, ECVI =
15.160.
One-factor model
χ2 (1377) = 4554.45; χ2/df = 3.308; CFI = .918; TLI = .912, ECVI =
21.682.
148
APPENDIX THREE
Examinationoftheitindicespointstotwomodels. Thebestittothedataisthenine-factormodel, withallitindicesbelow
recommended cutoff levels, and the second lowest ECVI statistic.
However, the ninefactor, three secondorder and two thirdorder
model(showninFigure4.1)alsoprovidesahighlysatisfactoryitto
the data, with an ECVI statistic (15.160) only slightly greater than
the simple ninefactor model (14.816). Therefore, the nested model
(seeigure4.1)ispreferredongroundsofparsimony. Thatis, the
model with the higher order factors (in this case technical quality
and service quality) provides a better description of the relationship
between the components of the model.