+1(978)310-4246 credencewriters@gmail.com


College of Administrative and Financial Sciences
Assignment 3
Deadline: 10/04/2021 @ 23:59
Course Name: Accounting Research
and practice
Student’s Name:
Course Code: ACCT 403
Student’s ID Number:
Semester: II
Academic Year: 1441/1442 H
For Instructor’s Use only
Instructor’s Name:
Students’ Grade: Marks Obtained/5
Level of Marks: High/Middle/Low
• The Assignment must be submitted on Blackboard (WORD format only) via allocated
• Assignments submitted through email will not be accepted.
• Students are advised to make their work clear and well presented, marks may be
reduced for poor presentation. This includes filling your information on the cover page.
• Students must mention question number clearly in their answer.
• Late submission will NOT be accepted.
• Avoid plagiarism, the work should be in your own words, copying from students or
other resources without proper referencing will result in ZERO marks. No exceptions.
• All answered must be typed using Times New Roman (size 12, double-spaced) font.
No pictures containing text will be accepted and will be considered plagiarism).
• Submissions without this cover page will NOT be accepted.
Question 1
What are the four guidelines for good experimental research?
[1 mark]
Question 2
Explain why the publication in a Refereed journal articles is more difficult than Professional
journal articles.
[2 marks]
Question 3
Explain why archival study will normally have more external validity than experimental or
simulation approaches?
[0.5 mark]
Question 4
What are the characteristics of a good sample?
[1.5 marks]
4/2/03 4:42 PM
Page i
SAGE Publications
London • Thousand Oaks • New Delhi
4/2/03 4:42 PM
Page ii
© Malcolm Smith 2003
First published 2003
Apart from any fair dealing for the purposes of research or
private study, or criticism or review, as permitted under the
Copyright, Designs and Patents Act, 1988, this publication
may be reproduced, stored or transmitted in any form, or by
any means, only with the prior permission in writing of the
publishers, or in the case of reprographic reproduction, in
accordance with the terms of licences issued by the Copyright
Licensing Agency. Enquiries concerning reproduction outside
those terms should be sent to the publishers.
SAGE Publications Ltd
6 Bonhill Street
London EC2A 4PU
SAGE Publications Inc
2455 Teller Road
Thousand Oaks, California 91320
SAGE Publications India Pvt Ltd
B-42, Panchsheel Enclave
Post Box 4109
New Delhi 110 017
British Library Cataloguing in Publication data
A catalogue record for this book is available
from the British Library
ISBN 0 7619 7146 7
ISBN 0 7619 7147 5 (pbk)
Library of Congress Control Number available
Typeset by C&M Digitals (P) Ltd., Chennai, India
Printed in Great Britain by The Cromwell Press Ltd, Trowbridge, Wiltshire
4/2/03 4:42 PM
Page iii
This book is dedicated to Beth, Cedric and Alice
4/2/03 4:42 PM
Page iv
4/2/03 4:42 PM
Page v
List of Figures
List of Tables
1 Introduction and Overview
Theory as testable explanation
A critical approach to accounting research
2 Developing the Research Idea
The research sequence
Emergence of the research topic
Conceptual frameworks
The structure of DNA: the development of new theory
The Bradman problem: the development of new strategies
The longitude problem: implementing solutions
Strategic management accounting
3 Theory, Literature and Hypotheses
Sources of theory
Searching the literature
Modelling the relationship
Developing the hypotheses
Validity concerns
4 Data Collection and Analysis
Sample selection
Measurement issues
Data management
Descriptive statistics
Differences in sample means
Measures of association
Analysis of variance
Multivariate model building
4/2/03 4:42 PM
Page vi
5 Research Ethics in Accounting
The ethics quiz
Informed consent
Ethical guidelines
6 Experimental Research
The problem statement
Theory and context
Experimental design
The validity trade-off
Quasi-experimental research
7 Survey Research
Mail surveys
Design and planning issues
Pilot testing
Data collection
Measurement error
Interview methods
8 Fieldwork
Case study methods
The qualitative analysis protocol
Grounded theory
Verbal protocol analysis
9 Archival Research
Cross-section data
Time-series data
The validity trade-off in archival research
Content analysis
Critical analysis
10 Supervision and Examination Processes
The role of the supervisor
Examiner profiles
The examination process
11 Turning Research into Publications
Why publish?
Where to publish?
What to publish?
How to publish?
Concluding remarks
4/2/03 4:42 PM
Page vii
Appendix 1: Ranking of Accounting Journals
Appendix 2: Sample Paper (1)
Appendix 3: Sample Paper (2)
4/2/03 4:42 PM
Page viii
4/2/03 4:42 PM
Page ix
List of Figures
The research sequence
The positivist approach
Alternative research methods
Kolb’s Learning Cycle
The deductive process
The conceptual schema
Measurement issues
The Harvey-Jones approach to problem-solving
Generalised process–improvement sequence
Searching for construct validity
Voluntary disclosure and reciprocal causality
Intervening variables and causality
Non-financial focus as an intervening variable
Moderated causal relationship
Size, industry and culture moderating performance
Influence of extraneous variables
Economic conditions as extraneous variables
Multiple independent variables
Modelling the recruitment process
Significance of test of proportion
χ2-test for difference in frequencies
Test of significance of correlation coefficient
t-test for difference in means
t-test for paired-case difference in means
Mann–Whitney U-test for difference in means
Product–moment correlation coefficient (Pearson’s r)
Coefficient of rank correlation (Spearman’s rho)
Measure of association within contingency tables
One-way analysis of variance (ANOVA)
The Kruskal–Wallis multiple sample test
Summary of regression results for charity
shops case study
4/2/03 4:42 PM
Page x
List of Tables
Table 1.1
Three alternative approaches
Summary of significance tests by measurement level
Measures of association by measurement level
ANOVA summary table
Summary of multivariate model-building methods
by measurement level
Table 10.1 Expectations of supervisor and candidate
4/2/03 4:42 PM
Page xi
This book could not have been written without the help of numerous
colleagues, most notably Bev Schutt and Paul Martin from the University of
South Australia. Thanks are also owed to Glen Lehman, Linley Hartmann,
Bruce Brown and Bruce Gurd at UniSA, and to David Russell, Ashok Patel
and Elaine Harris at the Leicester Business School. However, as is normal, all
errors and omissions remain the responsibility of the author.
4/2/03 4:42 PM
Page xii
This book aims to provide an insider’s view of the research process, by focusing
on actual choices made in the conduct of accounting research projects, together
with a realistic perception of what might go wrong, even with careful planning.
We must, however, acknowledge that no single author can be an expert in all
research methods; this author is no exception. My own publications will readily reveal a preponderance of studies concerning experimental methods and the
use of archival data; there are fewer instances of studies using survey and field
study methods. It would be unwise of me to claim expertise in the implementation of all such methods, so this book must necessarily lean heavily on the
work of others. For the same reasons, and because of pressure of space, this
book does not address issues of finance, capital markets research, or stockprice-related accounting research on the fringes of finance.
Most other texts in this area are long, over-theoretical and not particularly user-friendly. This book aims to address these issues by adopting a practical approach which takes the reader from the initiation of the research idea
right through to the publication of the research findings. The intended readership is wide, embracing instructors, doctoral candidates, and academics starting, or re-starting, their research careers. Although the focus, and examples,
are mainly accounting based, much of the material will also be relevant to
more general business applications, of particular interest to those pursuing a
Doctorate of Business Administration (DBA) qualification. The practical
examples employed are usually UK or Australia-based, these being the two
countries in which I have extensive, current experience of teaching, supervision and examining, but the principles should normally adapt easily to
alternative environments.
An early distinction between ‘methods’ and ‘methodologies’ in research
is essential because the two are so often confused, or else used interchangeably. Research methods are concerned with the technical issues associated
with the conduct of research; research methodology is concerned with the
philosophies associated with the choice of research method. This book is
almost exclusively concerned with the former and, after Chapter 1, deliberately neglects the philosophical foundations of research except where reference thereto is unavoidable.
Chapter 2 examines the research idea, and the documentary sources
which might aid their development. A number of examples, many from nonaccounting environments, are used to illustrate the research sequence, and to
examine research that is seeking either to improve outcomes, to explain
improved outcomes through new theory, or to examine the improvement
4/2/03 4:42 PM
Page xiii
process itself. Theoretical frameworks and research models are used extensively
here to help the reader to picture the key variables and relationships underlying their research.
Theory is the focus of Chapter 3, on the basis that ‘good research is
founded on good theory’. The chapter addresses the sources of the theory
widely applied in accounting research, but drawn from other disciplines. In
the space available it can only hope to give a flavour of the diversity which is
available; indeed, it prompts us to suggest that ‘theory in accounting
research’ might provide a suitable follow-up text in its own right! Recognition of the importance of theory, reliability and validity as desirable characteristics of accounting research lead, in Chapter 4, to the issues of data
collection, management and analysis necessary to conduct hypothesis testing.
This chapter is unashamedly quantitative in nature, but the relative strengths
of qualitative analysis are addressed in subsequent chapters.
Chapter 5 addresses the increasingly important ethical considerations
which underpin the conduct of accounting research, and the subsequent
publication of research findings. It highlights the confusion which is still
apparent among many academics as to what constitutes unethical conduct,
and specifies the necessary guidelines for good practice.
Chapters 6 to 9 are devoted, respectively, to the core forms of accounting research: experimental, survey-based, fieldwork and archival. Numerous
examples are used to demonstrate the relative advantages of alternative methods
so that researchers can both make an informed choice and justify their preferred approach. Research can be based on quantitative or qualitative methods,
and both should be equally acceptable as long as the most appropriate
method has been chosen. Richardson (1996, p. 4) notes that ‘work on how
science really gets done’ (such as that described in Chapter 2 based on Watson,
1968) shows that even though we are in an extreme ‘positivist’ domain, interpretive knowledge is still important in the development of new theory.
The majority of the readers of this text will likely be doctoral candidates
so Chapter 10 is devoted to supervisor–candidate relationships, highlighting
the mutual responsibilities of both parties to the supervision process, from
the outset right up to the examination process.
Publication is the natural target output of the research process, and
Chapter 11 addresses the complexity of the publication process. In doing so it
recognises that we are working in a dynamic process; what was once acceptable in accounting research is no longer so because of a more appropriate
emphasis on research ethics; what is publishable, at all or in specific journals,
changes too, both with the passage of time and the passing of particular journal editors. Many journals remain very conservative in the type of research
they will publish, often on the grounds that it is difficult to demonstrate that
‘new’ methods constitute ‘good’ research in the same way as the traditional
methods. But this situation is changing gradually – the wider opportunities for
publishing case-based research in recent years provides evidence of this. However, the renewed emphasis on journal and university rankings, and associated
funding systems based on the quality of publications, provides fresh difficulties.
4/2/03 4:42 PM
Page xiv
The provision of ‘acceptable journals’ listings by many universities, and the
prohibition of publication elsewhere, perpetuates the position of the wellestablished journals, while making it extremely difficult for the editors of other
journals to attract quality submissions. The opportunities for innovative new
journals are also severely diminished in such circumstances.
Contributions to the profession by academic accountants are generally
not well regarded, either by one’s colleagues or by government bodies providing funding based on publications performance, even though, arguably,
the education of the potential employers of our students might be seen as an
important part of our jobs. So journalistic pieces in practitioner magazines
and workshops to professional audiences count for close to nought – even
though the individuals concerned would never read a refereed journal or
attend an academic conference. We need to exploit the available media to get
our message, and the power of research findings, over to those implementing
change in an unbiased way, before the consultants get in on the act! This
process must be of mutual benefit to all parties, but if the practitioners feel
they are being short-changed, or even used, then future collaborative efforts
will be threatened. It is just such attitudes which generate the ‘them and us’
cultures leading to accusations of academics being out of touch with reality.
In this context it is interesting to note the changes taking place within professional journals: there were once two such journals called Management
Accounting, but now there are none, the US version becoming Strategic
Finance and the UK one Financial Management. With moves to term
‘accounting’ as ‘assurance services’ it will surely not be long before some of
the professional bodies themselves follow their journals with the removal of
the word ‘accounting’ from their titles.
Communication problems also remain. The timeliness and relevance of
much of the content of the refereed literature does little more than suggest
that it is written by academics, only for the consumption of other academics!
Most practitioners do not have an appreciation of research methods, nor do
they read the refereed literature, so important findings and recommendations
often do not reach the individuals who can make sure it has the greatest
impact. A number of journals have emerged with the express intention of
providing readable research for practitioner audiences, but even these have
tended to become more academic and less readable over relatively short time
periods. This book aims to provide a treatment of research methods that will
be of use to both accounting practitioners and those contemplating the conduct
of research projects.
Space restrictions mean that this slim volume cannot hope to tackle all
of the detail of the application of different research methods, or the associated intricacies of complex quantitative methods. But if its use causes one
paper to be published that would otherwise have gone unpublished, then all
will have been worthwhile.
Malcolm Smith
Introduction and Overview
• Theory as testable explanation
• A critical approach to accounting research
A number of authors (e.g., Brownell, 1995, p. 2) describe accounting
researchers as ‘parasites’ who prey on the work of others to generate their
findings. The term may be an overstatement, but as with most rash generalisations it contains more than a germ of truth: accounting researchers have
little theory of their own (they rely on economics, finance, psychology, sociology and organisational behaviour as their major sources); they have no
methods of their own (they are all adapted from the natural and social
sciences); and they have few instruments of their own (with many of these
originating in or adapted from the organisational behaviour literature). Merchant (quoted in Brownell, 1995, p. 140) even suggests that organisational
behaviourists are much better at developing survey instruments than their
accounting counterparts.
The overall aim of this book is to facilitate the conduct of applied
research studies in accounting, and to do this we must recognise our reliance
on work in other disciplines. To accomplish this aim, a number of subordinate
objectives may be identified, all of which will contribute to the overall goal:
• an understanding of contemporary research ideas in accounting, so that
readers can identify and define research problems and prepare strategies
for their solution;
• an awareness of alternative research methods, to facilitate the selection of
the most appropriate method for addressing particular research questions;
• an ability to review existing research and to offer critiques of articles
published in refereed journals; and
• an appreciation of the ethical constraints on the conduct of accounting
Research Methods in Accounting
Research in accounting is concerned with solving problems, investigating
relationships and building a body of knowledge. Because we rely to such a
great extent on prior research in the natural and social sciences to do so, this
volume will take a similar approach in leaning on work in other disciplines
where it helps to inform accounting research.
Bennett (1991) identifies four basic levels of research:
• Description – concerned with the collection and reporting of data related
to what is, or was, the case. This would include means and standard deviations of individual variables, and correlations between pairs of variables.
• Classification – still descriptive, but easing the reporting process, and
highlighting similarities and clustering through grouping and classifying
(e.g., through the familiar cross-tabulation facility in most basic statistical
• Explanation – an attempt to make sense of observations by explaining the
relationships observed and attributing causality based on some appropriate
• Prediction – going beyond the understanding and explaining of the prior
stage, to model observations in a way that allows testable predictions to
be made of unknown events.
We return to this structure in Chapter 4 when discussing alternative quantitative methods, but an early distinction between ‘explanation’ and ‘prediction’
is appropriate here, because, as in the natural sciences, we are able to make
excellent predictions of accounting behaviour without the backing of a sound
underpinning theory. Bankruptcy prediction modelling provides an excellent
example. A number of researchers (e.g., Altman, 1968; Taffler, 1983) have
developed models that have proved very successful in identifying ‘distressed’
companies – those companies that will fail in the short term. These models are
statistically excellent but the theory underpinning their content, in terms of the
ratios to be used and the variables they represent, is extremely weak; the essential problem is that such theories as we have (e.g., Blum, 1974; Myers, 1977;
Scott, 1981) do not generate very good predictive models!
Good research generates the sound evidence needed to overturn or revise
existing theories. These assertions will, in turn, yield to revised theories based
on better evidence, so that healthy competition between rival ideas will lead
to better explanations and more reliable predictions. Two major processes of
reasoning, ‘deductive’ (theory to observation) and ‘inductive’ (observation to
theory), are important for theory construction and observation testing.
Inductive reasoning starts with specific observations (data) from which
theories can be generated; a generalisable pattern may emerge from further
observations and repeated testing for compliance. The natural sciences, for
example astronomy, provide numerous examples of inductive reasoning, thus
Hawking (1998) provides a number of fascinating examples of theories
revised, or still in question, with implications for the progress of accounting
research. However, he notes that generalisations made on the basis of induction
Introduction and Overview
can never be regarded as ‘certain’, since just one contrary instance can cause
them to be overturned:
• Big Bang versus Steady State. From the late 1940s to the mid-1960s two
competing theories were prominent in offering alternative explanations of
the origins of the universe. The ‘Big Bang’ theory recognised a singular
event as causing an ever-expanding universe in which matter (notably
galaxies) becomes continuously more widely dispersed. The ‘Steady State’
theory, attributed to Bondi, Gold and Hoyle, on the other hand, suggested
that matter was continuously being created to fill the gaps between existing galaxies. They argued that the universe had no beginning, and had
been forever expanding, with new matter being created out of apparently
empty space. The Steady State theory importantly provided testable
hypotheses in suggesting that the universe should look the same at all
times and from wherever it was viewed. But surveys of radio waves in the
early 1960s showed that sources were more numerous in the past, and
that there were many more weak (distant) sources than strong (close)
ones. Further, microwave radiation studies in 1965 demonstrated that the
universe did not have a common density – it had been much denser in the
past. These observations provided disconfirmations of the Steady State
theory, causing its abandonment.
• Newton’s Laws of Physics. New theory emerges when a new observation arises which does not correspond with existing theory. Once the technology permitted accurate observations of the planet Mercury to be made,
it was clear that there were small differences between its observed motion
and that expected under Newton’s Theory of Gravity. Einstein’s general
theory of relativity matched the observed motions of the planet in a
manner that Newton’s theory did not, providing confirmation for the new
• The Wave Theory of Light. We can attempt to explain the behaviour of
light in terms of its being composed of either ‘waves’ or ‘particles’. Each
view produces a plausible explanation of behaviour – both of which are
needed to affirm existing properties – but they are incompatible explanations
which cannot exist simultaneously. New theories are required (possibly
those associated with parallel universes) for a complete understanding of
the incompatibility.
Deductive reasoning, on the other hand, starts with the theory and proceeds
to generate specific predictions which follow from its application. The predictions can be verified, or otherwise, from subsequent observation. For
example, in his seminal paper, Healy (1985) used agency theory to develop a
bonus hypothesis which could be substantially verified through observations
of how managers manipulated their accounting earnings to optimise their
short-term bonus performance.
Research Methods in Accounting
However, such a strict division of reasoning processes is not always
helpful because interdependencies almost always exist: induction will usually
imply some knowledge of theory in order to select the data to be observed (a
common criticism of grounded theory advanced in Chapter 8); deduction will
be dependent on the selection of the initial hypotheses for testing.
Even without such problems, the scientific position of ‘objective measurement’ has come under repeated attack, in both natural and social sciences,
because the act of observation is itself ‘theory-laden’ and influenced by the
motives and preferences of the observer. For example, Hopwood (1987), in
management accounting, and Hines (1988), in financial accounting, argue that
accounting helps to create the ‘facts’ that it is supposedly reporting. More radical approaches (e.g., Tinker and Niemark, 1987) suggest that accounting distorts practice in a systematic manner. Such concerns have aided the development
of new approaches: an interpretive perspective and a critical perspective.
• An interpretive perspective – From an interpretive perspective, human
actions are the result of external influences. These actions have both intentions and reflections, and take place within a structure of rules which
binds the participants. The task of the researcher goes beyond measurement to developing an understanding of the situation. To do this effectively, active participation, rather than detached observation, may be
required. Since the ‘action’ may be interpreted ambiguously when taken
out of context, this perspective places the fundamental emphasis on the
understanding of the process. In an accounting context, Arrington and
Francis (1989) provide an example of this approach.
• A critical perspective – The critical approach expands on the scope of the
interpretive approach by focusing on the ownership of knowledge and the
associated social, economic and political implications. An empirical approach
is criticised on the grounds that the research process is value-laden, and
that the acquisition of knowledge provides the opportunity to oppress
those being researched. In an accounting context, Tinker (1980) provides
an example of this approach.
Table 1.1 summarises the differences in research assumptions, process and
outcomes associated with each of these three major approaches.
Kuhn (1970) suggests that researchers are concerned with problem-solving
within a single framework of widely accepted beliefs, values, assumptions
and techniques. This shared framework, or view of the world, he termed a
paradigm, so that a ‘paradigm shift’ corresponds with some revolution where
the existing framework and theories can no longer cope with the volume of
disconfirming evidence. Kuhn neatly illustrates such a shift by reference to a
simple psychology experiment:
Subjects viewed cards from a deck. The deck included some unusual cards,
including black hearts and red spades, but the subjects were not informed in
Introduction and Overview
TA B L E 1 . 1
Three alternative approaches (adapted from Connole, 1993, p. 37)
What is the approach
modelled on?
Classical investigation
founded in the physical
What does it assume
about reality?
Reality is unitary and it
can only be understood by
empirical and analytic
methods, i.e., the scientific
What is the foundation of
Disciplined rules for
How is observation done?
Through clear and
unambiguous rules which
are not modified by the
setting and are totally
independent of it.
What is generated?
Evidence and generalisable
laws which are not
affected by contexts and
have nothing to do with
the way in which they
were discovered in the
first place. Objectivity
depends upon the removal
of error and bias which is
related specifically to the
logic of observation and
What interests are inherent?
Prediction and control,
technically exploitable
knowledge, and
What values are inherent?
Science and scientific
knowledge are inherently
Historical, literary and
existential studies in which the
subjective understandings of
subjects are significant.
Marxist and interpretive studies
which focus on the insights and
judgements of the subjects.
There are multiple realities
which require multiple methods
for understanding them.
There are multiple realities which
are made problematic through
distorted communication.
Meanings are the basis of data:
meaning precedes logic and fact.
Meanings are found in
language and social behaviour
and they precede logic and fact.
Through the social, linguistic
and cognitive skills of the
Interpretive methods, plus
critical self-reflection concerning
the grounds of observation.
Knowledge which is dependent
on the process of discovery.
The integrity of the findings
depends upon the quality of
the social, linguistic and
cognitive skills of the
researcher in the production of
data analyses and conclusions.
Knowledge which falls within
the interpretive framework,
but which also serves the
purposes of assisting personal
liberation and understanding,
and emancipation from forces
constraining the rational
independence of individuals.
Understanding at the level of
ordinary language and action.
Discovering the meanings and
beliefs underlying the actions
of others.
Interpretive interests and those
which underliey other forms of
inquiry. Radically improving
human existence. Practical
and public involvement in
knowledge formation and use.
Science and scientific knowledge
have both to be interpreted in
terms of values they represent.
Science and knowledge are never
value-neutral: they always
represent certain interests.
Research Methods in Accounting
advance about their presence. Initially the subjects saw only ‘hearts’ and
‘spades’, because they believed that only ‘red hearts’ and ‘black spades’
existed; only with repeated viewing did they grasp that these cards were not
typical of a normal deck. Then they could recognise the cards that existed
rather than the ones they were expecting.
In accounting research the parallels might be the paradigm shifts associated
with the ideas introduced by Ball and Brown (1968) and the difficulty they
had in getting a paper published which questioned the existing paradigm by
showing a link between stock prices and accounting earnings, through the
abnormal performance index. A similar, though perhaps less radical, movement is associated with Watts and Zimmerman (1978) and their popularisation of agency theory in an accounting environment.
What is inescapable is that we are dealing with people, and in the
research community that means individuals with their own agenda and with
reputations to build and protect. The natural sciences are littered with character assassinations of individuals and their work, by others who have been
less than willing to accept the impact of new findings on their own fiefdoms.
Sir Humphrey Appleby, in Lynn and Jay (1987), outlines the four stages
of the process necessary to discredit an unwelcome report. The parallels
between the fictitious Department of Public Administration and academia
are uncomfortable, where unwelcome findings might arise from academic
Refuse to accept the findings on the basis that they could be misinterpreted, and that a wider and more detailed study is required.
Discredit the evidence on the basis that it is inconclusive and the figures
are open to other interpretations, or that the findings are contradictory
and leave important questions unanswered.
Undermine the recommendations because they say nothing new, and
provide insufficient information on which to draw valid conclusions.
Discredit the researcher by questioning his or her integrity, competence
and methods employed.
We thus have doubts about the researchers, their research questions, their
research methods, the means of data collection and analysis, and the validity
of the interpretation and recommendations – all issues to which we will
Theory as testable explanation
Faced with a set of diverse observations, we can establish a set of tentative
explanations which help to make sense of the diversity. Such explanations
Introduction and Overview
constitute theory. In any set of circumstances there will usually be multiple
theories available to explain the observations. The systematic collection of
further data allows for the testing of the alternative theories so that we can
establish which of the exiting theories best explains the facts. A layman’s perspective of ‘theory’ is cynically expressed in Michael Crichton’s The Lost
World as: ‘A theory is nothing more than a substitute for experience put
forward by someone who does not know what they are talking about’ (1995,
p. 67).
The data collection itself allows only a descriptive approach (e.g.,
means, standard deviations, ranges, correlations); we cannot attempt to
attribute causation in any meaningful way without recourse to an explanatory theory. We are always looking for another theory which may fit better,
so that, as Popper (1959, p. 104) suggests, a ‘genuine test of a theory is an
attempt to falsify it or refute it’. We look for disconfirmations rather than
In the short term this may not be successful. In accounting, we witness
the frequent and numerous ‘anomalies’ to which the Efficient Markets
Hypothesis (EMH) is subject, but we have no other widely accepted theory
of the manner in which stock prices react to the availability of relevant
Popper’s suggestions are very attractive in providing a powerful empirical methodology for subjecting theories to attempts to refute them. However,
this position is not always ideal because the process of ‘observation’ in itself
may be fallible. Thus Hawking (1998) reports Heisenberg’s Uncertainty
If we are to predict the future position and speed of any particle, then we
require accurate measurement of both its present position and current speed.
Heisenberg did this in 1926 by shining light on a particle, and observing the
resultant scattering of light in order to reveal its position. However, to determine the position of the particle accurately an amount of light needed to be
used which changed the speed of the particle in an unpredictable way:
the more accurately he tried to measure position, the less accurate was the
measurement of speed!
The Uncertainty Principle has wide implications for research conducted in
any environment, where it is impossible to measure the size and speed of a
particle without altering all other characteristics in the process of measurement. We have a parallel situation in accounting research where the actions
of the participants in ethnographic, experimental, survey or fieldwork
impacts on the outcomes of the measurement process.
Three fundamental criteria exist to judge whether theory fits observation:
Co-variation – even where no causality exists we would expect the two
variables to move together so that a high degree of correlation exists
between the two variables. Where there is no co-variation it will be
difficult to establish a causal link.
Research Methods in Accounting
Cause prior to effect – if a causal link is to be established then the ‘causal
event’ should occur before the ‘effect event’. The sequence of events can
therefore help to establish an explanatory direction.
Absence of plausible rival hypotheses – the third rule seeks to eliminate
alternative explanations of the events as being implausible. This may
only be possible in the present, because future researchers may develop
competing explanations of the events from a re-analysis of the data.
Consider, for example, the voluntary disclosure of information in corporate
reports and analyst following (i.e., the number of analysts examining the
performance and reporting on the disclosures of large companies). There is
a relationship between these two variables – they co-vary: the volume of
voluntary disclosures and the number of analysts reporting move together.
But which is causing which? Rival hypotheses suggest:
companies are supplying more information voluntarily to the market to
signal their intentions and reputation, attracting the attention of more
investment analysis;
investment analysts are focusing their attention on particular companies
and demanding more information and more detailed disclosures.
The existing empirical evidence is less than convincing: Lang and Lundholm
(1996) find (a); but Walker and Tsalta (2001) provide only weak evidence for
(a) but stronger evidence to support (b). Clearly more empirical work is
required to clarify the direction of causation.
A critical approach to accounting research
Researchers must demonstrate a healthy scepticism towards both their own
findings and those of other researchers. They must adopt a critical posture,
questioning everything that they read until sufficient evidence has been provided for them to be satisfied with the quality of the outcomes. The development of critical appraisal skills is a fundamental requirement in researchers,
so that they can distinguish between good and bad research, and clearly identify flaws of argument, methodology and analysis.
Honest and transparent reporting of research practice is an ethical duty
of those participating. Researchers should report everything that they did,
why they did it, and how they did it. If they have doubts about any stage of
the procedure, then these should be stated, along with their likely implications and what, if anything, has been done to overcome these doubts. Where
researchers have been ‘economical with the truth’, this is usually apparent in
their papers and is often an indicator of bad research.
Introduction and Overview
Students frequently struggle initially when they are asked to critique
published articles. They are often in awe of the reputation of the authors, or
doubt whether they are able to offer sensible criticism of papers which, after
all, have already undergone editorial scrutiny and double-blind review.
Despite the above, some flawed papers do get published, and these are not
always in lower-tier journals (see Hartmann and Moers, 1999, for their critique of 28 papers on contingency analysis in three top accounting journals –
Accounting Organizations and Society (AOS), The Accounting Review (AR)
and Journal of Accounting Research (JAR) – in which they identify problems
in the design and analysis of 27 of the studies!). With appropriate guidelines
as to the appropriate questions to ask, students can quickly develop
some confidence in their ability to spot flaws and omissions. For example,
Abernethy, et al. (1999) provide a stimulating critique of the three subsequent papers in the same, outstanding, edition of the journal Accounting &
We would usually want to address the following:
Why is this article interesting/important? The paper must offer some new
insights which constitute a contribution to knowledge. These insights
should be non-trivial, so that they can be embraced in either further
theory development or recommendations for improvement.
Are the outcomes important? Effectively, does the paper pass the ‘so
what’ test? Will anyone be interested in the outcomes of this research, or
will it have any implications for future practice? Would the scope of the
research be well regarded by competitive grant authorities? This has
important implications for those papers which produce ‘negative’ findings, that is, they test reasonable hypotheses based on the research literature, but their datasets fail to support any of the expectations. These
findings still make a contribution in that they demonstrate that findings
from elsewhere (often other disciplines) do not hold in accounting, but
their negativity may restrict their publication opportunities.
What motivates the authors to write this article now? The paper may be
clearly addressing issues of contemporary concern; on the other hand, it
may be addressing more historical issues and/or be using ‘old’ data. If we
have the latter, we may be dealing with an old paper recently recycled, or
a paper which has been through many iterations at several different journals before being deemed ‘publishable’.
What is the research problem/question? We are looking for a clear statement of the problem very early on in the paper, so that its objectives are
readily apparent. If we reach page 11, say, of the paper without a clear
idea of its direction, or any sort of research model, then perhaps the
authors need to readdress the fundamental purpose of the research.
Research Methods in Accounting
What theory or theoretical framework underpins the research? Without
some theoretical foundations we have a problem-solving exercise or a
consultancy project, neither of which should be gracing the pages of a
refereed journal. There must be some theoretical justification for the
question being addressed and the research approach adopted. Theory
will often not come first in the research process – it will frequently be preceded by an interesting idea or a perplexing observation. But we require
some theoretical explanation for the relationships under investigation
before we have the basics of a refereed journal article. Observed deficiencies in this area usually fall into one of four categories:
the underlying theory is either non-existent or extremely thin;
the theoretical context is there but appears to have been tacked on as
an afterthought – usually at the beginning of the paper and often
written by a co-author. Examination of writing styles suggests that we
frequently do not have a seamless divide between ‘theory’ and
‘conduct of research’;
the theoretical arguments are unconvincing, so that there are
competing theories that may reasonably have been adopted in the
paper but have been overlooked;
a sound theoretical framework but findings which are totally at odds
with theory. Apparently, a competing theory may be more appropriate,
although this is unknown to the authors at the time.
What are the key motivating literatures on which the study depends?
There will normally be a small number of seminal pieces of literature
which are driving the research. If any of these are themselves unreliable,
it may cast doubt on the state of the foundations on which the paper is
based. If one of the papers is an unpublished conference or working
paper from several years before, then alarm bells ring to question why
that piece has not itself been published in the refereed literature. If key
seminal pieces of literature have been overlooked, then again the integrity
of the findings is reduced.
Which research method has been chosen? There should be a justification
for the chosen method, and a clear preference over alternatives. The
method should be consistent with both theory and literature and, ideally,
prior empirical studies in the field will have adopted similar methods.
Most importantly, we want to see a research method that has evolved
rather than one that has been selected first, even before the research question has been fully developed. The use of survey methods should always
be questioned in this way since frequently they seem to have been selected
without explanation of the elimination of alternatives. Ideally, we should
be able to trace through the emergence of abstract concepts, from theory,
through their operationalisation and measurement, so that any hypotheses are entirely consistent with both theory and literature.
Introduction and Overview
How has the sample been selected? Details on sample selection are often
sketchy in many articles, perhaps because the authors feel vulnerable
about the procedures adopted. Sometimes (see, for example, Young,
1996) the actual sample size employed is omitted, as is the response
rate. Both omissions should be regarded as bad news. It is usually clear
that scientific methods have not been adopted (unfortunately far too
commonly in accounting research) where there is an over-reliance on
convenience samples. What may be apparent is an attempt by the authors
to obfuscate in this regard, to overlook detail and try to create an impression that the sample selection is more systematic than it has actually
How have questions of validity been addressed? Choice of research
method should address issues of validity. Where experimental methods
have been employed we would anticipate questions of internal validity to
be paramount; where field studies are involved we would expect issues of
external validity to be addressed. For survey methods we would anticipate the focus to be more on the reliability of the test instrument and the
rigour of the subsequent statistical analysis, rather than on validity
How have the results been analysed? We want to see the simplest analysis of the results consistent with the relationships being explored. We do
not wish to see unnecessary complexity; this will make the paper less readable and tend to mask the findings and their significance. On the other
hand, most academic accountants are only ‘amateur’ statisticians; if the
level of their analysis is inadequate, then they may need to bring in a
statistician as co-author (evidenced by the number of ‘quant jocks’
appearing as third or fourth authors on accounting papers to satisfy the
reviewers). Importantly, we do not wish to see the method of analysis
driving the study. In just the same way as the research method should not
precede the research question, then neither should the method of analysis.
For example, I recall a paper of my own (M. Smith, 1992) presented at a
conference but never published. It attempted to show the advantages of
using multidimensional scaling (MDS, then a little-used technique in the
accounting literature) for problem-solving, but the journal referees rightly
observed that the method was inappropriately sophisticated for a relatively simple research question. MDS was abandoned, simpler methods
instituted and the revised paper eventually published as Smith (1996).
Are the conclusions and recommendations consistent with the findings?
Effectively, does the paper hold together? Is the title appropriate? Do the
abstract and introduction lead us to expect what we find at the end of the
paper? In many papers the final sections are the weakest and may not do
justice to the breadth of the research conducted. We look for explanations, limitations and a future research agenda.
Research Methods in Accounting
Let us now consider how this framework may be applied to a critique of a
published piece. Naturally, I chose one of my own publications (Smith et al.,
2001) for the treatment because a knowledge of the history of the development of the paper, from an insider’s perspective, can be most instructive.
Readers will be able to make the most of the subsequent discussion if they are
first able to read a copy of the paper, and for this purpose the complete paper
is reproduced as Appendix 2.
Interesting new insights: The paper posits an interesting connection
between (1) audit firm; (2) manner of conduct of the audit; and (3) classification of audit firms based on their procedures and culture. The paper
also attempts to impose a global perspective by employing findings from
the USA, the UK and Australia. But neither the data nor the supporting
literature is new, and it compromises the originality of the paper.
Importance: The paper is important if it makes a contribution to knowledge. This may be a contribution to theory development or implications
for business practice. If the paper can demonstrate a relationship between
‘auditor’ and the manner in which the audit has been conducted, then
this makes a contribution, even though it may only be of historical relevance. Such a relationship is shown for 1987/88 data, but evidence is also
presented to suggest that this relationship no longer holds. The absence
of a current relationship suggests that the paper has no implications for
current auditing practice. The reasons why a relationship between the
audit firm and its propensity towards tolerance of particular accounting
policies among its clients is by no means clear.
Motivation: The timing of the paper is problematic. It is published in
2001 but uses data predominantly from 1987/88. There is a danger of its
being regarded as a historical piece with little relevance to current practice. The authors justify the use of this dataset in that the Kinney classification, the target test of the paper, is based on data relating to the Big
8 group of accountants, with 1988 being the last year of existence of the
Big 8 in Australia, prior to extensive merger activity in the sector. There
is the suggestion, though, both from the paper itself and the references
cited, that the data have been used primarily to generate failure prediction models for the Western Australian government (i.e., Houghton and
Smith, 1991) and that the further use of this data in this paper may be
incidental and opportunistic.
Problem statement: The problem statement is quite clearly stated as:
Accounting Policy Changes = f {auditing firm}, where both sides of this
equation are elaborated and measured for a large number of companies:
Accounting policy changes: discretionary/mandatory; income
increasing, income reducing, neutral.
Introduction and Overview
Auditing firm: by individual name, and by grouping according to
classifications developed by Kinney (1986) and Moizer (1998).
A number of extraneous variables (notably firm size, financial
performance and industry) are also examined to determine their impact.
Theoretical framework: This remains something of a problem with
this paper, despite strenuous efforts to overcome omissions. The literature demonstrates that there are differences between auditors, and
in the procedures that they adopted for audit in 1988 (i.e., Cushing
and Loebekke, 1986; Sullivan, 1984). However, why these procedural differences between auditors translate into differing tolerances
towards income-impacting accounting policy changes is unclear, and
is largely attributable to unpublished anecdotes from practising
auditors and the discussion arising in a single paper (Dirsmith and
Haskins, 1991).
Motivating literatures: Relatively few articles, noted above (i.e., Sullivan,
Kinney, Cushing and Loebekke, Dirsmith and Haskins) motivate this
paper, while Terry Smith (1992) and Peter Moizer (1998) provide the
opportunity for UK comparisons. The pivotal paper is Dirsmith and
Haskins (1991), published after the conduct of the data collection; there
is thus a strong suspicion that interesting findings have arisen from data
mining operations in 1988, for which Kinney (1986) provides a conceptual framework, but that publication must wait for a suitable theory.
There is very little other supporting literature, though self-citation by the
authors is also revealing:
• Houghton and Smith (1991) relates to failure prediction models
constructed with the same data and is employed here to provide a
measure of overall financial performance;
• Smith (1998a) reports current UK findings linking auditor with attitude
to accounting policy change;
• Smith and Kestel (1999) update the present study with a time series
analysis, but the results are apparently insufficiently interesting to
constitute publication in a refereed journal;
• Brown (1988) reports on the most appropriate means of conducting
statistical tests with contingency tables.
Research method: Archival methods are employed, since they are the
only realistic alternative given the nature of the data: namely, historical, documentary, and covering many companies that are no longer in
existence. The authors’ access to a dataset comprising the population
of Western Australian public companies is a considerable strength of
the paper. Data collection is meticulous and involves checks for consistency both between individual researchers and for temporal validity.
Research Methods in Accounting
Sample selection: The paper accesses the annual reports of all 463 publicly
quoted companies in Western Australian (WA), so does not encounter
any sampling issues other than a restriction on the nature of statistical
tests that may be employed because of using a population rather than a
sample from a normal distribution.
Validity issues: There are potential internal validity threats consequent
upon the failure to consider competing theoretical explanations for the
observations. The incidence of accounting policy change is apparently
associated with auditing firm, but both the direction of causation for the
relationship and alternative auditor motivations might be considered.
The authors acknowledge the lack of external validity in the study – the
applicability of the findings to other time periods and other datasets – in
that conditions have changed so substantially since the data collection
period that the procedures adopted by all auditors are now very similar.
Analysis: The fundamental analysis is relatively unsophisticated, involving
the comparison of ‘observed’ and ‘expected’ frequencies through a chisquared test. A variation on the traditional approach is introduced to take
account of an ordering effect in the contingency tables, the power of the
tests being increased with the use of Kendall’s-tau. (A co-author with specialist statistical publications has been included to address testing issues,
potentially in response to reviewer concerns on previous versions). A comparative fundamental analysis for UK data (alluded to in Smith, 1998a) is
apparently not possible, and further analysis is restricted to tertiary sources.
Conclusions: There are no formal conclusions or recommendations,
rather a discussion of other interesting findings in related fields which
may impact on the integrity of the outcomes. The findings of this study
are linked to merger activity in the Big 8, showing a pattern with considerable similarities to past successes. The paper suggests that future
merger activity in the sector may be influenced by the organisational culture aspects of the Kinney classification and the clustering of companies
generated by Moizer (1998); thus if we were looking at potential suitors
for Arthur Andersen, say, then the analysis suggests that Ernst & Young
would provide potentially the most successful alternative.
Such a critique is revealing, giving glimpses of a less-than-optimum approach
adopted in the development of this particular paper. Data were collected
for the specific purpose of generating failure prediction models for the WA
government, and corporate monitoring of distressed enterprises (i.e.,
Houghton and Smith, 1991). The interesting auditor findings were generated
at the same time, but there was no substantive theory to justify the observed
relationship – and consequently no research paper. Only with the emergence
of new theories (e.g., Dirsmith and Haskins, 1991), which might motivate the
study, could further development towards a publishable paper proceed.
Introduction and Overview
Clearly research is not always simple, systematic and clean – despite the
sanitised versions that we read in the published journals. The research process
can be both chaotic and exciting, and very rarely proceeds exactly according
to plan. Unfortunately, this impression is rarely created by what we read
because published pieces usually have happy endings – positive findings and
co-operative participants. For a more realistic version of events we must rely
on books like this, conference presentations and research workshops!
Armed with a critical and sceptical approach to the research of others,
we can now start to develop the skills required to conduct competent research
of our own, and commence a sequence which will eventually result in the
publication of our research findings.
Developing the Research Idea
The research sequence
Emergence of the research topic
Conceptual frameworks
The structure of DNA: the development of new theory
The Bradman problem: the development of new strategies
The longitude problem: implementing solutions
Strategic management accounting
We recognise in Chapter 1 that research processes are usually neither simple,
systematic nor clean because research rarely proceeds exactly to plan. However, this should not deter us from planning thoroughly in the first place to
specify how, in an ideal world, we would like the research to be conducted.
The research sequence
Figure 2.1 specifies the typical research sequence described by Howard and
Sharp (1983) as a series of stages we would expect to progress through in
most forms of accounting research, while moving from original idea to eventual publication.
The following chapters of this book address each of these stages, and
detail the constraints we might anticipate.
Identify broad area: Narrow the focus from accounting in general to a
stream associated with financial accounting, management accounting,
auditing, accounting education or accounting information systems.
Developing the Research Idea
Identify Broad Area
Select Topic
Decide Approach
Formulate Plan
Collect Information
Analyse Data
Present Findings
The research sequence
Select topic: Specification of a sub-area to provide a tighter focus, and
one for which supervision capacity is available, but one which may be
modified in the light of subsequent developments.
Decide approach: Early thoughts regarding the approach to be adopted
will revolve around the resources available, and in particular access
to the necessary data sources. A detailed specification of research methods to be adopted must wait until the literature review has been
conducted and theoretical foundations and outline hypotheses have been
Formulate plan: Milestones and targets should be established at the outset so that it is clear how the research will progress over an extended
period. This is particularly important for part-time researchers who may
be contemplating study over six or seven years. A Gantt chart, or similar, is very helpful in clarifying the extent of the programme of work and
the mutual expectations of those involved, especially if this concerns the
relationship between candidate and supervisor in doctoral work. This
plan should include target conferences where preliminary findings may
be presented, especially where deadlines are important to the candidate.
The commencement period can often cause anxiety because of the
Research Methods in Accounting
perceived need to make swift progress. It cannot be emphasised enough
how important an extended period of reading is to ensure that effort is
not wasted performing experiments or surveys which subsequently
emerge as being unnecessary or fatally flawed.
Collect information: Data collection can safely proceed only when we
recognise exactly what we want to know, and for what purpose. The
planning stage should highlight the period over which we want to collect
data; this usually effectively precludes most longitudinal studies, partly
because it takes too long to collect data and partly because of the
increased vulnerability associated with extended site access. We may
require access to commercial databases; if these are an essential requirement then permissions should be sought immediately.
Analyse data: Methods of data analysis and software requirements
should be apparent early in the research process.
Present findings: Preliminary findings will normally be presented at
university workshops and seminars, and then at specialist conferences.
These provide the precursor to publication in the refereed literature,
which may take place before completion of any associated doctoral dissertation. Publication in the professional literature will bring important
findings to the attention of interested practitioners.
The research sequence above can easily be squeezed to provide the elements
of the traditional positivist approach as devised by House (1970), and illustrated in Figure 2.2. Again this approach assumes the presence of specific
• the specification of a priori hypotheses: formulated on the basis of
theory and literature before any data is collected or fieldwork contemplated;
• the specification of a priori criteria: to measure the acceptability of
the hypotheses, most commonly in the form of standard statistical
• the isolation and control of the variables to be investigated: determination
of which variable(s) will be treated as dependent, which will be independent (explanatory), and which will be held constant, matched or ignored;
• the verification of the methods for measuring and the variables: specification of which variables can be measured directly, and how, and those
which will require the use of proxy variables, or measurement instruments, of some form.
However, we have to acknowledge that there is no single method which
necessarily applies to research in all situations. Thus while the positivist
tradition remains the most prominent in the accounting literature, non-positivist
approaches have become increasingly acceptable. (Even so, some of the top
Developing the Research Idea
Literature Review
Hypothesis Development
The positivist approach
US journals are still unmoved in their attitude to non-positivist approaches.
Baker and Bettner (1997), among others, observe that most of the top
journals are devoid of interpretive and critical research studies.) However,
management-oriented investigations of change (e.g., the implementation of
accounting innovations) may be particularly unsuitable to a scientific
approach. Where people are involved and multiple variables are beyond the
control of the researchers, including management’s own motivation and
agenda, scientific approaches are of questionable validity. Checkland (1981,
p. 316) observes: ‘attempts to apply scientific methodology to real world,
essentially social problems, have been responsible for the limited success of
management science’.
Thus we can stay within our original ‘research sequence’ framework but
extend out beyond the positivist approach. Figure 2.3 illustrates the range of
possibilities. As we move from top to bottom in the figure, we move from the
traditional positivist approaches (archival and experimental studies) through
field studies towards a case-based approach typically associated with ethnographic studies. This movement corresponds with an increase in the number
of uncontrolled variables, with our increasing inability to formulate testable
hypotheses, and with the increasing prominence of the ‘human’ element.
Research Methods in Accounting
Archival Research
Laboratory Experiments
Survey Methods
Action Research
• hypothesis testing using standard
• quantitative data
• deductive positivism
• quasi-research methods
• field-based surveys/interviews
Field Research
Ethnographic Methods
• subjective accounts ‘inside’
• no a priori hypotheses
• qualitative data
Alternative research methods
Emergence of the research topic
We should begin by choosing a research topic which is of interest both to the
researcher and the supervisor, where the project is contributing to a doctoral
qualification. The topic should generate enthusiasm in the researcher at the
outset, otherwise he or she is unlikely to last the course of a protracted period
of study in which motivation is bound to wane, even temporarily. The source
of the topic can be from anywhere, but most commonly:
• a problem at work with potentially wider implications;
• a problem or application spotted in the newspaper or from television;
• a conference presentation revealing the directions being explored by other
• working papers and completed theses elsewhere – the contents of which
are usually at least two years away from publication;
• textbooks – particularly in management-related areas – which are a
constant source of untested theories;
Developing the Research Idea
• review articles, analysis of the literature in a particular area, to reveal the
current boundaries of knowledge and a potential research agenda. The
Journal of Accounting Literature is a particularly useful source in this
regard (e.g., Bauman, 1996; Dunk and Nouri, 1998; Gramling and Stone,
2001; Jones and Shoemaker, 1994). Also good is Accounting Organizations and Society (e.g., Hartmann and Moers, 1999; Langfield-Smith,
1997; Libby and Luft, 1993).
• Review monographs (e.g., Ashton and Ashton, 1995, information processing; Brownell, 1995, management accounting; Smith and Gurd, 2000,
behavioural issues; Trotman, 1996, auditing) are also helpful, as is Foster’s
(1986) excellent book, which unfortunately has never been updated since
this second edition.
• Refereed journal articles, particularly the final sections, revealing flaws in
existing research, gaps in our knowledge and research opportunities.
The ideas will rarely emerge, therefore, from a ‘spark’ of original thought.
More likely, the thought development will have emerged elsewhere, with the
originator having either discarded his or her ideas or not seen their full value.
It may fit an ‘added value’ concept in the same way as a successful innovation may be remote from the inventor. This may be in the form of relating
two concepts from different disciplines, in a manner which provides an application opportunity in the accounting environment. Here the ‘success’ is the
publishing of research findings in a respectable journal. To do so we will
inevitably be building on the work of others.
The common element in each of the above approaches is ‘reading’ – hence
the common advice given to doctoral candidates of ‘reading, reading and yet
more reading’ to know an area and spot the opportunities. Candidates usually
have a much greater commitment to a topic if they have developed it themselves, yet many find idea-generation an extremely difficult process. Thus it is
not uncommon for the supervisor to be the source of the research idea,
because active and experienced researchers usually have far more ideas than
they are capable of exploring by themselves. As Gill and Johnson (1997, 2002)
observe, topic selection can be risky if left entirely in the hands of the candidate; the chosen topic may prove to be too small, too large, or simply not
feasible in the time frame (especially for longitudinal studies). There may need
to be a trade-off between the ownership/commitment associated with a studentselected topic and the practicability/timeliness expected of a supervisordirected preference. Commonly, candidates will contemplate studies which
involve the implementation of an accounting change, in order to monitor the
change process and the resulting impact on financial outcomes. Such a model
is rarely feasible because, apart from the access problems, involvement and
data collection will be necessary over a period usually extending beyond that
permitted within a standard candidature.
Once the germ of an idea is forthcoming it must be worked over to see
if it really constitutes ‘research’. For example, are we sure that it is more
than a consultancy project? Is it more than a trivial problem with no wider
Research Methods in Accounting
implications? Is it more than a replication of something someone else has
done before, or done in a different industry or different country? If we are
happy in this regard then several other questions emerge:
• Is the project ‘doable’ in a reasonable time frame (e.g., the period of
• Will the project fit the NIRD acronym (usually attributed to Rashad
Abdel-Khalik during his tenure as editor of The Accounting Review) in
that it is new, interesting, replicable and defendable? Unsurprisingly, the
acronym fits a positivist outlook since replicability may be impossible to
guarantee in field study settings.
• Will the data required be readily available? If site visits are required, will
access be available over a sufficient time period and to a sufficient depth?
This last scenario is of great practical concern and difficult to control.
Young and Selto (1993) report on a study whose depth was seriously
curtailed because management changed its mind and restricted the access
to personnel due to be interviewed. Worse than that, there are numerous
cases of change of company ownership during the data collection period,
resulting in further site access being permanently denied.
Once the general topic area has been determined, it may be refined by formal
methods (e.g., brainstorming, attribute listing, etc.) to identify possible fruitful directions and potentially interesting relationships, and to eliminate blind
alleys. Diagrammatic aids, particularly white-boards, are very helpful at this
stage for mapping ideas, variables, relationships and processes.
If you are in any doubt as to what constitutes ‘research’ and what constitutes a study that is published in a reputable refereed journal, then the
reader is referred to recent examples from the professional literature. Often
the professional journal will have interesting ideas and will convey a message
with implications for practitioners (its essential purpose), but will not have
the essential conceptual framework of a refereed research piece. Consider, for
example, the contribution to the professional literature made by Smith and
Briggs (1999). The paper is both amusing and interesting in making observations in a number of areas, and is reproduced in its entirety as Appendix 3. It
• the traditional stereotype of the accountant as a boring individual with
poor interpersonal skills;
• the portrayal of the accountant in the media (particularly in films), initially as a comic incompetent, but latterly as an unethical, unprofessional
manipulator with criminal tendencies;
• the accountant, almost exclusively, seen as a white male occupying a subordinate position;
• the concern in the accounting profession about its failure to attract the
best candidates, who are lost to other professions;
Developing the Research Idea
• the concern that the characteristics which the profession tries to convey
about accountants (financial expertise, interpersonal skills, ethical conduct, openness to all) are at odds with those conveyed by the media.
The paper speculates that there might be some link between these issues and
suggests that recruitment to the profession is being damaged by the accountant’s image. However, it provides no evidence to support this relationship,
nor does it provide any evidence for the manner in which this image may be
being influenced by the popular media. Furthermore, there is as yet no theoretical explanation to justify these relationships. Without these foundations
we have an interesting idea but no basis for a research project whose findings
would interest the refereed journals. However, the idea could potentially
yield rewarding research projects, were answers to fundamental questions to
be found, and these are considered in more detail later in this chapter.
Attention to the fundamental requirements of the refereed literature will
allow the researcher to produce an outline research proposal, one that is continually revised during the reading period and may nevertheless have to be
revised further during the conduct of the research itself due to unforeseen
circumstances. A typical research proposal will include the following elements:
Title – should make it clear what you are trying to do.
Abstract – should summarise the problem, objectives and expected outcomes.
Issues – why they are interesting and important.
Objectives – how the study relates to the problem.
Literature – review of relevant, themed publications.
Method – the how and why of the process.
Benefits – the anticipated outcomes that make this study worthwhile.
A working title is important to clarify the topic, especially if external grant
income is being sought to support the project. However, the final title is rarely
the original one and there are plenty of opportunities to make changes.
The abstract is an important departure at this stage, because it allows the
researcher to speculate on what the outcomes of the research might be should
everything go according to plan. The abstract can be ambitious initially, but
will require revision (perhaps radical revision) as problems and constraints
emerge in the research process.
The contribution of the paper and the way that it aims to address important issues in a systematic manner are fundamental to its success. Internal
consistency and overall coherence should ensure that the objectives and the
intended approach are appropriate.
The outline literature review may be incomplete but it must none the less
identify the key motivating literatures and theories. Research candidates
must recognise that the literature review is a constantly moveable feast and
something that will be added to right up to final presentation of the research
findings. One of the major deficiencies of both papers and dissertations is
Research Methods in Accounting
that they frequently overlook the most recent relevant publications: it can be
a heart-stopping moment when one is about to submit and the latest issue of
a journal appears to report the outcomes of a research project very similar to
one’s own! At the very least this new paper must be cited. A common complaint from inexperienced researchers is that there is ‘no literature’ available.
If this is true, it may mean that the projected topic may be too trivial to consider. More likely, however, is that the literature review should drill down
further and search on different keywords. That there is a dearth of recent
literature on a topic may foreshadow problems. For example, papers on ‘decision-making heuristics’ were common in the late 1970s and early 1980s, and
papers on ‘group decision-making’ common in the mid-1980s, but progress in
both of these research areas has slowed, and publications are relatively rare
because the psychological theories underlying the research have not developed
sufficiently to facilitate new approaches. With respect to apparently new
projects, for example, research into ‘e-commerce’ related topics, students must
recognise that e-commerce is just a new way of doing business, and that their
review must address the implementation of prior business innovations.
Discussion of method should address the alternatives available in order
to demonstrate that the preferred choice is the most appropriate. The proposal should also echo the benefits of the research, in particular its contribution to knowledge and the potential implications for business practice.
The research proposal would normally form a central feature of any
application for ethics approval, and must therefore demonstrate both the
value of the research, the integrity of the methods employed and the extent
of the involvement of human participants.
Conceptual frameworks
A valuable part of the initial planning process is the development of a conceptual representation of the research project. This can help to clarify the
important relationships (and the need for supporting theory), the explanatory
and intervening variables, as well as the demonstration of causation.
The inductive and deductive approaches identified in Chapter 1 provide
an objective alternative to the conduct of research, but neither allows the
opportunity for human interaction: the inductive approach, where new
theory is developed on the basis of fresh observations (as is most commonly
the case in hard sciences, like astronomy), and the deductive approach, where
theory provides the basis for the testing of empirical observations (and which
is the most common form of positivist accounting research). The deductive
approach is suitable in a highly structured environment, involving the empirical
testing of theoretical models, so that its reliability is dependent on the integrity
of quantitative and statistical methods. However, the causal relationships
Developing the Research Idea
Concrete Experiences
Testing Implications
of Concepts in New
Observations and
Formation of Abstract
Concepts and
Kolb’s Learning Cycle
explored rely on an internal logic and take no account of the human
relationships present. The application of the inductive approach in the accounting environment necessitates a variation to the traditional model such as that
provided by Kolb’s Experiential Learning Cycle (Kolb et al., 1979, p. 38),
which is illustrated in Figure 2.4. Recognition of the importance of internal
processes and human relationships to the inductive approach allows for the
existence of human subjectivity without distorting research findings, even
though they may be qualitative and not replicable. Where human relationships are central to an understanding of accounting behaviour the approach
exploits the subjective environment.
Although both models provide opportunities in accounting research, the
deductive approach offers greater possibilities for the implementation of
scientific methods, since it facilitates arguably more reliable measurement
and control. Grounded theory (discussed in more detail in Chapter 8) illustrates the potential for inductive methods in accounting research.
We can therefore develop the model of the deductive process (see Figure 2.5)
so that it corresponds with Popper’s (1959) defining characteristics of scientific
• the theory is capable of empirical testing;
• scientists (and researchers) should make rigorous attempts at falsifying
• science advances as falsified propositions are left behind, leaving a core of
theory still to be disproved.
Research Methods in Accounting
of Concepts
Test Theory
Falsified, So
Discard Theory
Not Falsified, So
Predict Future
The deductive process
The conceptual schema
The basic conceptual schema (see Figure 2.6) provides a powerful tool
for the examination of causal relationships in a positivist environment. By
establishing the key variables of interest, and the other potentially influential
factors, we can form a better impression of the breadth of the problem.
If, for example, we return to our earlier article on ‘accounting stereotypes’, we can see how the content of a professional piece may be used to
develop a conceptual schema for more systematic investigation. As a starting
point we need to identify dependent variable(s) – one or more outcome variables whose value may be influenced by a number of (independent) explanatory factors.
The scenario outlined extends over a long period of time – potentially
from the final years of secondary school through undergraduate years to
recruitment to the accounting profession; quite possibly in excess of five years
even for straightforward cases. The length of this period, the process of maturation and the incidence of a multitude of uncontrolled variables makes it
extremely unlikely that we will be able to predict behaviour in year five based
on attributes in year zero. In trying to establish a causal relationship for
investigation, it is much more realistic, therefore, to break up the process in
Developing the Research Idea
order to be able to focus on events during more finite time periods. Such a
reorientation allows us to focus on two ‘choices’ which will permit the
suggestion of two measurable dependent variables:
• the selection of commerce/accounting as the preferred undergraduate
course by school leavers; and
• the decision to enter the accounting profession by those nearing the end of
their university course.
The major dependent variable, in each case, as suggested by the professional
article, is the ‘image’ of the accountant, though it is unclear how this concept
is to be either defined or measured. In each case there is likely to be a potentially large list of influential variables that may impose conditional relationships:
for the former, these include teachers, parents, schoolmates, perceptions of
other courses, job prospects, etc.; and for the latter, these include income,
career prospects, perceptions of competing professions, etc. Both of these
‘models’ in their simplest form ignore how the perceptions of the accounting
profession have been formed (e.g., did film and media have any influence
whatsoever, or are there other more important influences?), and whether
entry into the profession can be regarded as a single decision, irrespective of
the claims of the competing recruiters (e.g., individual members of the Big 5,
the public sector, commerce).
In terms of our example(s), we can only sample relatively small proportions of both school-leavers and potential recruits to the profession. In
neither case will we have access to population statistics. In addition, there
may be problems with interviewing school-leavers prior to university
entrance, especially if we are contemplating the conduct of psychometric
testing to examine personality issues and different cognitive styles. Such an
investigation is better conducted when the decision has been made, early in
the first year of university courses. This has the advantage of providing
access to mature entrants and of looking at an actual decision rather than a
choice intention, but it has the disadvantage of potential hindsight bias.
Data collection at this time may determine the motives for choice of entry
into competing professional courses (e.g., accounting, law, medicine). For
the decision point prior to entry into the accounting profession, some collaboration with the major recruiters (e.g., the Big 5, state government) is
advisable; they are all interested in attracting the ‘best’ students and make
clear statements on what they are looking for among successful applicants.
The researcher may gain access to student information, with appropriate
permissions and participation, to identify those selected for interview and
those subsequently offered positions. A comparison of recruiter decisions
with student profiles should facilitate an examination of the transparency of
recruitment policies. The design problems associated with the identification
and measurement of variables means that the basic schema can be modified
(see Figure 2.7).
Research Methods in Accounting
Identify and
Identify and
Identify and
Measurement issues
In terms of our example, we have serious problems in defining and
measuring the dependent and independent variables, and in controlling influential variables:
• ‘Image’ cannot be measured directly. The literature will help in identifying
those variables which contribute to the creation of the concept. This
process may allow the identification of variable(s) that can be measured
directly and used as a proxy for ‘image’. This is a far from satisfactory
arrangement but may be the best we can do if the investigation of these
relationships is to proceed.
• Decision choices will have been influenced by many variables over long
periods of time. We must restrict our focus to those variables cited in the
literature as being potentially influential, and try to gauge their relative
importance to decision-making in this context, possibly through pilot surveys or interviews after the event. The use of matching procedures (e.g.,
sampling candidates from the same schools, of the same age and gender,
and with equivalent academic profiles at entry) will help to provide some
control over the influence of extraneous variables.
We clearly have some way to go before this interesting research idea becomes a
feasible research project. Questions relating to theory and measurement remain
problematic, but a search of the literature specifies a number of theories with
precise implications for this scenario. Friedman and Lyne (2001) examine the
development of the ‘beancounter’ stereotype and its implications for the future
of the accounting profession. They identify a number of sociological theories
which could assist the establishment of our conceptual framework:
• Hamilton and Troiler (1986, p. 13) suggest that stereotyping is determined by cognitive, motivational and social learning processes. Social
Developing the Research Idea
learning theory suggests that stereotypes are mostly acquired through
socialisation, via variables such as parents, schools and mass media, but
that they may be modified by subsequent experience.
• Campbell (1965) suggests that realistic conflict theory may explain how
stereotypes result from inter-group competition for scarce physical
• Tajfel and Turner (1985) develop a social identity theory which modifies
conflict theory to take account of individual achievement motives so that
conflict can also occur over scarce social resources like prestige and status.
• Stroebe and Insko (1989, p. 14) recommend the simultaneous consideration of the three approaches: social learning theory, social identity theory
and conflict theory.
The implications of these theoretical approaches to our research idea are
• The perception of a negative accounting stereotype is unlikely to fade
among school students or accounting undergraduates while it continues to
be widely perpetuated (social learning theory).
• The negative associations are likely to impact on candidates’ choice of
career profession and the associated income and prestige (social identity
• The inter-group conflict between the professions, when each is trying to
attract the ‘best and brightest’ candidates, means that they may seek to
perpetuate the negative stereotypes of other professions to their own
resource advantage (realistic conflict theory). The inference here is that we
may extend the research to consider the advertising material used by the
competing professions (and by competitors within the accounting profession) and exposure of students to these. We may anticipate that recruitment at school level involves the denigration of the accounting profession
by other professions, and at university level the denigration of some members of the profession by fellow members from rival organisations.
What is also apparent is that the simple conceptual schema of Figure 2.6 has
outlived its usefulness in this context. It was helpful to begin with but must
now be modified to allow the further development of hypotheses and data
collection methods. Chapter 3 focuses on these issues in the development of
the important links between theory, literature, hypotheses and methods.
In Chapter 1 we acknowledged our debt to researchers in the natural and
social sciences. It is now helpful to turn to popular literature, which provides
highly readable explanations of complex situations, for an insight into the
emergence of research ideas, the development of research questions and the
airing of potential solutions for testing, to illustrate the research sequence in
practice. It is instructive to consider the different scope of original research in
solving practical problems by examining the particular aspects of three welldocumented stories from non-accounting environments. Thus we explore the
Research Methods in Accounting
development of new theory in a chemical environment in ‘The Structure of
DNA’, the development and testing of alternative strategies to address a
sporting issue in ‘The Bradman Problem’, and the solution of apparently
insuperable implementation issues to a problem where the ‘answer’ was well
known in ‘The Longitude Problem’.
The structure of DNA: the development of new theory
James D. Watson’s (1968) The Double Helix (subsequently filmed as Life
Story) provides a brilliant description of the exciting process of discovery in
scientific research, even if the approach adopted is rather unorthodox. The
development of theory and conceptual modelling, from systematic deductions
based on the empirical findings of others, is conducted in a competitive environment where the ultimate prize for winning the ‘race’ is the Nobel Prize.
(Watson, together with Francis Crick, both of Cambridge University, and
Maurice Wilkins, of King’s College, London, were awarded the 1962 Nobel
Prize for Physiology of Medicine for their pioneering work during 1951–52.)
Sir Lawrence Bragg reflects on these achievements by researchers in his
Cavendish Laboratory through a revealing preface to Watson’s book, with
implications for research ethics:
He knows that a colleague has been working for years on a problem and has
accumulated a mass of hard-won evidence, which has not yet been published because it is anticipated that success is just around the corner. He has
seen this evidence and has good reason to believe that a method of attack
which he can envisage, perhaps merely a new point of view, will lead straight
to the solution. An offer of collaboration at this stage might well be regarded
as trespass. Should he go ahead on his own? (Bragg, in Watson, 1968, p. vii)
That Watson and Crick, at the Cavendish, did proceed with a belated
relationship – though collaboration is too strong a word – with Wilkins and
Rosalind Franklin is a matter of history. The course of their investigations,
and the factors leading to their success in developing a new theoretical model,
have implications for all research. Essentially, they proceed to develop a
model which fits all of the evidence currently available to them, and await
confirmation or disconfirmation of their framework from the empirical findings of others:
• The work of Linus Pauling in the USA on a helix formation for polypeptide chains suggested that DNA (deoxyribose nucleic acid) too had a
helical structure. Pauling’s early attempts at modelling though, without
crystallographic evidence, had produced stereochemically impossible components. Early evidence from X-ray crystallographic diffraction presented
by Wilkins also seemed to suggest a helical structure, but there was no
Developing the Research Idea
evidence of whether a single, double or triple strand helical configuration
was most appropriate. Crick and Watson apparently proceeded on the
basis of an educated guess favouring the double helix because most things
biological come in twos!
• Ernst Chargaff had produced vital evidence on the ratios of constituent
bases, and particularly the equalities existing between adenine (A) = thymine
(T), and guanine (G) = cytosine (C). The A–T, G–C flat hydrogen-bonded
base pairs formed the core of the Crick and Watson structure, rather like
a spiral staircase in which the bases form the steps.
• Important advice from a structural chemist colleague (who just happened
to be sharing the same office with Crick and Watson) suggested that the
normal textbook formulation of the A–T, G–C bases was incorrect and
that they should work with an alternative ‘keto’ form. Without this
important questioning of textbook content, and accepted knowledge, their
structure would not have held together. The impact of this finding was
that a given chain could contain both purines and pyrimidines (with the
capacity to carry the genetic material for self-replication) and that the
backbones of the chains should run in opposite directions.
Thus Crick and Watson were able to construct a physical model comprising
two intertwining helically coiled chains of nucleotides, right-handed and
running in opposite directions, with complementary sequences of hydrogenbonded bases. The resulting structure was stereochemically possible, and
subsequent X-ray evidence from Franklin confirmed that the sugar-phosphate
backbone was indeed on the outside of the molecule.
The more general implications for researchers are constant vigilance and
a questioning attitude to the work of others and existing publications.
The Bradman problem: the development
of new strategies
Test match cricket in the 1930s was dominated by a single, outstanding
individual whose unique gifts of batsmanship threatened to change the way the
game was played. Donald Bradman scored so many runs, and scored them
so quickly, that he was likely to win a game almost single-handedly. Ashes
test series, once so closely contested, threatened to be one-sided affairs, with
Australia the perennial victor. In the 1930 England v. Australia test matches
Bradman totalled a record 974 runs at an average of 139.14 per innings, and
recorded separate scores of 334, 254 and 232. At the commencement of the
1932/33 series he averaged 112.29 in all test matches, and had already posted
six scores in excess of 200 in only 24 completed innings.
Bradman was a phenomenon, some would say a ‘freak’, and curtailing
his dominance became a pressing question for successive England captains.
Research Methods in Accounting
‘The Bradman Problem’ is capably detailed by Lawrence Le Quesne in his
book The Bodyline Controversy (1983), and presents an intriguing research
question, investigating a number of alternatives that might provide successful
1. Changing the playing conditions
Cricket was played on hard, fast wickets, largely true, though with the occasional unpredictable bounce. If bowlers did not take wickets in the first few
overs, when the ball was still shiny and swinging, then they might have a lot
of overs to bowl before they got a replacement ball, by which time 200 runs
had been scored. Wickets were uncovered and exposed to the elements once
a game had started and could become unplayable as a hot sun dried out a wet
pitch; these ‘sticky’ wickets provided a possible solution to the problem
because Bradman was nowhere near as prolific on bad wickets as many of his
contemporaries were. However, captains could hardly rely on this occurrence
to blunt Bradman’s genius on a regular basis.
2. Changing the rules of the game
Test matches were timeless in the 1930s and played to a finish. Declarations
were rare and slow play very common. Further, the leg-before-wicket (LBW)
rule made it difficult for the batsman to get out in that way – he had to be
caught in front of the stumps by a ball pitching in line, wicket-to-wicket. Life
was difficult for bowlers, and a number of other batsmen (notably Ponsford
for Australia and Hammond for England) regularly completed double and
triple centuries, although none with the regularity, reliability or speed of
Reduction of the specified playing time to five (or four) days would
mitigate slow play and dull the impact of less talented batsmen content to
occupy the crease, but would not affect Bradman – witness his 309 in a day
against England at Leeds in 1930. However, timeless tests were outlawed
before the end of the inter-war period. Similarly, the LBW law was changed
so that batsmen could be given out to a ball breaking back from outside the
off stump and striking the pads in front of the wicket.
3. Changing the bowling
Bowlers had often sought to restrict the batsman by bowling outside of the
leg stump (e.g., Hirst and Foster in the early 1900s), ostensibly to cut out offside shots but also to restrict the on-drive through bowling just short of a
length. An accurate bowler could therefore depress the scoring rate by keeping the batsman to deflected singles in the arc between the wicketkeeper and
square leg. As a consequence the whole game is slowed to a snail’s pace – a
Developing the Research Idea
strategy still used today to curb an aggressive batsman’s dominance. But
Bradman was too good to fall for such tricks and used his agility, quick
reflexes and nimble footwork to move to leg, outside the line of the ball, so
giving himself room to hit through the off-side field. Such a strategy would
have made less gifted batsmen vulnerable because it entailed the exposure of
all three stumps to the bowler, but Bradman could execute the manoeuvre
without undue risk.
4. Changing the field placings
An orthodox field will involve the placing of fieldsmen on both sides of the
wicket, with a majority coincident with the bowler’s planned line of attack.
This means plenty of gaps in the field and, for inaccurate bowlers, plenty of
scoring opportunities. Clearly, a preferred strategy would be to place all the
fielders in a tightly confined space and have the bowler deliver a line and
length that forces the batsman to play the ball there. This is a sound strategy
for an accurate bowler – witness Laker’s devastating use of a packed leg-trap
to a turning ball in 1956 – but for a batsman of Bradman’s ingenuity we now
have a vacant off-side to be penetrated by inventive, unorthodox shots.
5. Introducing a ‘bodyline’ attack
While a combination of numbers (3) and (4) above provides a partial solution in restricting scoring opportunities, only impatience will induce eventual
dismissal. For Bradman it may do neither. However, together they provide
the basis for a potentially successful solution: ‘leg theory’ will dictate the line
of delivery, but we need also to control the height of the delivery to induce
the batsman to give chances from playing the ball in the air. Herein lies the
Jardine–Larwood proposition, initiated by Douglas Jardine, England’s captain
in 1932 and executed brilliantly by Harold Larwood, the fastest and most
accurate bowler of his time.
Extreme fast bowling does not provide even Bradman with the time to
move outside the line consistently without risk. The introduction of a high
proportion of short-pitched balls rearing towards the throat or the rib-cage
of the batsman makes scoring without risk extremely difficult. Batsmen are
likely to fend…
Purchase answer to see full

error: Content is protected !!