Accounting Research: A

Practical Guide
ACCRESM (Accounting Research Methods)
2nd Semester FY 2023-2024
Instructor: Daniel Nemeses C. Espina, CPA, MMA, CDPO, CFMP, CC
 1. INTRODUCTION
• 1.1 What is a dissertation? A dissertation is not an essay or an
encyclopaedia article. It must go beyond mere description and include
a level of critical evaluation.
• A dissertation requires the researcher to identify a problem or issue
for resolution in a manner that adds original knowledge to the subject
area.
• All statements should be supported by reference to previous research.
There should be no unsupported conjecture in a thesis.
Completing a dissertation involves a number of distinct stages or
phases. As shown below, this book deals with each of these stages.
Research students are required to:
• Identify a problem or issue for research / select a topic (chapter 2);
• Identify the objectives of the study;
• Review the literature previously written in the area (chapter 3);
• Develop hypotheses for testing the problem/issue being researched (chapter 5);
• Devise appropriate research designs and methodology to test researchable
hypotheses (chapters 4 and 5);
• Collect data (chapter 5);
• Apply appropriate methods of analysis to analyse data (chapter 5);
• Present results of the research (chapter 6);
• Summarise the research project and make final concluding comments (chapter
7).
Expressing this in another way, the research should be able to
clearly answer the following questions:

• What was the problem? (chapters 1 and 2);

• What have other researchers found/concluded about this problem? (chapter 3);
• How did you study the problem? (chapters 4 and 5);
• What did you find? (chapter 6);
• What do these findings mean? (chapters 6 and 7).

The research and resulting dissertation should contribute to the total knowledge
relating to the issue or problem researched.
1.2 Starting out
• Start writing immediately, building up the dissertation from the
beginning.

• Write creatively, tell an interesting story and at the same time

maintain a clear structure. Develop your ideas in a logical way and
make sure the reader can follow and easily observe such
development. Support arguments with references to previous
literature.
1.3 Planning the research
• Start by first preparing a timetable/schedule for yourself setting
deadlines, leaving some allowance for overrun. An outline of the
steps required for completion of a dissertation or research report is
shown in Table 1.1.
1.4 Liaison with supervisor
• Students should liaise closely with their supervisor during the period
of their research. Arrangements to meet should be made in advance,
especially if the research is being conducted out of term.
1.5 Keeping records
• This book is based on the philosophy (outlined in section 1.2) that
students should start writing (on computer i.e. word-processing)
immediately. This has two benefits: (i) to assist students in structuring
their work and (ii) as a means of keeping a continuous record of work
done during the period of the research.
1.6 Benefits of writing a dissertation
• Learning how to conduct a comprehensive analysis of an accounting issue;
• Analysing the theory and implications of an accounting issue more thoroughly
than before;
• Understanding and appreciating the complexity and importance of
accounting issues;
• Understanding and experiencing the research process;
• Appreciating the role of theory and concepts learnt in research and applied to
the real world;
• Examining and judging the quality of accounting and business research and
comment - the world is full of unsubstantiated assertions, dubious logic in
arguments, biased discussions etc.;
• Learning to critically analyse what others say and write;
• Learning to improve your work by avoiding such errors and refining
your ability to find weaknesses in other researchers’ arguments;
• Learning to access a variety of information sources in libraries,
including electronic databases.
• Developing report writing skills including marshalling thoughts in a
coherent way and putting forward arguments logically;
1.7 Regulations for completing a
dissertation
• Most universities and colleges have regulations and guidelines for
completing a dissertation. These regulations will specify matters such
as the deadline for completion of the dissertation. They may also
contain instructions on style of the dissertation including the format,
layout, referencing system to be followed.
 2. Identifying the Research Question
• Deciding on a problem or issue for research can be the most difficult
part of the research process.
• The process of finding a research question or problem for research
involves a number of stages, each a refinement of the previous stage.
As a first step, students should select a general area of interest.
• Students should then focus on a specific topic in the area selected.
• Many students find the third stage very difficult - to refine the topic
chosen to one or more researchable questions.
The following example illustrates this
process:

• Research area: Accounting education

• Research topic: Accounting ability of male and female accounting
students
• Research question: Are male accounting students better at accounting
than female students?
Table 2.1: Suggested areas and topics for
research in accounting
SPSPS Research Agenda
• The research program of Saint Paul School of Professional Studies
seeks to identify aspects of human endeavor that can be augmented
by case studies, surveys, and in-depth interviews for the development
of new tools for research and consultancy. This includes:
1. Quality and Excellence
2. Relevance and Responsiveness
3. Access and Equity
4. Effectiveness and Efficiency
5. Information and Communication Technology
6. Environment
Research Areas and Themes
2.1 Factors influencing choice of topic
• The most important consideration in selecting a topic is whether it is
of interest to the student.
• Researchers are advised to choose a topic that plays to their
strengths.
• students with good language skills might consider doing research in
the general area of international accounting, focussing on the country
of their language expertise.
• Connections in the business world might suggest opportunities for
access to data sources not available to the average researcher.
• Another good starting point might be to replicate a piece of research
done elsewhere (e.g. UK, US). What starts out as a replication study
will end up as a unique and original piece of research.
• Practical considerations in choosing a research topic include:
1. Time and effort of student;
2. Library facilities;
3. Computing facilities;
4. Data sources;
5. Access to data (e.g. for case studies etc.).
2.2 Characteristics of a good research topic
• The topic selected should be narrow in scope.
• A good topic should be capable of expression in the form of one or more questions
• It is essential for a good dissertation that the objectives of the study and the
questions that need to be answered must be explicitly and concisely stated and
systematically presented.
• The topic should be interesting to you and to other researchers - it shouldn’t beg
the question “so what?”.
• The topic selected and the research question for testing should have a symmetrical
outcome i.e. a reasonable chance of being disproved as proved
• The research question should be amenable to research. The elements in the
research question (variables) should be measurable.
Table 2.2 summarises the characteristics of
a good research topic.
2.3 Writing a research proposal
• Many universities require students to prepare an initial research proposal in
advance of undertaking the main body of research, which may have to be
approved.
• This should be short and at least should cover the following:
1. Clear statement of the research questions and research objectives;
2. Brief discussion of the reasons and justification for selecting the topic;
3. Brief introduction to the literature on the topic; • Research methodology
in outline;
4. Initial references;
5. Outline of chapter contents.
 3. LITERATURE REVIEW
• The ability to carry out a literature review is an important skill for a
researcher. Researchers need to find out what work has already been
done in the area and how previous research compares with the
project on hand. In addition, the literature review will provide a
framework for the current research project.
3.1 Nature and purpose of a literature
review
• The purpose of a literature review is to summarise clearly and critically previous
research, and to put your research in the context of previous work in the field.
• Four functions of a literature review:
1. To give reasons why the topic is of sufficient importance for it to be
researched;
2. To provide a brief up-to-date account and discussion of literature on relevant
issues;
3. To provide a conceptual and theoretical context as background;
4. To discuss relevant research the same or similar topics.
The literature review is not a compilation of facts but is part of a coherent
argument leading to a description of the study.
• It is easier to organise the literature review into a coherent framework
if the research is well structured from the outset.
• Most of the literature review will take place at the earlier stages of
the research. The design of your research project depends on how
well you review previous literature. Previous literature will suggest
research questions and hypotheses for testing and methods of testing
them.
3.2 What constitutes the literature?
• A wide variety of publications are suitable for inclusion in the
literature review. These range from academic journals to professional
journals, professional regulations, articles in the business press,
textbooks, case studies, conference papers etc.
• Many students spend considerable time reading professional
accounting journals such as Accountancy Ireland, Accountancy,
Accountancy Age, The Accountants’ Journal etc. These are what
academics refer to as ‘non-refereed’ journals (i.e. the articles are not
subject to review by another academic prior to being accepted for
publication).
3.2 What constitutes the literature?
• Similarly most textbooks are not refereed and should only be used
selectively.
• Your literature review should be up to date. It is important to include
the most recent academic papers / articles on your topic.(Ideally, last
5 years)
• You should conduct a thorough examination of the literature.
3.3 Literature and information sources
• Information sources are described in research as ‘primary’ or
‘secondary’. Primary sources of information involve the researcher
generating the data. Examples of collection of data by researchers
include data obtained by survey, interview or experiment.
• Secondary sources are where the information has been collected,
organised and structured by someone else.
• Students will rely on the following main sources of information for
secondary data: 1.) Databases; 2.) Libraries; 3.) World Wide Web
3.3.1 Libraries
• Students should familiarise themselves with the facilities available in
their university/college library.
• Alternatively, the librarian may, on an individual basis, demonstrate
how best to use the library facilities.
• Most students know that libraries hold books and professional and
academic journals. However, in addition, students should find out
what directories, trade directories, indices, abstracts, microfiche and
microfilm material are available in their subject area.
3.3.2 Databases
• Students should find out what electronic databases are available in
their library.

In addition, electronic databases may be available. Examples include:

• Investext: Brokers’ reports for European listed companies;
• F & S Index Plus: Trade and current marketing information;
• FT Information: Equity research, company research and company
analysis;
• FT Graphite: Financial information.
3.3.3 The World Wide Web
• The World Wide Web is becoming an increasingly popular method of
find information useful for accounting research.
• For example, CAROL (Company Annual Reports On Line) is a site for
accessing annual reports and accounts and other financial information.
It has links to all the Financial Times Stock Exchange (FTSE) company
annual reports on the Internet. The US stock quotes and financial
information home page 20 Yahoo Finance is a vast source of
information.
• Yahoo Finance is a vast source of information. It has the ‘latest market
news’, features on ‘stock chat’ and seven reference sites on various
categories of financial information.
3.4 Academic accounting journals
3.5 How to read an academic article
• Table 3.4 lists suggested headings to be used for summarising an
academic article. Applying these headings will help to make academic
articles more digestible and understandable. Although it is time
consuming to summarise articles, good summaries will pay dividends
in the end.
3.6 Use of citations
• A literature review is not a grouping of unacknowledged quotations.
Previous research findings should be put in your own words and be
used to support the issues specific to your thesis.
• Statements should be supported by references. If possible, do not
make unsupported statements, and try to avoid including personal
opinion, as this would not be considered appropriate to rigorous
research methods.
• There is a format or convention for citing material from other
publications in a dissertation.
• Citations should give the surname (only) and year of publication.
• Any articles or papers referred to in your dissertation must be articles
or papers you have read. If you could not get the article or paper but
know of its findings from another article or paper this should be made
clear in the way you write your literature review. Here is an example
from my own dissertation. I was unable to get a copy of Spero (1979)
in the time available as it was an unpublished doctoral dissertation.
• ‘Marston and Shrives (1991) quote Spero (1979) as reporting that
attaching weightings to disclosure scores is irrelevant, as firms that
are better at disclosing ‘important items’ are also better at disclosing
‘less important items’.’ (Brennan, 1995)
3.7 References/bibliography
• The dissertation should finish with a list of references relating to
articles or papers etc. quoted therein. A bibliography (as opposed to a
list of references) cites additional reading on the topic.
3.8 Referencing systems
• As material is being collected and read, a careful note of the exact
reference should be made.
• References should follow the style guidelines of your chosen journal.
Follow the format exactly as laid out in the style guidelines. Be careful
to note exact punctuation etc. – capitals, punctuation and the sequence
of information are important. To illustrate, the referencing system in
The Accounting Review is summarised in Table 3.7. Appendix 1
reproduces the style guidelines of the Accounting Review which include
examples of the application of the guidelines to a selection of
references.
• For easy referencing, you may visit www.citationmachine.net
 4. Methods of Accounting Research
• Accounting research is becoming increasingly rigorous and scientific.
Many research studies start with a theory about a phenomenon or
behaviour.
• For example, some researchers (such as economists) will use
mathematical and analytical techniques to solve a particular problem,
whereas others will collect data and subject it to empirical testing.
• Research methodology refers to the way problems are approached
with a view to seeking answers. The methodology describes how an
individual piece of research is conducted. The terms ‘research
methodology’ and ‘research design’ are often used interchangeably.
• Accounting research can be categorised as either qualitative or quantitative.
• Qualitative research is often suitable for datasets that are too small to allow
generalisations of results from the sample to the population as a whole.
Such research is likely to lead to an in-depth understanding of a process
rather than making predictions about the causal relationships between sets
of variables.
• Quantitative research involves statistical analysis of large datasets.
Generally, the aim is to assess the extent to which changes in one variable
(the dependent variable) are caused by changes in the other variables
(independent variables). Information on the dependent and independent
variables must be collected by the researcher either directly (primary
sources) or from sources generated by others (secondary data). Such
quantitative research 30 may allow for generalisations from the dataset to
the population as a whole, depending on how representative the dataset.
4.1 Descriptive research
• Descriptive research includes detailed analysis of accounting problems by
focussing on accounting theory, published research, accounting regulations
(current and proposed) and current practice. Care must be taken with this
type of research that the subject matter is dealt with rigorously and
logically and that the resulting report is a dissertation and not an essay.
• Descriptive research may include sample statistics describing reported
accounting items, for example: • Financial reporting practices, trends,
levels; • Management accounting practices; • Sample average of
accounting amounts such as sales, total assets etc.
• For more advanced academic research, description is usually insufficient –
explanation is also needed which requires additional research methods.
4.2. Behavioural research
• Behavioural research examines the way people behave - for example,
people’s decision processes and the impact of organisations and
systems on behaviour. Some of this research derives from psychology.
This kind of research is suitable in many areas of accounting. For
example, it is common in studies of behavioural aspects of auditing
such as building expert systems and group versus individual decisions.
Behavioural research is suitable for experimental work in which
subjects are used to simulate real life decision-making (see 4.6
below).
4.3 Questionnaires
• Questionnaires are used extensively in business research and by
research students. They are a convenient way of obtaining views of
large numbers of individuals quickly and economically.
• All respondents are asked the same questions. Question wording has
to be chosen carefully. A good questionnaire will have been pilot
tested in advance on a small number of people to ensure the
questions are well drafted, unambiguous, not prone to multiple
interpretations and obtain the information required by the researcher.
• Questionnaires can be administered in different ways. Many researchers
distribute questionnaires by post. Response rates from mailed questionnaires
are often poor. An alternative approach is to have an interviewer administer the
questionnaire. This generally provides a higher response rate but the
interviewer may inadvertently introduce bias into the responses.
• Many questionnaires are poorly designed:
1. Questionnaires may be biased in advance;
2. Questionnaires may not ask the right questions;
3.Questions may not be clear and are consequently prone to multiple
interpretations;
4. The response rate to mailed questionnaires is generally extremely low (often
less than 20%) which could lead to biased results;
5. Nowadays people in business are inundated with these research instruments
and are irritated with the number they are asked to complete.
• Completion of a questionnaire takes a long time, involving preparing the
questionnaire, identifying the sample to be surveyed, copying the
questionnaire, preparing the mail shot, following up non-respondents etc.
• Questionnaires are an effective means of collecting facts from subjects but
are less successful at collecting opinions/value judgements such as views on
the relative importance of a particular item of accounting information. It
should be noted that causal relationships are rarely proved conclusively by
questionnaires.
• Questionnaires have some advantages over interview research. Respondents
can answer questions freely unhampered by considerations of the views of
the interviewer. Conversely questionnaires have limitations. Only simple
straightforward questions can be asked which significantly reduces the depth
of the analysis that can be undertaken.
• The validity of the results can be seriously impaired by low response rates.
The researcher cannot follow up responses to questions.
4.4 Interview research
• Interviews aim to find out the views of individuals such as investors,
analysts and auditors. They allow more in-depth discussion of
individuals’ views than questionnaires but are limited by the number
of individuals that can be interviewed. Also interviewees may react to
the presence of the interviewer and change their responses to the
questions. three basic approaches to collecting data through
interview. Unstructured interviews follow an informal conversational
approach. They have the advantage of flexibility but it can be difficult
to aggregate and report the results as each interview takes 33 its own
unique path. In accounting research interviews will generally be semi-
structured or structured.
• As the interview outline or questionnaire addresses the research
questions being examined in the thesis, it should therefore be
structured in a way that is complementary to the structure of the
dissertation.
• Interviews should be carefully recorded either in writing or by tape
and should be typed up immediately afterwards.
• Results of interviews should be summarised, possibly in a table. The
main, most interesting, points coming from interviews should be
emphasised by quotes (put in inverted commas and/or italics) from
interviewees.
4.5 Survey research
• Surveys in this context are distinguished from questionnaire-based
research (sometimes called survey research) whose purpose is to find
out views of individuals.
• Surveys in accounting frequently refers to surveys of accounting
practices such as: 1. Financial accounting treatment as evidenced in,
say, published accounts; 2. Management accounting techniques; 3.
Practices in internal audit departments; 4. Operation of audit
committees.
• Such survey research can be very suitable for a minor dissertation.
However, students should be careful to add new knowledge and not
just replicate a survey done before which might result in an
uninteresting and banal dissertation.
4.6 Laboratory experiments
• In this kind of research subjects are used to simulate decision-makers.
Subjects are presented with information on which they are asked to
make decisions similar to real life decisions.
• Students are often used in this type of research as cheap accessible
surrogates for real decision-makers. Results from such research are
only valid if the students behave in the same way as the decision-
makers they are simulating.
4.7 Field work and case studies
• Fieldwork is usually taken to mean studies of social practices in the field of activity in
which they take place. A case study implies a single unit of analysis. Both these
techniques enable study of accounting in its practical setting.
• A case study involves researching a problem, focussing on a single case such as an
individual company. A case study allows in-depth study of issues in a small number of
cases or firms. This type of research is not as rigorous as other types but it can provide
insights into the real world not possible using other research techniques. It is vital
that case study data is collected systematically and that the study is methodically
planned.
• Case study research involves selecting cases and gaining access to them. The firms
selected must be described and put into their organisational and social context.
Record keeping, interpretation and reporting of case studies in an objective manner
are essential to the quality of the research findings.
4.8 Accounting models
• Some studies are highly theoretical, deriving from economics, and try
to develop mathematical models to explain events. Most of these
mathematical models use very simplistic assumptions and are difficult
to apply to real world situations. There is a trade off between
simplicity of these models and their validity.
4.9 Empirical research
• Social scientific research (such as accounting) is more difficult than scientific
research (chemistry, physics) in that it is difficult to set up controlled experiments.
For example, it is not easy to vary treatment across subjects. Frequently accounting
research examines what occurs after the event in an uncontrolled environment.
Many variables are difficult to measure.
• Empirical studies use a more scientific approach to accounting research. Empirical
research adapts the problem being researched into a proposition testable by
available research methods. Testable research hypotheses are developed and
specified in advance, usually based on, and derived from, firm theoretical
foundations.
• A hypothesis specifies a relationship, or no relationship (referred to as null
hypotheses), between two or more variables - the dependent variable and
independent variables.
• Hypotheses are better stated in positive than in null form. The research
design/methodology attempts to confirm or reject these hypotheses.
• It is difficult, if not impossible, to prove a hypothesis – to do so you
would need to observe the complete population of events.
• However, research can provide evidence rejecting a hypothesis or
confirming it for the sample examined.
• Empirical research methodology requires dependent, independent and
control variables to be measured quantitatively.
• Statistical techniques must be applied to determine the significance of
the results.
• This type of research needs a large amount of data and clear causal
relationships to enable statistically valid conclusions to be drawn.
• The research design has to be sufficiently well specified to enable
another researcher to replicate the study. Thus the population,
sample and measurement of variables must be described in exact
detail.
• H: The examination performance of male accounting students is
better than that of female accounting students. The table starts by
showing the stages (being continuously refined) from choosing an
area for research to articulating a hypothesis for testing.
• Table 4.1 illustrates that many choices can be made in measuring the
population, sample and variables.
• The dependent variable, examination performance, can be measured by
reference to second level, third level or professional accounting
examinations. In measuring this variable, a decision has to be made
about the time period to be covered.
• For example examination performance could be measured as (i)
Pass/Fail, (ii) Honours/Pass/Fail, (iii) Result in ranges e.g. 0-10%, 11-20%
etc. or (iv) the examination marks.
• Control variables included in the model for testing depend on the
dependent variable. For example, if the dependent variable is
examination performance at third level, whether the student studied
accounting at school would need to be controlled for as this would
influence performance in accounting at third level.
• The population, sample, variables and statistical technique must be
sufficiently well described to enable the study to be replicated.
• To summarise, the empirical approach to accounting research involves:
1. Specifying hypotheses for testing;
2. Designing models to test hypotheses;
3. Operationalising and measuring variables for examination in the model;
4. Controlling for other factors;
5. Applying statistical tests to the model examined.
An example of a rich vein of empirical research in accounting is capital
markets research which tests for a relation between share prices and
accounting information such as earnings, its components (revenues,
depreciation, asset write-offs, pension liabilities etc.). This research is
based on the Efficient Markets Hypothesis and on the Capital Assets Pricing
Model from Finance Theory. The relationships between firms’ share
returns and their earnings can be statistically tested.
 5. Research questions and methodology
• Do Irish accounting standards comply fully with international
accounting standards?
• Are Irish accounting regulations as comprehensive as US accounting
regulations?
• What are the differences in accounting treatment of goodwill in
Ireland compared with the rest of the world?
A good research project will contain a clear statement of the research
problem or issue for research.
5.1.1 Converting research questions into
hypotheses
• Not all research projects will contain hypotheses for testing.
• In smaller scale projects, clearly articulated problems for research may be
more appropriate than explicit hypotheses.
• A hypothesis is a tentative proposition which is subject to verification through
subsequent investigation.
• Hypotheses make statements (conjectures or ‘hunches’) about the
relationship between variables.
• If you are using hypotheses (a good thesis will specify clearly the hypotheses
or propositions being tested by the research), your methodology should be
able to prove/disprove the hypotheses.
• There should be good linkage between research questions and hypotheses.
• Hypotheses are better put positively than in the null form.
• Wallace (1991) discusses a number of desirable attributes of
hypotheses. They should be:
1. Testable;
2. In harmony with existing knowledge;
3. Parsimonious;
4. Relevant to the research problem or question of interest;
5. Susceptible to quantification;
6. Yield a large number of consequences;
7. Logical and simple;
8. General in scope.
5.2 Data for research
• Data for research may be collected or generated by researchers
(primary data) or may be generated by others (secondary data
sources).
5.3 Research methodology
• Different topics and issues being researched require different research
methodologies (some of which were discussed in chapter 4 on
different methods of accounting research). A practical approach to
research design and research methodology, particularly for students
doing research for the first time, is to copy what other researchers
have done. This is particularly the case for measurement of variables
and for choosing appropriate statistical techniques. It is important
therefore when reading the literature to carefully note the research
methodology used by other researchers, while asking yourself the
question “Would this be appropriate/do-able in my thesis?”
• advisable to pilot test your design in advance on a small sample to
ensure the design is satisfactory.
5.4 Statistical analysis
• Statistical analysis is only feasible when there are sufficient data points
to enable statistical conclusions to be validly made.
• Categorical (in categories)
• Nominal: Data classified into categories e.g. Industry analysed into
three categories Manufacturing, Non-manufacturing and Financial; •
Ordinal: Data ordered or ranked e.g. Small, Medium, Large; Continuous
(measured as a continuous number) • Interval: Data for which physical
unit of measurement in numerical terms is possible e.g. length in feet,
heat in degrees centigrade/Fahrenheit etc. • Ratio: Continuous numerical
measure with absolute or non-arbitrary zero point e.g. profits measured
in £s.
• There are broadly two types of statistical test. Non-parametric tests
make no assumptions about the behaviour of the data; parametric
tests assume the data are distributed normally.
Table 5.1 shows some of the main
statistical tests used in research.
• Univariate statistical tests examine a single variable on its own e.g.
company size. The univariate statistical test chosen depends on how
the variable, company size (say), is measured – whether as a
continuous numerical variable (e.g. net asset value in £s) or as a
categorical variable (Small, Medium, Large).
• Bivariate tests examine the relationship between two variables.
Multivariate tests examine the relationship between a dependent
variable and more than one independent variable. The test chosen
depends on the behaviour of the data (whether normally distributed
or not) and on the measurement level (categorical, ordinal, interval,
ratio).
• Computer software is now available to carry out most statistical tests
used in research. The most common packages are SPSS (Statistical
Package for the Social Sciences), SAS (Statistical Analysis Software)
and Minitab. There are many books and manuals available for these
packages.
 6. Results
• Students often spend considerable time and effort in collecting raw data.
Unless this is well recorded, analysed and interpreted the reader may not
fully appreciate the research done. In presenting the results of the research,
the researcher needs to look for similarities, differences, groupings,
patterns and items of particular significance. The information needs to be
organised, understood and presented in such as way as to allow conclusions
to be drawn.
• It can considerably assist the design of your research if you conduct a small
pilot study and write up the results in advance of the larger data collection
exercise. There is nothing worse than finding at the analysis of results stage
that there is a vital piece of information you omitted to collect.
• It is important to distinguish between results and conclusions. Results
are the raw information from the research project. Conclusions are
what can be deduced, inferred and argued from the results.
Discussion may include some speculation about the reasons and
causes for the results. The quality of the discussion and argument of
the results, and inferences therefrom, will make the difference
between a good and bad dissertation.
6.1 Analysis of data
• The method of analysis depends to a large extent on the nature of the
research project. Regardless of the method of analysis, a certain
amount of trial and error is required in trying out different
approaches to analysing the data. Experimentation is required until
you find the approach that best presents the data to the reader.
6.2 Presentation of results
• Excessive use of un-interrupted narrative is boring for the reader and an
inefficient way of presenting results.
• 6.2.1 Tables
You should make good use of well laid out tables. Tables may be suitable for
verbal as well as numerical data. For example, a two-column table might be
an efficient way of comparing and contrasting two phenomena.
Tables should be introduced to the reader by a concise commentary in the
text. The table number should be referred to in the text. The commentary
should state what the table is about and briefly refer to the significant
findings. Numerical tables should be accurate and should not contain totting
errors.
• There may be circumstance where two tables presenting results might
be easier for the reader to follow than putting a large amount of data
into a single table. Alternatively, opportunities for efficiencies in
presentation should also be sought, whereby data in a number of
tables could be better summarised in a single table.
6.2.2 Other ways of presenting results
• Other methods for presenting verbal and numerical data include
family trees, algorithms, flow charts, graphs and histograms.
6.3 Interpretation of results
• Themes and patterns across results should be identified. Likewise, any
inconsistencies should be pointed out. Reasons for these themes,
patterns and inconsistencies should be put forward. Comparison of
the results with prior research findings should be made. The sources
of any conflicts with expectations derived from the literature review,
or with previous research findings, should be examined. The reasons
for such differences need to be fully explored. Commentary should be
succinct.
• The results should make sense and should be intuitively appealing.
Are they what you expected? If not, why? Could your research
methodology be the problem? Be careful not to read more into the
results than they are telling you.
• Conclusions should be drawn by reference to the specific aims of the
project. Conclusions need to be expressed clearly and unambiguously.
Some recommendations should be derived from the research.
 7. Writing the research report /
dissertation
• The report / dissertation Research reports and dissertations generally
have (subject to topic) a similar structure which is quite distinctive
from other types of business reports.
• General structure
• Research reports and dissertations contain introductory material
followed by a number of chapters and ending with references and
appendices. Page numbers in roman numerals (e.g. i, ii, iii..) often
distinguishes introductory material from the main text.
• The body of the text should all be in chapters, and the material should
be organised in the appropriate chapter. For example, the literature
review should be in the literature review chapter, the hypotheses in the
hypotheses chapter etc. Material included should be directly relevant to
the dissertation. Incidental material padding the dissertation will only
confuse, if not irritate, the reader.
• Preface material. The dissertation should commence with a title page.
This should contain the title which should be clear and concise,
including only essential words. All words in the title should be chosen
with care, paying attention to syntax.
• The title page should also contain, in the correct format, the
submission details which include: 1. Full title of the dissertation; 2.
Name of the author, followed, if desired, by any qualifications and
distinctions; 3. Qualification for which the dissertation is submitted; 4.
Whether the dissertation is being submitted in partial fulfilment for
the requirements of the degree; 5. Name of the institution to which
the dissertation is submitted; 6. Month and year of submission.
7.1.3 The first chapter
• The most difficult, and some would say the most important, chapters are
the first and last. First impressions are very important and create an initial
reaction to your work which may be difficult to change thereafter. The last
chapter is where you tell the final story - the findings and implications of
your work.
• These chapters should be written bearing in mind that some readers may
only look at the first and last chapters. These chapters will therefore have to
contain material that may be repeated elsewhere. Some repetition is
necessary in a research report or dissertation. For most readers coming to
the research for the first time, constant reminder of the project, the themes
and objectives underlying the research, and the research methodology, is
necessary.
7.1.4 Body of the report / dissertation
• For the written report to hang together as a coherent story, there should be good
linkages between each chapter. Every chapter should start with introductory
paragraphs and should end with a summary and conclusions section.
• The introductory paragraphs should introduce the reader to the chapter and briefly
summarise the chapter contents. The chapter introduction may refer to material in
previous chapters to emphasise the linkages between what has come before and the
material in the current chapter.
• The summary and conclusions sections should briefly summarise the chapter,
following its structure and sequence. It should end with an introduction to the next
chapter.
• Text in chapters should be sub-divided as much as possible by headings and
subheadings. There should be a logical and obvious structure to the material which
should be clearly evident by looking at the table of contents.
7.1.5 The final chapter
• The final chapter is essentially a summary and conclusion of all
chapters preceding it. It should briefly summarise the research project
(i.e. issues addressed, methodology) and the research findings.
7.1.6 References
• The chapters should be followed by the list of references. This was
dealt with in section 3.8 earlier. To summarise - References should be
in alphabetical order and should consistently follow a referencing
style (APA style). The list should only contain references cited in the
research report or dissertation.
7.1.7 Appendices
• Appendices generally contain detailed information which is not
essential in the main chapters of the report but which need to be
reproduced. Examples of material suitable for appendices include a
copy of any questionnaire or interview outline used in the research, a
full listing of the sample used (such as names of companies in the
sample), background information on companies used in case study,
list of abbreviations used in the report etc.
7.2 Methods of presentation
• 7.2.1 Writing style
1. As well as carrying out a piece of research competently you need to communicate the
work to the reader. Remember you are telling a story - the more eloquently, logically and
persuasively you tell that story the better you will persuade the reader that your research is
good.
2. No matter how well you carry out the research, if you cannot tell the story well you
cannot expect the reader to believe the research is well done. Logical argument and
structure are very important.
3. A well-constructed table of contents is vital to this. Make sure to present information in a
logical sequence. Start your story at the beginning. Move from the general to the particular.
4. Try and write as you would speak – avoid complex words and language that you would
not use in everyday life. Language and tone should be professional/academic - overly
personal remarks and jokes usually grate on the reader.
The following specific points about style
are worth noting:
• The style of writing should be professional and business like.
• Generally use passive voice (“This dissertation was written..”). It is probably better to
use the passive voice as some academics do not like the active voice in academic
research;
• Numbers up to ten should be spelled out (e.g. eight), numbers greater than ten
should be in figures (e.g. 18);
• Be careful about using capitals
• Keep sentences short;
• Use plain language;
• Do not include words where their omission would make no difference to the
meaning of the sentence e.g. ‘The disclosure of profit forecasts provides useful
information to the shareholders’ - omitting ‘the’ twice improves readability;
• Avoid repeating the same word or phrase in a sentence; Similarly,
avoid starting successive sentences with the same word or phrase;
• Footnotes should be kept to a minimum or not used at all.
7.2.2 Styles of presentation
• Text should be properly paragraphed.
• You should decide whether to use capitals, bold, underlining or italics in
headings - once you decide, be consistent throughout the dissertation.
• Vary your presentation (not font styles – see later comment) to make the
dissertation more interesting and readable. Use of bullets, tables, and
diagrams etc. helps to summarise issues and focus the readers’
attention, especially where there is a lot of text to be read.
• Use the most up-to-date technology available i.e. word processor (e.g.
Microsoft Word) and a good quality printer. Limited and appropriate use
of colour can improve appearance.
7.2.3 Length and format
• Maximum length, in terms of number of words, is usually specified in
the dissertation guidelines. There is no necessary relationship
between quality and length of dissertation. However, there is a trend
towards concise, brief dissertations. Briefer dissertations are more
likely to reflect sharp and clear thinking (Allen, 1973).
• Choice of font and spacing (single, space and a half, double spacing)
will influence the number of pages e.g. 10,000 words, space and a
half, in New Times Roman 12 point font will be approximately 45-50
pages.
• The dissertation should include page numbers. The introductory
material should be numbered using lower case roman numerals (i, ii,
iii etc). The first page of chapter 1 should be numbered page 1. Page
numbers should be centred at the bottom of the page and look best in
10 point.
• 7.2.4 Layout and presentation of main text Modern computers have a
wide range of font styles and sizes to chose from. You should choose a
font that is plain and not too fussy (for example, New Times Roman).
Font size for normal text should be 12 point but larger font sizes might
be used on (say) the title page, whereas smaller font sizes can look
well for text in tables. Do not use too many font sizes/types as this can
irritate. Text should be well spaced. Space and a half is usually
adequate, but some researchers use double spacing. Text should also
be centre justified.
• Use plenty of headings and sub-headings to divide up text so that the
reader is not presented with un-interrupted pages of narrative.
Headings might be presented in bold or 58 in italics for emphasis (but
don’t over do this – for example, headings in capitals, bold and
underline can be excessive on the eye). Similarly, use bullets where
appropriate to list points especially where an overview or summary is
provided.
• Headings and sub-headings should be concise.
7.2.5 Presentation of tables and figures
• Tables in dissertations can be used for two purposes: (i) to summarise the
literature / regulatory framework and (ii) to summarise results of research.
Some dissertations include tables and figures at the end of the dissertation
(this is also the format required for articles submitted to journals). However
dissertations are more readable if tables are integrated with the main text.
• Commentary should come before (not after) tables and figures.
• Tables and figures should start with a number and title.
• Headings and captions should briefly explain the numbers and other data in
tables and figures.
• Inserting a frame or textbox around a table prevents the table from
appearing on more than one page.
7.3 Producing the final document
• Leave plenty of time at the end for the production process.
Dissertations need to be carefully proof-read - at least twice, by
calling it over with another person - preferably someone experienced
and with a good eye for detail.
• Only then can you put in page numbers (and be certain they won’t
change!) and print the table of contents.
HAPPY WRITING!