Journal of Orthodontics, Vol.

31, 2004, 144–148

FEATURES How to write a thesis

SECTION
S. J. Cunningham
Eastman Dental Institute for Oral Health Care Sciences, London, UK

The prospect of writing a thesis can be intimidating. However, there are certain formats that the writer should follow in order to
make life much easier. This article covers a logical approach to presenting research findings. Also included are suggestions for a
last minute checklist.

Key words: Publishing, writing a thesis, writing a dissertation

So, you think the hard part is over now that your actual • Your full name and qualifications.
research is completed, but then you realize it is only just • The name of the institution to which the thesis is being
beginning! The thesis has to be written up! However, submitted.
there are many practical tips that can make writing-up as • The degree for which the thesis is submitted and the
stress-free as possible. year of submission.
The ‘thinking’ part of writing is the hardest part of any
thesis.1 Most graduate schools have their own guide to
writing a thesis or dissertation, and it is important that Abstract
this is obtained (and read!) prior to commencing the
The abstract is placed immediately following the title
research. Such guides offer information on a wide spec-
page and is generally short (up to a side of A4 usually).
trum of research-related issues, but more importantly,
The abstract should include a brief introduction and
will also give other vital information, such as word limits,
statement of the problem, as well as a summary of the
etc.
methodology, findings and conclusions. A structured
The recommended format for theses is similar for most
abstract may be used if preferred.
institutions, with some local minor variations:
• Title page
• Abstract Acknowledgements
• Acknowledgements It is important to acknowledge your supervisors, statisti-
• Table of contents cal advisers and others who have helped (e.g. colleagues
• Review of the literature — concluding with a summary who have offered support or provided data). In addition,
• Statement of the problem, aims of the study, hypoth- any funding bodies should be mentioned. It is also
esis customary to refer to those who have had to ‘put up’ with
• Materials and methods (including statistical analysis) you during the process of writing the thesis — usually
• Results family and friends!!
• Discussion
• Conclusions
• Appendices Table of contents
• References
Conventionally, there should be a table of contents,
followed by a list of tables and a list of figures. The table
Title page of contents includes all the major divisions of the thesis,
including subsections. The relationship between major
The title page should include the following information:
divisions and minor subdivisions should be shown by
• Title of the thesis — usually centred on the page and in an appropriate use of capitals and indentations.2 The
upper case letters. preliminary pages (Abstract, Acknowledgements and

Address for correspondence: S. J. Cunningham, Department of Orthodontics,

Eastman Dental Institute for Oral Health Care Sciences, University College
London, 256 Gray’s Inn Road, London WC1X 8LD, UK.
Email: [email protected]
© 2004 British Orthodontic Society DOI 10.1179/146531204225020445
 JO June 2004 Features Section How to write a thesis 145

Contents) are numbered using roman numerals (i, ii, ii Materials and method
. . . .) and the remainder of the thesis is numbered using
Arabic numerals (1, 2, 3 . . .). The title page is not Full details should be included in the materials and
numbered. It is sensible to start writing a table of con- methods so that the reader could repeat the study if they
tents as you write the thesis; however, the page numbers wished. For this reason, details should also be provided
can only be added when you have completed the final of any equipment and materials used. These can be
draft. At this stage, always cross-check the page numbers placed in a separate appendix if necessary (see later). It is
between the table of contents and the main body of the also useful to start writing this section as you undertake
text — and ask someone to double check them for you. your research, not only will everything be very fresh in
The list of tables or figures follow the same format, with your mind at this stage, but it will also mean you have less
to do at a later stage.
each table/figure number in Arabic numerals, the exact
This section allows readers to understand the objectives
title as it appears in the main text and also the page
of the study and to judge whether the methods used were
number.
appropriate. The methods may actually be modified
during the process of the study, in which case, the final
Review of the literature form of the method is discussed (unless advised otherwise
by your supervisor). The most serious error in this section
Beginning this section can be a daunting task, particu- is to omit important information required by the reader
larly if your research area is extensive. This should be one to see the logic of the study and to judge the adequacy of
of the first things you do, and if the majority of this work the method.1
is undertaken relatively early in the project, it makes later Everything to do with the validity of the method should
stages very much easier. The review of the literature be included in this section, including design, instrumenta-
should be as comprehensive as possible in the space avail- tion, analysis, etc. This section should also contain
able. Include only material that gives an appropriate evidence of the reliability of the method.
background to your study and avoid the temptation to There is a great deal of individual variation in the
include everything you have ever read on the subject! subheadings used in this section, often depending on the
Obviously, this review must be constantly updated to type of research undertaken. However, possible examples
ensure you do not miss important additions to the litera- include:
ture and a final search should always be undertaken • subjects;
immediately prior to submission, and also just before the • design;
viva in case an important work has been published in the • materials, apparatus and procedures;
time between submission and examination. • statistical analysis;
Make sure you acknowledge which paper something is • validity and reliability of method.
cited in if you did not actually read it yourself. An exam-
iner may find it hard to believe that you managed to get a Subjects
copy of a 1932 Russian orthodontic journal, let alone
managed to translate it yourself! There are two particularly important aspects to be
Great care is required in the way the references are considered here, both of which determine the generality
quoted in the text, especially with respect to punctuation, of the findings:1
etc. If several papers are quoted at the same point, they • The number of subjects at each stage (this establishes
should always be in chronological order. the statistical power).
• Freedom from bias in the way the subjects are
obtained/recruited. In clinical studies, we frequently
Statement of the problem, aims and
fail to achieve random selection for obvious reasons
hypothesis and it should be acknowledged if this is so. At this stage
The literature review is followed by a statement of the it is also important to consider that, particularly when
problem, then the aims of the study and the hypothesis. comparing two groups, the groups may not be bal-
Traditionally, null hypotheses were used because they anced. If this is the case, you will need to take this into
have their basis in inferential statistics. However, it is account and select an appropriate statistical analysis.
now much more acceptable to pose a hypothesis in its Full details of subjects should be included: number, age
positive form. When writing up the results and discus- group, gender, where they were recruited from, inclusion
sion, it is important to make sure that this hypothesis is and exclusion criteria, how many subjects were asked to
actually answered. take part and how many agreed, drop-out rate, etc. It is
 146 S. J. Cunningham Features Section JO June 2004

also important to discuss power and sample size at this to precede each sub-section of results with a summary
point and explain how you arrived at the estimated paragraph, but avoid duplication. Tables and figures
number of subjects/samples, etc. should be as simple as possible and the use of very
complicated graphics or obscure colour combinations
avoided — the examiner will not thank you! The table or
Design
figure should not repeat information covered in the main
This section describes exactly what type of study was text, it should augment it. Every table and figure should
undertaken. For example, a prospective randomized have a title that is a concise explanation of what is being
controlled trial comparing the effect of early treatment presented. If abbreviations are used, it is important that
with a functional appliance. they are explained fully.
Tables where p-values are quoted should give the actual
p-value, rather than p<0.05; p<0.01, p<0.001. With the
Materials, apparatus and procedures widespread use of computerized statistical packages, the
actual p-value can be found with relative ease.
This section includes all procedural elements that will
allow the reader to understand what happened. It may be
worth using diagrams or photographs if complex appara- Discussion
tus was used. Likewise, a flow chart can make a complex
procedure seem much more straightforward. Some Researchers often find the discussion the most difficult
aspects may be included in an Appendix, e.g. information part to write — it really is the ‘thinking part’ of the thesis.
sheets given to patients, details of the manufacturers of The examiner wants to know several things when they
apparatus or chemical reagents used. read this section:
• What did you find? Do you understand why you found
Validity and reliability this and can you suggest reasons why?
• Why is it important to you? Is it likely to affect current
The reliability and validity of any method should be clinical practice?
addressed at this stage. At this point it is also appropriate • How do your findings compare with previous studies,
to include any threats to the validity of the method (e.g. particularly with those which have been mentioned in
loss of subjects in a longitudinal study). the review of the literature?
Limitations of the method should be discussed fully • Did you encounter any problems with the methodol-
in the Discussion section. It is better to be honest ogy (for example, problems with recruitment/retention
and acknowledge any issues, rather than allowing the of subjects that meant the study was under-powered)?
examiner to point them out to you. There are problems Were the reliability and validity of the method
associated with all studies and examiners do realize that. acceptable?
It is also customary to follow the discussion with a small
Statistical analysis section on ‘Suggestions for further study’. This allows
you to make suggestions on continuation of the study or
It is a failing of many research studies that statistical amendments to the existing methodology.
advice was obtained too late to be of any real use and the
writing-up stage is certainly not the time to consult your
statistician for the first time! They should have formed Conclusions
part of the research team from the outset, and this is the Finally, the salient conclusions from your work should
time for them to provide support as you undertake be summarized and these frequently have more impact
the necessary analyses and to offer advice regarding if they are presented in bullet format. Try to avoid
interpretation of data. repetition and making this section too long; it is supposed
The statistical tests should be discussed fully, along to represent the most important findings — not every
with any criteria that were fulfilled in order to use these single finding!
tests (e.g. Normality of data).

Appendices
Results
In most universities, the appendices do not form part of
The results should be presented in a clear, concise manner the main word count, but avoid using them just to put
and avoid the temptation to present in more than one data in because you run out of space! They can, however,
format (e.g. in both tables and graphs). It is often helpful be used for data that is not required in the main body of
 JO June 2004 Features Section How to write a thesis 147

the text or for questionnaires that were used, consent Word limits
forms and information sheets given to patients, etc. If
you have a great deal of data that is not required in the Ensure you know the word limit before starting. Univer-
main text, but you wish the examiners to have access to it, sity regulations can vary markedly and finding out that
consider using a CD or disk, and putting it in the back of you have written 5000 words more than the limit, with
the thesis. only a week to go, will do absolutely nothing for your
stress levels!!

References
When commencing the research for the thesis, ensure
MPhil and PhD
you enter all references on a computer-based reference Larger theses are now frequently chapter based, in which
manager. This minimizes the errors, although does not case each chapter may have a review of the literature,
completely avoid them. If you do not have access to a material and methods, results and discussion. If this is the
computer-based system for any reason, keep index cards case, start the thesis with an overall introduction and
of all the references you have read so that you can go end with overall conclusions that bring the whole thesis
back and double check them. This avoids a great deal of
together.
work at a later stage. Always get someone else to double
check your references and cross-check that all those in the
text are in the references and vice versa. Be consistent
Dealing with your supervisor
with the style of the references, for example, either abbre-
viate all journal titles with correct abbreviations or write Try to make the supervisor/postgraduate relationship as
them all in full. easy as possible during your research and throughout the
writing-up phase. Remember that, although your thesis is
the most important thing to you, your supervisor may
Publications
well be involved with several others as well!
Include details of any papers, and the paper itself if Always keep to deadlines that are set for you and,
already published, that may have been accepted from the equally, attempt to set your supervisor a deadline on
study. This reflects well on your research, and shows that returning drafts. It is also sensible to keep a copy of the
it has already been peer reviewed and judged to be of a supervisor’s corrections, so that when they are done the
standard that is acceptable to a national/international supervisor is not tempted to ask for more! Most supervi-
journal. This is particularly important for MPhil or PhD sors will also appreciate being given the thesis in sections
theses. as they are written, as this breaks up the workload. They
will obviously want to see the entire thesis at a later stage,
but most of the corrections should have been done by
Consistency of style
that stage.
Consistency of style is important (the style of tables;
headings of sections, etc.). It may seem a small point, but
it does make the thesis look better if you have obviously Final checking
paid attention to detail.
The following should prove useful as a final check list:2

Additional points Readability

If you include diagrams or data from other publication,
Always ask someone else to read the thesis to check for
acknowledge them as such. Plagiarism is something that
readability, grammar, etc. This does not necessarily have
universities take very seriously.
With the increased use of desktop publishing, colour to be someone who fully understands it. Indeed, family
printing, etc., theses can be very professional. If colours members can often be willing volunteers!
are being used, choose sensible combinations and avoid
making it too garish! Page numbering
With the availability of spell check, there is absolutely
no excuse for spelling errors. However, do make sure Check the numbering of the pages to ensure it is consecu-
you choose the correct language (e.g. either US or UK tive. Likewise, also check the numbering of tables and
English) to check against! figures.
 148 S. J. Cunningham Features Section JO June 2004

Margins Reference list

Check that all of the margins are in accordance with the Has every work cited been included in the references?
regulations for your university (often 4 cm for the left Have the rules for alphabetical and chronological
margin to allow binding and 2.5 cm for the right margin). ordering of references been followed? Does each book
Also ensure that the right margin is justified throughout reference include the author(s) and/or editor(s), date of
as it looks much neater. publication, title, chapter number and title, publisher and
place of publication? Does each journal reference include
author(s), date of publication, title, name of journal,
Headings and subheadings
volume number and inclusive page numbers?
Check that all of the chapter headings and subheadings
in the main text match those in the table of contents, Binding of the thesis
and ensure that all of the headings and subheadings are
formatted consistently throughout the thesis. Have you checked the university requirements regarding
number of copies; type of binding for the submitted
version (e.g. do you need a hard bound or soft bound
Tables
copy at initial submission?); colour of cover and the
Has the data been checked for accuracy? Are sufficient lettering on the cover?
details given to interpret the data? Is a consistent format
used for all tables? Are units of measurement stated? Are
Summary
all abbreviations explained?
So, you have got to the final stage and the thesis is ready
Figures to go to the binders. Getting to that point is something
to be proud of — nobody will ever under-estimate the
Has the accuracy of the figure been checked? Is the figure amount of effort that has been expended in the process.
self-explanatory? Is the zero position shown on the axes However, following the steps in this paper should make
of graphs? Are the units of measurement clearly shown that effort as pain free as possible.
on the axes? Are all abbreviations explained?

References
Appendices
1. Van Wagenen K. Writing a Thesis: substance and style.
Is the appendix warranted? Is it referred to in the text? Englewood Cliffs: Prentice Hall, 1991.
Are sufficient details given to make the appendix useful? 2. Anderson J, Poole M. Assignment and Thesis Writing, 3rd
Does each appendix start a new page? edn. Brisbane: Wiley, 1998.