Research Dissertation Guidance
Presentation of text
Sections of text should be clearly set out, with tables and diagrams numbered consistently, and a thorough check made on spelling and layout. The supervisor’s job is to advise on the specialised aspects of analysis, not matters which should have been dealt with earlier. Do not spend too much time on elaborate computer-generated diagrams. Neat diagrams supplemented by captions and labels in word-processed text are sufficient. It is more important to concentrate on the commentary and analysis of results.
Dissertations should be typed/word-processed single spaced, with generous margins on A4-size paper. Leave a margin of at least 1 inch (2.5 cm) on the left hand side for binding. The main bound copy should be reproduced in single-sided form, but the second may be double-sided.
Text should be single-spaced, 12 point, in a commonly-used typeface (this one is in Times New Roman). Text should be fully justified (i.e. flush to right and left margins, like this document).
The first four pages should comprise:
– A title page (including your name, course and year)
– Abstract and acknowledgements (total word count should also be shown on this
page)
– A list of contents (with tables and diagrams listed separately, as well as chapter
headings)
– A glossary of abbreviations and technical terms used.
1. Internal structure of a dissertation
The precise structure of chapters will vary according to the subject matter, but as a general guide the following may be appropriate:
1. Introduction, setting out the issues to be examined.
2. Literature review – relevant published work, and other material on the subject
3. Methodology to be used
4. Description of specific case study to be covered (e.g. particular firms, locations,
user groups, etc.).
5. Surveys and data collected (where applicable)
6. Analysis of evidence and discussion
7. Summary and conclusions
Overall length should be around 15,000 to 20,000 words, including tables, contents list and abstract but excluding the acknowledgements, bibliography, appendices and diagrams. The absolute maximum is +10% on the above i.e. 22,000 words. Please include a word count of the main text on the abstract page.
1.1 Abstract
A dissertation should begin with an abstract of up to one page (about 300 words). This should set out the main aims of the dissertation (e.g. to examine the impacts of airport privatisation in Britain), and the major stages in the work (e.g. gathering of statistical data, literature review, interviews with managers). The main findings should also be emphasised. It may be easier to draft out an abstract at an intermediate stage of the work, then finalise it when the work is completed (so that the conclusions match!).
1.2. Introduction
An introductory chapter helps to set the scene for the reader. This should refer to the general situation being examined, and recent developments prior to the dissertation work. For example, in the case of airline deregulation, this would cover the main events (in Britain, or whichever country was being examined) – government or EU acts, changes in ownership, changes in structure of the industry, overall trends in passenger and freight volumes, etc. Where a particular case study is being carried out, e.g. of an individual airline, some background information could be given on the area served, etc.
1.3 Literature review and citation of sources
An important stage in any dissertation is reviewing the existing ‘literature’ in the subject. This shows that you have checked what is already published on the subject. It may help you to identify particular issues for further investigation as part of your own original work. It may also indicate similar work with which your own work and results may be compared, e.g. a case study of the same issue in another region or country.
Guidance will be provided from the supervisor on main sources but you should also use your own initiative to find further material.
In an applied field such as air transport such literature is not only that in academic journals, but also in the industry press (e.g. Airline Business), and ‘grey literature’ reports produced by transport operators, consultancies, etc. However, care should be taken to ensure that reputable sources are being quoted for factual evidence. Many websites in particular may be set up by pressure groups or individuals – as such they may represent useful examples of viewpoints that can be quoted (e.g. on attitudes toward airport expansion) but not necessarily of detailed factual evidence.
All sources should be indicated. As far as possible you should put your findings into your own words, i.e. summarise the relevant material from the literature you have read. In a few cases, direct quotes may be appropriate, to illustrate particular points of views, or statements of policy that have been made. The following examples (from a general transport context but the principle is identical) demonstrate this:
On 24 October 2003 Richard Bowker, the Chairman of the then Strategic Rail Authority (SRA) issued a statement regarding the decision by Network Rail to bring in-house the maintenance work currently undertaken by private contractors. The following is a direct quote:
“This is the final piece in the jigsaw of reshaping the privatised industry. This is a stable base from which to just get on and deliver” (Strategic Rail Authority, 2003).
And in the bibliography as
Strategic Rail Authority (2003) Press release Friday 24 October 2003, downloaded from website <www.sra.gov.uk/news/20031014_final_piece_railway_jigsaw.tt2> on 26 October 2003.
In this case the exact words are shown in quotation marks, followed by source used. The website address is shown in the bibliography, and the date on which it was accessed. Note that ‘Strategic Rail Authority’ was introduced in full with its usual abbreviation (SRA). Subsequently, only the abbreviation needs to be used in the text. It is also useful to provide a glossary of such terms in an appendix.
Another approach could be to summarise in your own words, and then use a reference to indicate the source, e.g.
The chairman of the SRA issued a statement on 24 October 2003 supporting Network Rail’s decision to bring track maintenance in-house, instead of using private contractors (Strategic Rail Authority, 2003).
Referencing should follow the ‘Harvard system’ of referencing, i.e. to place the author’s name and date of the publication in brackets after the part of your text which mentions it. An alphabetical list is then provided at the end of the chapter or whole dissertation giving the full reference. For example:
As Gwilliam has argued, developing countries often face major institutional problems in improving their urban transport systems. Multilateral banks and aid agencies may be able to make very useful contributions to development by assisting such countries in overcoming these institutional problems (Gwilliam, 2003).
And then shown in the later reference list as:
Gwilliam, K. (2003) ‘Urban transport in developing countries’ Transport Reviews, Vol 23, no 2, pp 197-216.
Where an author has published more than one paper or report in the same year these should be shown as ‘2003a’, ‘2003b’, etc.
Where government reports etc. are used, don’t just give the overall title, but also show the table number from which any statistics are quoted, page numbers for statements made, etc.
Where no author is quoted in the original source document, use the publication name instead e.g. (The Daily Telegraph, 2009)
It is essential that all such sources are clearly cited. The University has very strict rules on plagiarism, i.e. making use of other people’s work without acknowledgement (see section 5.3 of the course handbook).
In addition to literature surveys specific to the topic you are writing about, it is also appropriate to refer to relevant general theories. For example, in discussing effects of airport privatisation there may be references on privatisation in general as well as material specific to aviation.
1.4 Methodology
The methodology of your work should be described as a separate chapter. For example, if a sample survey is to be undertaken, how the ‘population’ was defined, the sample selected, and statistical techniques to be used. These should be specific rather than referring to entire packages (e.g. ‘use of the chi-squared test’ not just ‘use of SPSS’). The degree of detail appropriate here will vary greatly according to the particular topic at which you are looking.
Where structured interview outlines, self-completion questionnaires, etc. have been used these should be described briefly in the main text, and then included as appendices for reference.
1.5 Case study description
Having introduced your case study briefly in the introduction, a much fuller description should be given, especially of any special features that might make it untypical, or affect interpretation of results (for example, if an air carrier served an area experiencing rapid economic change this would affect the volume of travel, as well as the impacts of changes in fares, service levels etc.).
1.6 Surveys and data collection (where applicable)
A clear indication should be given of the sample which was contacted (e.g. number of questionnaires sent out for reply by post) and the response (both the absolute number, and as a percentage of those contacted). In the case of face-to-face interviews, a ’strike rate’ should be estimated (i.e. proportion of those approached who were willing to participate). Where secondary data is used, a clear indication should be given of sources, and definitions employed. The use of such sources should be justified.
1.7 Analysis of evidence
This is the most important part of the dissertation. A clear description should be given of any assumptions made, e.g. grossing up from a sample to give a total. Make sure that sufficient time is allowed in the dissertation work for a full discussion of results, i.e. don’t just present results or report main findings but suggest why certain differences or outcomes have been observed. This may include some (intelligent) speculation on reasons behind this, e.g. why different age groups in the population exhibit different trip rates. Interpretation is required to draw the reader’s attention to the outcomes of the research. If using statistical tests and statistically significant differences are found, indicate where these occur and discuss why.
1.8 Summary and conclusions
Summarise your main findings, making comparisons with other studies and authors where appropriate. Summary tables and diagrams help to present these to the reader. Include recommendations for government, industry etc and any suggestions for further research. Do not introduce new evidence at this stage. Conclusions should match the abstract.
2. Some general issues in research methodology
There is a need for a systematic approach to study, covering aspects such as:
– statement of issue to be addressed/hypothesis to be tested/main question.
– identification of appropriate data sources
– selection of case studies (e.g. number, type, conditions they cover)
– is selection representative of wider population? (case studies may still be valid in their own right if not, but care is needed in drawing conclusions).
– consistency in data series (especially over time)
– forms of survey/observation to be used
– role of structured interviews/ study of management.
– justification of approach taken, and alternatives also considered
– Self-critique in light of outcome
3. The concept of ‘experimentation’ in air transport
In physical sciences, use is made of ‘laboratory’ experiments, in which ‘control cases’ are defined, to compare outcomes (generally similar to experimental cases, but not subject to deliberate variation – e.g. change in temperature). Loose parallels may be drawn in air transport, e.g. US domestic market v intra-European market. Is there an ability to control the operating environment?
‘Double blind’ concept is used in medicine (i.e. active treatment and placebo, randomly assigned to patients). A direct equivalent is virtually impossible in air transport, but we can ensure randomisation in sampling.
Some explicitly experimental work is carried out in air transport in perception and physical responses (e.g. safety issues, aircraft boarding procedures). Stated Preference work has a role roughly equivalent to laboratory experiments in ‘pure’ sciences.
In air transport, changes occur mainly in the ‘outside world’:
– operators/authorities making changes in any case, e.g. new route, new ATM procedures, contracting in versus outsourcing, partial/temporary network closures
– experimental situation may itself cause bias in response (e.g. trial of self-applied bag tags). Aviation taxes (introduced and then withdrawn in e.g. The Netherlands).
– it may be appropriate to treat real change as if experiments, i.e. compare with overall trend, paired comparisons.
– seasonality and time period issues. Quality of data prior to ‘experimental’ change.
4. Primary versus secondary data – and data collection
‘Primary’ data is that obtained directly by the researcher, e.g. through passenger surveys, management interviews, counts of air traffic. ‘Secondary’ is data already compiled and available from some other source, e.g. financial reports, traffic statistics and existing surveys from which further analysis can be made (e.g. CAA).
There is no fixed pattern to the way in which such data is used, since this will vary greatly according to the topic being investigated. Strong dissertations often combine the two, e.g. use of existing secondary data, followed by direct primary data collection (such as interviews with senior managers, or a survey of potential passengers).
Where a direct survey is being carried out by one individual, the feasible sample size will be limited. In some cases, especially part-timers employed in the industry, additional resources may be available to carry out routine data collection work. In such cases, the dissertation text must indicate clearly the degree of original work undertaken, and extent to which assistance was available. For example, where a passenger survey is carried out the wording of a questionnaire, sampling method, and analysis should all be carried out by the author of the dissertation. The greater sample size made possible than if an individual were doing all the survey work should enable a much more robust sample size to be obtained for subsequent statistical testing, etc.
5. Ethical issues
Increased importance is being given to technical issues in research, especially:
– Giving full attribution of data sources used, other work quoted, etc. (except where confidentiality is requested, and/or presumed, by the originator/author)
– Retaining confidentiality of individual respondents, especially in personal or household surveys. No indication of names or addresses should be given in any published work, or in copies placed on open access (e.g. library, internet). Any names given in illustrative examples should be assumed (not actual) and described as such. Classificatory variables may be used for purposes of analysis (e.g. age group, country of journey origin), but be careful to ensure cross-tabulation within smaller samples could not result in inadvertent disclosure.
– Confidentiality of commercial and/or politically sensitive data (but always check what may already be in the public domain, e.g. company annual reports, technical press coverage, etc.). Omit verbatim transcripts or replies to e-mails, etc, in final dissertation text. Do not indicate explicitly refusals to respond. Where interviews have taken place with decision makers, managers etc. check if they are willing to be quoted by name. If not, they should be described by their role (e.g. ‘senior manager in an airport company’).
– Particularly strong concerns arise in experimental situations where individual respondents are affected directly, e.g. in some medical or psychological work. In general this does not arise in air transport (the ‘experiment’ usually being of the form of changes affecting many people at the same time).
– The university has adopted a ‘Code of Practice Governing the Ethical Conduct of Investigations, Demonstrations, Research and Experiments’.
This requires you to login to the University’s Virtual Research Environment (VRE).
The student should complete the necessary forms consulting with their supervisor regarding the details if necessary.
Please try to do this at an early stage before the research becomes too far advanced!
6. Further reading
Most reference material will be specific to the subject being considered. However a good general guide is:
Tony Greenfield (ed) Research methods for Postgraduates. Arnold, 2002 (price approx £25, can be ordered from campus bookshop).
Some of its material is related to PhD level work and to medical or physical sciences, but most of the principles apply equally well to MSc dissertations. Particularly relevant parts/chapters are:
Part One Chapters 1, 4, 5 (re presentation), Chap 6 (ethics)
Part Two Chapter 9
Part Three Chapters 10, 12-14 (research tools)
Part Four Chapters 16-18 (role of creativity)
Part Five Chapter 22 (surveys)
Part Six Chapters 23-27 (sampling, interviewing and measurement)
Part Seven Chapters 29, 30, 32 (analysis of data)
Part Nine Chapters 37-39 (presentation)