Constructing a Dissertation
Introduction – The overall structure of a dissertation, and the weight given to various parts, should be a personal decision – otherwise dissertations would be depressingly formulaic. There are some useful guidelines and structural principles which you can use to guide your decisions – use these guidelines as suggestions only. Plan the structure of your dissertation with your personal tutor reasonably early
next year – this will help establish the goals of your work.
With a good structural plan you will be able to produce useful drafts of sections at many points through the year, enabling you to make progress when other aspects of your work get stuck.
Some components of a dissertation are standard, and have to appear in some guise in whatever structure you choose. Here are the most common:
• title page and declaration (regulation)
• methodology and research design
• theoretical background/literature survey
• analysis of results
• conclusions/future directions
• glossary (?)
There are also standard components for the more formulaic types of dissertation such as a software construction project or a methodologically-guided analysis, note that these always end with an ‘analysis’ section appropriate to the investigation which overlays the ‘analysis of results’ in the list above:
A software construction project might contain:
• design goals
• highlights of construction process: design/code/test (structured around whatever life- cycle was used)
• evaluation of software against goals
A methodologically-guided analysis might contain:
• organisation and its problems
• operation of methodology: investigation/ modelling/ improved design (structured around whatever methodology was used)
• recommendations to organisation
Similar standard components might apply in the case of standard experimental or investigative processes such as the evaluation of a user interface using a well-defined approach.
A Closer Look at Some Components.
• used to orient someone who is approaching your work ‘blind’ (your dissertation is written for an intelligent peer, who doesn’t know the specifics of your work)
• should be written in a clear ‘tutorial’ style, you are trying to give someone insight and motivation, convincing them that your problem is interesting and worthwhile
• you are saying how your problem arises, for whom it is a problem, and what you think might be done about it
• this can be built around your ‘research question and hypothesis’, formulated in the RM workbook – it’s a carefully-explained expansion of it
methodology and research design:
• you are setting out your plan of campaign, what you intend to do about your problem, how you are going to answer your research question using established methodologies and a well-designed strategy
• this is a careful statement of intent which the reader will see fulfilled in the remainder of the dissertation
• this isn’t just setting out a chosen analysis method or lifecycle model, it’s your entire approach (of which specific methods are a part)
• this can be built around the corresponding section of the RM workbook – again, it’s a carefully-explained expansion of it
theoretical background/literature survey:
• your work will be based on a reading of background theory which an intelligent peer cannot be expected to have done, this section provides a summary
• this section sets out the foundation for the content-specific sections which are to come
• this section should be built around explanations, quotations and references; there should be enough information for someone to reconstruct your thought processes
• the language of this section should be scholarly and precise, you are trying to compress a lot of thought into a small space
• this section can be written anywhere, perhaps in the Xmas vacation when you have time for quiet reflection(!) – make sure you have the books or photocopies to do so
• some research books give the impression that literature surveys must be complete, and give recognition to all previous work. For an undergraduate, that is hardly realistic; we will be very content with evidence of broad reading and serious intent
analysis of results:
• not just a presentation, but a logical analysis of what you have found out
• explain the significance of your results and draw precise technical conclusions from them
• the kind of analysis carried out depends on the domain of the investigation and the kinds of results obtained – it might range from precise quantititative analysis to an informed, discursive analysis
• this analysis is the ’rounding off’ of your methodology and research design section – where all that planning pays off
• a summing-up of what you have achieved, not precise technical conclusions (they should have come earlier)
• can be written in a personal, reflective style; say what you have learned, what you might have done and didn’t, how you think the work might be developed in the future, etc.
• a sort of ‘mirror image’ of the overview and method sections, reflecting on whether you think your question was well-chosen and your overall strategy sound, how you might have refined them
• don’t feel constrained to be up-beat, the most is often learned from poorly-framed questions and flawed strategies, honesty and realism are key.
• references, list all of the works referred to in the text, using some standard system (the library produces a leaflet – I would suggest the Harvard system)
• bibliography, list general background reading which you did, but which you did not reference specifically in the text
• don’t list works just for the sake of it, but a healthy list (20-30 entries at the very least, more for the kind of wide-ranging studies that might occur in the more cultural reaches of multimedia, or the more socially-engaged parts of IS) should be the natural outcome of well-conducted research (this also applies in technical areas such as a software construction project)
• self-contained units of background information, referenced in the main body of the text
• may be complex diagrams or tables, fragments of interview or survey data, test results, etc.
• you should never place anything explanatory or orginal in an appendix – but it can be a useful repository if you have exceeded the permitted word count
• might be useful if you are writing in a technical area with many acronyms