Writing Technical Articles

The notes below apply to technical papers in computer science and electrical engineering, with emphasis on papers in systems and networks.

Read Strunk and White, Elements of Style. Again.

Give the paper to somebody else to read. If you can, find two people: one person familiar with the technical matter, another only generally familiar with the area.

Papers can be divided roughly into two categories, namely original research papers and survey papers. There are papers that combine the two elements, but most publication venues either only accept one or the other type or require the author to identify whether the paper should be evaluated as a research contribution or a survey paper. (Most research papers contain a "related work" section that can be considered a survey, but it is usually brief compared to the rest of the paper and only addresses a much narrower slice of the field.)

Research Papers

A good research paper has a clear statement of the problem the paper is addressing, the proposed solution(s), and results achieved. It describes clearly what has been done before on the problem, and what is new.

The goal of a paper is to describe novel technical results. There are four types of technical results:

1. An algorithm;
2. A system construct: such as hardware design, software system, protocol, etc.;
One goal of the paper is to ensure that the next person who designs a system like yours doesn't make the same mistakes and takes advantage of some of your best solutions. So make sure that the hard problems (and their solutions) are discussed and the non-obvious mistakes (and how to avoid them) are discussed. (Craig Partridge)
3. A performance evaluation: obtained through analyses, simulation or measurements;
4. A theory: consisting of a collection of theorems.

A paper should focus on
> describing the results in sufficient details to establish their validity;
> identifying the novel aspects of the results, i.e., what new knowledge is reported and what makes it non-obvious;
> identifying the significance of the results: what improvements and impact do they suggest.

Paper Structure

> Typical outline of a paper is:
-- Abstract, typically not more than 100-150 words;
-- Introduction (brief!): introduce problem, outline solution; the statement of the problem should include a clear statement why the problem is important (or interesting).
-- Related Work (or before summary). Hint: In the case of a conference, make sure to cite the work of the PC co-chairs and as many other PC members as are remotely plausible, as well as from anything relevant from the previous two proceedings. In the case of a journal or magazine, cite anything relevant from last 2-3 years or so volumes.
-- Outline of the rest of the paper: "The remainder of the paper is organized as follows. In Section 2, we introduce ..Section 3 describes ... Finally, we describe future work in Section 5." [Note that Section is capitalized. Also, vary your expression between "section" being the subject of the sentence, as in "Section 2 discusses ..." and "In Section, we discuss ...".]
-- Body of paper
.. problem
.. approach, architecture
.. results

The body should contain sufficient motivation, with at least one example scenario, preferably two, with illustrating figures, followed by a crisp generic problem statement model, i.e., functionality, particularly emphasizing "new" functionality. The paper may or may not include formalisms. General evaluations of your algorithm or architecture, e.g., material proving that the algorithm is O(log N), go here, not in the evaluation section.

> Architecture of proposed system(s) to achieve this model should be more generic than your own peculiar implementation. Always include at least one figure.

> Realization: contains actual implementation details when implementing architecture isn't totally straightforward. Mention briefly implementation language, platform, location, dependencies on other packages and minimum resource usage if pertinent.

> Evaluation: How does it really work in practice? Provide real or simulated performance metrics, end-user studies, mention external technology adoptors, if any, etc.

-- Related work, if not done at the beginning
-- Summary and Future Work
.. often repeats the main result
-- Acknowledgements
-- Bibliography
-- Appendix (to be cut first if forced to):
.. detailed protocol descriptions
.. proofs with more than two lines
.. other low-level but important details

It is recommended that you write the approach and results sections first, which go together. Then problem section, if it is separate from the introduction. Then the conclusions, then the intro. Write the intro last since it glosses the conclusions in one of the last paragraphs. Finally, write the abstract. Last, give your paper a title.

Title

> Avoid all but the most readily understood abbreviations.

> Avoid common phrases like "novel", "performance evaluation" and "architecture", since almost every paper does a performance evaluation of some architecture and it better be novel. Unless somebody wants to see 10,000 Google results, nobody searches for these types of words.

Use adjectives that describe the distinctive features of your work, e.g., reliable, scalable, high-performance, robust, low-complexity, or low-cost. (There are obviously exceptions, e.g., when the performance evaluation is the core of the paper. Even in that case, something more specific is preferable, as in "Delay measurements of X" or "The quality of service for FedEx deliveries".)

> If you need inspiration for a paper title, you can consult the Automatic Systems Research Topic or Paper Title Generator.

Authors

> The IEEE policies (Section 6.4.1) used to state the following about authorship:
The IEEE affirms that authorship credit must be reserved for individuals who have met each of the following conditions: 1) made a significant intellectual contribution to the theoretical development, system or experimental design, prototype development, and/or the analysis and interpretation of data associated with the work contained in the manuscript, 2) contributed to drafting the article or reviewing and/or revising it for intellectual content, and 3) approved the final version of the manuscript, including references.

This has now moved to the IEEE PSPB Operations Manual, Section 8.2.1.

Abstract

> The abstract must not contain references, as it may be used without the main article. It is acceptable, although not common, to identify work by author, abbreviation or RFC number. (For example, "Our algorithm is based upon the work by Smith and Wesson.")

> Avoid use of "in this paper" in the abstract. What other paper would you be talking about here?

> Avoid general motivation in the abstract. You do not have to justify the importance of the Internet or explain what QoS is.

> Highlight not just the problem, but also the principal results. Many people read abstracts and then decide whether to bother with the rest of the paper.

> Since the abstract will be used by search engines, be sure that terms that identify your work are found there. In particular, the name of any protocol or system developed and the general area ("quality of service", "protocol verification", "service creation environment") should be contained in the abstract.

> Avoid equations and math. Exceptions: Your paper proposes E = m c 2.

Introduction

> Avoid stock and cliche phrases such as "recent advances in XYZ" or anything alluding to the growth of the Internet.

> Be sure that the introduction lets the reader know what this paper is about, not just how important your general area of research is. Readers won't stick with you for three pages to find out what you are talking about.

> The introduction must motivate your work by pinpointing the problem you are addressing and then give an overview of your approach and/or contributions (and perhaps even a general description of your results). In this way, the intro sets up my expectations for the rest of your paper -- it provides the context, and a preview.

> Repeating the abstract in the introduction is a waste of space.

> Example bad introduction:

Here at the institute for computer research, me and my colleagues have created the SUPERGP system and have applied it to several toy problems. We had previously fumbled with earlier versions of SUPERGPSYSTEM for a while. This system allows the programmer to easily try lots of parameters, and problems, but incorporates a special constraint system for parameter settings and LISP S-expression parenthesis counting.

The search space of GP is large and many things we are thinking about putting into the supergpsystem will make this space much more colorful.

> A pretty good introduction, drawn from Eric Siegel's class:
Many new domains for genetic programming require evolved programs to be executed for longer amounts of time. For example, it is beneficial to give evolved programs direct access to low-level data arrays, as in some approaches to signal processing \cite{teller5}, and protein segment classification \cite{handley,koza6}. This type of system automatically performs more problem-specific engineering than a system that accesses highly preprocessed data. However, evolved programs may require more time to execute, since they are solving a harder task.

Previous or obvious approach:
(Note that you can also have a related work section that gives more details about previous work.)) One way to control the execution time of evolved programs is to impose an absolute time limit. However, this is too constraining if some test cases require more processing time than others. To use computation time efficiently, evolved programs must take extra time when it is necessary to perform well, but also spend less time whenever possible.

Approach/solution/contribution:
The first sentence of a paragraph like this should say what the contribution is. Also gloss the results.

In this chapter, we introduce a method that gives evolved programs the incentive to strategically allocate computation time among fitness cases. Specifically, with an aggregate computation time ceiling imposed over a series of fitness cases, evolved programs dynamically choose when to stop processing each fitness case. We present experiments that show that programs evolved using this form of fitness take less time per test case on average, with minimal damage to domain performance. We also discuss the implications of such a time constraint, as well as its differences from other approaches to {\it multiobjective problems}. The dynamic use of resources other than computation time, e.g., memory or fuel, may also result from placing an aggregate limit over a series of fitness cases.

Overview:
The following section surveys related work in both optimizing the execution time of evolved programs and evolution over Turing-complete representations. Next we introduce the game Tetris as a test problem. This is followed by a description of the aggregate computation time ceiling, and its application to Tetris in particular. We then present experimental results, discuss other current efforts with Tetris, and end with conclusions and future work.

Body of Paper

Hints and common mistakes

Bibliography

> Avoid use of et al. in a bibliography unless list is very long (five or more authors). The author subsumed into et al. may be your advisor or the reviewer... Note punctuation of et al..

> If writing about networks or multimedia, use the network bibliography. All entries not found there should be sent to me. A listing of frequently-used references for networks is available.

> Internet drafts must be marked ``work in progress''. Make sure that they have been replaced by newer versions or RFCs. Any Internet Draft reference older than six months should automatically be suspicious since Internet Drafts expire after that time period.

> Book citations include publication years, but no ISBN number.

> It is now acceptable to include URLs to material, but it is probably bad form to include a URL pointing to the author's web page for papers published in IEEE and ACM publications, given the copyright situation. Use it for software and other non-library material. Avoid long URLs; it may be sufficient to point to the general page and let the reader find the material. General URLs are also less likely to change.

> Leave a space between first names and last name, i.e., "J. P. Doe", not "J.P.Doe".
References such as
John Doe, "Some paper on something", technical report.
are useless. Cite the source, date and other identifying information.

> For conference papers, you MUST name the conference location, month and the full conference name, not just some abbreviation. Page numbers are nice, but optional. All of this information is readily available via the IEEE or ACM digital libraries.

> Check if Internet drafts have been published as RFCs or if there's a newer version.

> Having a citation
Jane Doe, "Some random paper", to be published, 2003.
is useless, as the paper has presumably been published by now. Google or the ACM or IEEE digital libraries will help you find it.

Acknowledgements

> Acknowledge your funding sources. Some sources have specific wording requirements and may prefer that the grant number is listed. The NSF requires text like "This work was supported by the National Science Foundation under grant EIA NN-NNNNN."

> Generally, anonymous reviewers don't get acknowledged, unless they really provided an exceptional level of feedback or insight. Rather than "We thank X for helping us with Y", you might vary this as "X helped with Y.".

Reporting Numerical Results and Simulations

In all but extended abstracts, numerical results and simulations should be reported in enough detail that the reader can duplicate the results. This should include all parameters used, indications of the number of samples that contributed to the analysis and any initial conditions, if relevant.

When presenting simulation results, provide insight into the statistical confidence. If at all possible, provide confidence intervals. If there's a "strange" behavior in the graph (e.g., a dip, peak or change in slope), this behavior either needs to be explained or reasons must be given why this is simply due to statistical aberration. In the latter case, gathering more samples is probably advised.

Figures should be chosen wisely. You can never lay out the whole parameter space, so provide insight into which parameters are significant over what range and which ones are less important. It's not very entertaining to present lots of flat or linear lines.

The description of the graph should not just repeat the graphically obvious such as "the delay rises with the load", but explain, for example, how this increase relates to the load increase. Is it linear? Does it follow some well-known other system behaviors such as standard queueing systems?

LaTeX Considerations

There's no need to enclose numbers in $$ (math mode).
Use \cite{a,b,c}, not \cite{a} \cite{b} \cite{c}.
Use the \usepackage{times} option for LaTeX2e - it comes out much nicer on printers with different resolutions. Plus, compared to cmr, it probably squeezes an extra 10% of text out of your conference allotment.
Multi-letter subscripts are set in roman, not italics. For example,
x_{\mathrm max}
For uniformity, use the LaTeX2e graphics set, not the earlier psfigure set:
\usepackage{graphics}
...
\begin{figure}
\resizebox{!}{0.5\textheight}{\includegraphics{foo.eps}}
\caption{Some figure}
\label{fig:figure}
\end{figure}
Things to Avoid

Too much motivational material
Three reasons are enough -- and they should be described very briefly.
Describing the obvious parts of the result
"Obvious" is defined as any result that a graduate of our program would suggest as a solution if you pose the problem that the result solves.
Describing unnecessary details
A detail is unnecessary, if its omission will not harm the reader's ability to understand the important novel aspects of the result.
Spelling errors
With the availability of spell checkers, there is no reason to have spelling errors in a manuscript. If you as the author didn't take the time to spell-check your paper, why should the editor or reviewer take the time to read it or trust that your diligence in technical matters is any higher than your diligence in presentation? Note, however, that spell checkers don't catch all common errors, in particular word duplication ("the the"). If in doubt, consult a dictionary such as the (on line) Merriam Webster.
Text in Arial:
Arial and other sans-serif fonts are fine for slides and posters, but are harder to read in continuous text. Use Times Roman or similar serif fonts. Unusual fonts are less likely to be available at the recipient and may cause printing or display problems.
Guidelines for Experimental Papers

"Guidelines for Experimental Papers" set forth for researchers submitting articles to the journal, Machine Learning.
Papers that introduce a new learning "setting" or type of application should justify the relevance and importance of this setting, for example, based on its utility in applications, its appropriateness as a model of human or animal learning, or its importance in addressing fundamental questions in machine learning.
Papers describing a new algorithm should be clear, precise, and written in a way that allows the reader to compare the algorithm to other algorithms. For example, most learning algorithms can be viewed as optimizing (at least approximately) some measure of performance. A good way to describe a new algorithm is to make this performance measure explicit. Another useful way of describing an algorithm is to define the space of hypotheses that it searches when optimizing the performance measure.
Papers introducing a new algorithm should conduct experiments comparing it to state-of-the-art algorithms for the same or similar problems. Where possible, performance should also be compared against an absolute standard of ideal performance. Performance should also be compared against a naive standard (e.g., random guessing, guessing the most common class, etc.) as well. Unusual performance criteria should be carefully defined and justified.
All experiments must include measures of uncertainty of the conclusions. These typically take the form of confidence intervals, statistical tests, or estimates of standard error. Proper experimental methodology should be employed. For example, if "test sets" are used to measure generalization performance, no information from the test set should be available to the learning process.
Descriptions of the software and data sufficient to replicate the experiments must be included in the paper. Once the paper has appeared in Machine Learning, authors are strongly urged to make the data used in experiments available to other scientists wishing to replicate the experiments. An excellent way to achieve this is to deposit the data sets at the Irvine Repository of Machine Learning Databases. Another good option is to add your data sets to the DELVE benchmark collection at the University of Toronto. For proprietary data sets, authors are encouraged to develop synthetic data sets having the same statistical properties. These synthetic data sets can then be made freely available.
Conclusions drawn from a series of experimental runs should be clearly stated. Graphical display of experimental data can be very effective. Supporting tables of exact numerical results from experiments should be provided in an appendix.
Limitations of the algorithm should be described in detail. Interesting cases where an algorithm fails are important in clarifying the range of applicability of an algorithm.
Other Hints and Notes

From Bill Stewart (Slashdot, May 7, 2006), edited

Write like a newspaper reporter, not a grad student.
Your objective is clear communication to the reader, not beauty or eruditeness or narration of your discoveries and reasoning process. Don't waste their time, or at least don't waste it up front.
Hit the important conclusions in the first few sentences so your reader will read them. If you'd like to wrap up with them at the end of your memo, that's fine too, in case anybody's still reading by then, but conclusions come first.
If you're trying to express something complex, simplify your writing so it doesn't get in the way. For something simple, 10th grade language structures will do, but if it's really hairy stuff, back down to 8th grade or so.
Think about what your audience knows and doesn't know, and what they want and don't want. Express things in terms of what they know and want, not what you know.
From MarkusQ, Slashdot, May 7, 2006

Top down design Starting with an outline and working out the details is the normal way of tackling an engineering problem.
Checking your facts Engineers should be used to checking anything that is even remotely doubtful before committing to it. So should writers.
Failure mode analysis For each sentence ask yourself, could it be misread? How? What is the best way to fix it?
Dependency analysis Are the ideas presented in an order that assures that each point can be understood on the basis of the readers assumed knowledge and the information provided by preceding points?
Optimization Are there any unnecessary parts? Does the structure require the reader to remember to many details at once, before linking them?
Structured testing If you read what you have written assuming only the knowledge that the reader can be expected to have, does each part work the way you intended? If you read it aloud, does it sound the way you intended?
The Conference Review Process

It is hard to generalize the review process for conferences, but most reputable conferences operate according to these basic rules:

The paper is submitted to the technical program chair(s). Many current conferences require electronic submission, in either PostScript or PDF formats, occasionally in Word.
The technical program chair assigns the paper to one or more technical program committee members, hopefully experts in their field. The identity of this TPC member is kept secret.
The TPC member usually provides a review, but may also be asked to find between one and three reviewers who are not members of the TPC. They may be colleagues of the reviewer at the same institution, his or her graduate students or somebody listed in the references. The graduate student reviews can be quite helpful, since these reviewers often provide more detailed criticism rather than blanket dismissal. Any good conference will strive to provide at least three reviews, however, since conferences operate under tight deadlines and not all reviewers deliver as promised, it is not uncommon that you receive only two reviews.
In some conferences, there is an on-line discussion of papers among the reviewers for a particular paper. Usually, a lead TPC member drives the discussion and then recommends the paper for acceptance, rejection or discussion at the TPC meeting.
The technical program chair then collects the reviews and sorts the papers according to their average review scores.
The TPC (or, rather, the subset that can make the meeting), then meets in person or by phone conference. Usually, the bottom third and the top third are rejected and accepted, respectively, without (much) further discussion. The papers discussed are those in the middle of the range, or where a TPC member feels strongly that the paper ended up in the wrong bin, or where the review scores differ significantly. Papers that only received two reviews are also often discussed, maybe with a quick review by one of the TPC members as additional background. The rigor of the TPC meeting depends on the size and reputation of the conference. In some workshops and conferences, the TPC chairs may well make the final decision themselves, without involving the whole TPC.
Other References

How to give a good research talk and how to give a good research paper
How to read a paper, CCR, July 2007.
How to Write a PhD Thesis
Top 10 tips for writing a paper, by Jim Kurose
Bartleby has dictionaries, grammars, an encyclopedia, and Columbia Guide To Standard American English
The Free Dictionary, also an online dictionary
Dictionarist: multi-lingual dictionary
IEEE IEEE Computer Society style guide, including hints on how to reference Internet drafts and RFCs
USC editorial style guide
Berkeley Information Systems and Technology Publications style guide
Guide to Grammar and Style, by J. Lynch
alt.usage.english FAQ, addressing common grammar questions
Writing Tips
Key to common comments made on your papers
Drafting the Paper in an Academic Style
Religious Studies style sheet
Cisco style guide
Oded Goldreich wrote an essay entitled "How not to write a paper", with recommendations on style and organization.
Don Knuth has online the TeX source of a book on "Mathematical Writing" (also useful for Computer Science).
The structure of paper/report in Systems, by Michalis Faloutsos, U.C. Riverside
The Elements of Style. William Strunk Jr. and E.B. White. Macmillan Publishing Co., New York, 1979.
This is an amazing little book that you can read in a few hours. Your writing style will never be the same afterwards. This $8 book is the best investment you can ever make.
"A Guide to Writing as an Engineer"
Spring into Technical Writing for Engineers and Scientists
Bugs in Writing. Lyn Dupre', (2nd ed.)
This is a great book that expands on Strunk&White. It has more examples, and was written by an author who edited numerous technical publications, many of which were in computer science.
The Chicago Manual of Style, Univ. of Chicago Press.
This is the bible for American academic style. It's long and heavy, but has everything you ever want to know about style. When in doubt, or if you get conflicting stylistic advice, following The Chicago Manual of Style is your best choice.
A Handbook for Scholars by Mary Claire van Leunen; Alfred Knopf, Publisher.
This is another useful book written for publishing (computer) scientists.
The UIST Guide for Authors is geared towards a specific conference, but the general process and guidelines are similar to many other conferences.
The Science of Scientific Writing, George D. Gopen and Judith A. Swan, In American Scientist, Vol. 78, No. 6 (Nov-Dec, 1990), pp. 550-558.
This is a useful article that teaches scientists how to write single sentences and paragraphs, such as "follow a grammatical subject as soon as possible with its verb" and "articulate the action of every clause or sentence in its verb".
The Mayfield Handbook of Technical and Scientific Writing, Perelman, Paradis and Barrett, Mayfield, 1998.
It is an extensive resource explaining how to write papers, reports, memoranda and Ph.D. thesis, how to make high-performance slides and oral presentations, how to avoid common pitfalls and mistakes in English, etc., with many examples of "good" and "bad" practices.
Roy Levin and David D. Redell, An evaluation of the ninth SOSP submissions -or- How (and how not) to write a good systems paper, ACM SIGOPS Operating Systems Review 17(3):35-40 (July, 1983).
Alan Snyder, How to get your paper accepted at OOPSLA, OOPSLA '91 Proceedings, pp. 359-363.
Mark Allman, A Referee's Plea, 2001
Ralph Johnson et al, How to get a paper accepted at OOPSLA, Panel at OOPSLA'93, pp 429-436.
Craig Partridge, How to Increase the Chances Your Paper is Accepted at ACM SIGCOMML.
Generally useful advice that also applies to other networking conferences.
What kinds of papers does USENIX publish?
Alan Jay Smith, The Task of the Referee, IEEE Computer 23(4):65-71 (April 1990).
Grammar, Punctuation, and Capitalization, NASA SP-7084
Stylewriter software
Talks

"The Short Talk" (Charles Van Loan)
"Pointers on giving a talk" (D. Messerschmitt)
Tips for Preparing Delivering Scientific Talks and Using Visual Aids (ONR)
Miscellaneous

International Standard Paper Sizes
Contributors

This page contains material provided by Gail Kaiser, Craig Partridge, Sumit Roy, Eric Siegel, Sal Stolfo, Luca Trevisan, Yechiam Yemini, Erez Zadok.

Hints for PhD Proposal Defenses

PhD proposal defenses in Computer Science allow student audience; this is a good opportunity to find out what works and doesn't from your more senior colleagues.
Proposal defenses consist of four parts: first, the candidate introduces themselves, then presents a summary of their work, interrupted and followed by questions from the committee. Finally, the committee meets in private to discuss the presentation and the plan.

While most of the committee will have read most of your proposal, you cannot assume that everyone has read every page in detail.

Avoid high-level talks: "... they usually fail to convey the intellectual substance, creativity, ingenuity of the speakers' accomplishments - what takes the work out of the routine. Naturally, these comments apply to all of our speakers who want to impress people with their ability as opposed to the breadth of their knowledge or the size of their project." (Ed Coffman)

When presenting experimental work, be prepared to defend your methodology. What was your sample size? Confidence intervals?

Standard presentation guidelines apply:
Talk to your audience, not to your slides.
Project; speaking softly conveys the impression that you are unsure of what you are saying.

Make sure that all your graphs are readable. Check this in the actual presentation environment (using a video projector), not just on your laptop screen. A common problem is that the lines are too thin.

Avoid flashy or cheesy animations, such as animated GIFs, or PowerPoint word art. This is not a sales talk and these gimmicks distract from the message and make you look unprofessional.

Keep to the allotted time of no more than 45 minutes.
Your presentation needs to address the following:
What is the problem you are studying?
Why is it important?
What results have you achieved so far and why to they matter?
How is this substantially different from prior work?
What do you need to do to complete your work?

Your workplan should be sufficiently detailed so that the committee can judge whether it is realistic or not. You don't have to account for every day between the proposal and your thesis defense, but a roughly monthly or quarterly granularity is to be expected, depending on how far away your anticipated graduation date is. Specify the experiments you need to run, the software you need to write and the algorithms you want to try out. This should not just be one page that says "I will do miraculous things".

The committee should be handed a copy of your slides.
No more than 25 slides, plus "back up" slides with additional material in case of questions. The committee will get anxious once the presentation lasts longer than 35-40 minutes.

List your contributions early and explicitly. You don't want to create the impression that related work is yours, and vice versa.

One of the most important concerns during the proposal is to convince the audience that you are aware of all related work. Since some of your work may date back a few years, it is not sufficient to just copy the reference list from your first paper. Check common recent conferences to see whether any recent work applies to your thesis. If applicable, point out your work predates work presented by somebody else done more recently. (Given the duration of most theses, it is not uncommon that others pursue a direction after you have stopped working on it.)

When presenting your contributions, be sure to use "I" and not "we" so that the committee will know what aspects of the work where yours, and which were group projects.

You must convey a clear plan how you are going to evaluate your work systematically - by measurement, simulation, user experiments. This is a core part what makes computer science science and not just software-building.

Be prepared to back up any comparative statement with facts, in particular statements like "works better", "faster", "scalable" or "optimal". If you are presenting a protocol, how do you know that it works correctly? If your algorithm is optimal, can you prove that it is? (If not, avoid the term.)

Writing a Research Proposal

http://nottingham.ac.uk/business/phd/Proposal.html

A typical research proposal will be somewhere between one and two thousand words. While we do not insist on a definite format, we encourage students to keep the following in mind:
The proposal should begin by explaining the subject area in which the research is to be located, and providing an indication of the key theoretical, policy or empirical debates it plans to address.

The proposal should then present a brief review of the literature you plan to contribute to in conducting your own research. You need to demonstrate a familiarity with the relevant academic literature and theories relating to your research proposal, and an awareness of the major lines of argument that have been developed in your chosen research field. You then need to discuss the research questions you plan to address. Importantly, you need to demonstrate the manner in which your research questions emanate from: gaps in the existing empirical literature; from the application of a particular theory in a specific industry or national context; or from a synthesis of a number of bodies of literature, for example. Although not essential, research questions that are topical or have policy relevance will be particularly welcome.

You research proposal should be as specific and focused as possible. If your research is being driven by gaps in the existing literature, which of these gaps will you attempt to address? If your research is being driven by theoretical or policy debates, which specific points of these debates are you going to focus on?

The research proposal can also provide some explanation of what led you to the topic. Thus, if your topic emerges from a long-standing interest or from interests you developed while studying for a former qualification, do not hesitate to impart this information - it will help to convey your motivation for pursuing doctoral studies.

Depending on the nature of your subject, it is desirable to give some indication of the research methods that will be used to conduct the research. A variety of different research methods exists, so you should seek to identify the method that is most suited to your area of research - econometric modelling, participant observation, survey data analysis, case-study analysis, the analysis of historical records, for example.


Most doctoral work involves empirical research. The successful completion of doctoral work in the time allotted thus often depends on the ability to obtain the data needed. If your proposed research involves empirical work you should provide an indication of how you might collect any relevant data. For example, you might like to say something about access to particular sources of information (whether you will need on-line access to databases, whether you can use relevant archives etc). You could also comment on the country or geographical region in which the study will take place and explain why you have chosen this country/ region. It might also be worth saying something about unit of analysis for the research (whether you are looking at individuals, groups, workplaces, companies etc) and provide some justification for your choice of unit of analysis. Importantly, you need to explain the manner in which the data you collect will enable you to address your research questions.

You should give special attention to the feasibility of data collection. Your proposal may contain interesting and highly relevant research questions, and it may be well grounded in the literature, but it may not be a practical research enterprise. You must balance the scope of your proposal against the practical problems of data gathering. Does your research proposal call for special access to managers or organisations? How many potential variables or factors does it require you to address? Can you examine all of them? Students whose projects involve data gathering in countries other than the UK are advised to pay close attention to the issue of feasibility. Students whose projects depend on data that are not available in the UK should indicate in their proposal how they propose to finance their data gathering.

Although no indication of the research findings can be presented, it is often beneficial to conclude the research proposal by indicating the contribution you envisage that your research will make to the literature in your particular subject area, or by indicating the potential practical or policy implications of your research. This means providing an indication of the extent to which you feel your research will make an original contribution, suggesting how it may fill gaps in existing research, and showing how it may extend understanding of particular topics.

While your research proposal is judged mainly on content, it must also look professional. It should be typed and written in good English. If you are submitting a photocopy, make sure it is of the highest quality. Particular attention will be paid to clarity of expression and also the structure, coherence and flow of argument. Finally, always include a bibliography (in a standard format – e.g. Harvard) with your research proposal that lists books and articles to which you make reference in your proposal.

Finally, it is natural for ideas to evolve and change, so you will not be obliged to adhere to the specifics of your proposal if you are offered a place on the programme. However, the proposal is the foundation of your working relationship with your supervisors and thus it cannot be radically altered without discussion and consultation with your supervisors.

How to Write a PhD Thesis

This guide to thesis writing gives simple and practical advice on the problems of getting started, getting organised, dividing the huge task into less formidable pieces and working on those pieces. It also explains the practicalities of surviving the ordeal. It includes a suggested structure and a guide to what should go in each section. It was originally written for graduate students in physics, and most of the specific examples given are taken from that discipline. Nevertheless, the feedback from users indicates that it has been widely used and appreciated by graduate students in diverse fields in the sciences and humanities.

Getting Started

When you are about to begin, writing a thesis seems a long, difficult task. That is because it is a long, difficult task. Fortunately, it will seem less daunting once you have a couple of chapters done. Towards the end, you will even find yourself enjoying it---an enjoyment based on satisfaction in the achievement, pleasure in the improvement in your technical writing, and of course the approaching end. Like many tasks, thesis writing usually seems worst before you begin, so let us look at how you should make a start.
An outline

First make up a thesis outline: several pages containing chapter headings, sub-headings, some figure titles (to indicate which results go where) and perhaps some other notes and comments. There is a section on chapter order and thesis structure at the end of this text. Once you have a list of chapters and, under each chapter heading, a reasonably complete list of things to be reported or explained, you have struck a great blow against writer's block. When you sit down to type, your aim is no longer a thesis---a daunting goal---but something simpler. Your new aim is just to write a paragraph or section about one of your subheadings. It helps to start with an easy one: this gets you into the habit of writing and gives you self-confidence. In an experimental thesis, the Materials and Methods chapter is often the easiest to write – just write down what you did; carefully, formally and in a logical order.

How do you make an outline of a chapter? For most of them, you might try the method that I use for writing papers, and which I learned from my thesis adviser (Stjepan Marcelja): Assemble all the figures that you will use in it and put them in the order that you would use if you were going to explain to someone what they all meant. You might as well rehearse explaining it to someone else---after all you will probably give several talks based on your thesis work. Once you have found the most logical order, note down the key words of your explanation. These key words provide a skeleton for much of your chapter outline.

Once you have an outline, discuss it with your adviser. This step is important: s/he will have useful suggestions, but it also serves notice that s/he can expect a steady flow of chapter drafts that will make high priority demands on his/her time. Once you and your adviser have agreed on a logical structure, s/he will need a copy of this outline for reference when reading the chapters which you will probably present out of order. If you have a co-adviser, discuss the outline with him/her as well, and present all chapters to both advisers for comments.

Organisation

It is encouraging and helpful to start a filing system. Open a word-processor file for each chapter and one for the references. You can put notes in these files, as well as text. While doing something for Chapter n, you will think "Oh I must refer back to/discuss this in Chapter m" and so you put a note to do so in the file for Chapter m. Or you may think of something interesting or relevant for that chapter. When you come to work on Chapter m, the more such notes you have accumulated, the easier it will be to write.

Make a back-up of these files and do so every day at least (depending on the reliability of your computer and the age of your disk drive). Do not keep back-up close to the computer in case the hypothetical thief who fancies your computer decides that s/he could use some disks or membory as well.

A simple way of making a remote back-up is to send it as an email attachment to a consenting email correspondent, preferably one in a different location. You could also send it to yourself. In either case, be careful to dispose of superseded versions so that you don't waste disk space, especially if you have bitmap images or other large files.

You should also have a physical filing system: a collection of folders with chapter numbers on them. This will make you feel good about getting started and also help clean up your desk. Your files will contain not just the plots of results and pages of calculations, but all sorts of old notes, references, calibration curves, suppliers' addresses, specifications, speculations, letters from colleagues etc., which will suddenly strike you as relevant to one chapter or other. Stick them in that folder. Then put all the folders in a box or a filing cabinet. As you write bits and pieces of text, place the hard copy, the figures etc in these folders as well. Touch them and feel their thickness from time to time – ah, the thesis is taking shape.

If any of your data exist only on paper, copy them and keep the copy in a different location. Consider making a copy of your lab book. This has another purpose beyond security: usually the lab book stays in the lab, but you may want a copy for your own future use. Further, scientific ethics require you to keep lab books and original data for at least ten years, and a copy is more likely to be found if two copies exist.

If you haven't already done so, you should archive your electronic data, in an appropriate format. Spreadsheet and word processor files are not suitable for long term storage. Archiving data by Joseph Slater is a good guide.

While you are getting organised, you should deal with any university paperwork. Examiners have to be nominated and they have to agree to serve. Various forms are required by your department and by the university administration. Make sure that the rate limiting step is your production of the thesis, and not some minor bureaucratic problem.
A note about word processors

One of the big FAQs for scientists: is there a word processor, ideally one compatible with MS Word, but which allows you to type mathematical symbols and equations conveniently? One solution is LaTeX, which is powerful, elegant, reliable, fast and free from http://www.latex-project.org/ or http://www.miktex.org/. As far as I know, the only current equation editor for MS Word is slow and awkward. (If anyone knows a way of writing equations in this software without using the mouse, many people including this author would like to hear from you!) Another solution is to use old versions of commercial software. Word 5.1 allows equations to be typed comfortably: it is faster in this respect than LaTeX, with the added advantage of 'what you see is what you get' (WYSIWYG). (If anyone knows how to run Word 5.1 on OSX, please let me know!) A search will find sites that provide discontinued software, but, not knowing whether this is legal or not, I shan't link to them. (I am told that LyX, available free at http://www.lyx.org/, is a convenient front-end to LaTeX that has WYSIWYG. )

Commercial word processors have gradually become bigger, slower, less reliable and more awkward to use as they acquire more features. This is a general feature of commercial software and an important input to the computing industry. If software and operating system performance did not deteriorate, people would not need to buy new computers and profits would fall for makers of both hard- and soft-ware. Software vendors want it to look fancy and obvious in the demo, and they don't really care about its ease, speed and reliability to an expert user because the expert user has already bought it. In our example, it is much faster to type equations and to do formatting with embedded commands because you use your fingers independently rather than your hand and because your fingers don't leave the keyboard. However, click-on menus, although they are slow and cumbersome when typing, look easy to use in the shop.

A timetable

I strongly recommend sitting down with the adviser and making up a timetable for writing it: a list of dates for when you will give the first and second drafts of each chapter to your adviser(s). This structures your time and provides intermediate targets. If you merely aim "to have the whole thing done by [some distant date]", you can deceive yourself and procrastinate more easily. If you have told your adviser that you will deliver a first draft of chapter 3 on Wednesday, it focuses your attention.

You may want to make your timetable into a chart with items that you can check off as you have finished them. This is particularly useful towards the end of the thesis when you find there will be quite a few loose ends here and there.

Iterative solution

Whenever you sit down to write, it is very important to write something. So write something, even if it is just a set of notes or a few paragraphs of text that you would never show to anyone else. It would be nice if clear, precise prose leapt easily from the keyboard, but it usually does not. Most of us find it easier, however, to improve something that is already written than to produce text from nothing. So put down a draft (as rough as you like) for your own purposes, then clean it up for your adviser to read. Word-processors are wonderful in this regard: in the first draft you do not have to start at the beginning, you can leave gaps, you can put in little notes to yourself, and then you can clean it all up later.

Your adviser will expect to read each chapter in draft form. S/he will then return it to you with suggestions and comments. Do not be upset if a chapter---especially the first one you write--- returns covered in red ink (or its electronic equivalent). Your adviser will want your thesis to be as good as possible, because his/her reputation as well as yours is affected. Scientific writing is a difficult art, and it takes a while to learn. As a consequence, there will be many ways in which your first draft can be improved. So take a positive attitude to all the scribbles with which your adviser decorates your text: each comment tells you a way in which you can make your thesis better.

As you write your thesis, your scientific writing is almost certain to improve. Even for native speakers of English who write very well in other styles, one notices an enormous improvement in the first drafts from the first to the last chapter written. The process of writing the thesis is like a course in scientific writing, and in that sense each chapter is like an assignment in which you are taught, but not assessed. Remember, only the final draft is assessed: the more comments your adviser adds to first or second draft, the better.

Before you submit a draft to your adviser, run a spell check so that s/he does not waste time on those. If you have any characteristic grammatical failings, check for them.

What is a thesis? For whom is it written? How should it be written?

Your thesis is a research report. The report concerns a problem or series of problems in your area of research and it should describe what was known about it previously, what you did towards solving it, what you think your results mean, and where or how further progress in the field can be made. Do not carry over your ideas from undergraduate assessment: a thesis is not an answer to an assignment question. One important difference is this: the reader of an assignment is usually the one who has set it. S/he already knows the answer (or one of the answers), not to mention the background, the literature, the assumptions and theories and the strengths and weaknesses of them. The readers of a thesis do not know what the "answer" is. If the thesis is for a PhD, the university requires that it make an original contribution to human knowledge: your research must discover something hitherto unknown.
Obviously your examiners will read the thesis. They will be experts in the general field of your thesis but, on the exact topic of your thesis, you are the world expert. Keep this in mind: you should write to make the topic clear to a reader who has not spent most of the last three years thinking about it.

Your thesis will also be used as a scientific report and consulted by future workers in your laboratory who will want to know, in detail, what you did. Theses are occasionally consulted by people from other institutions, and the library sends microfilm versions if requested (yes, still). More commonly theses are now stored in an entirely digital form. These may be stored as .pdf files on a server at your university. The advantage is that your thesis can be consulted much more easily by researchers around the world. (See e.g. Australian digital thesis project for the digital availability of research theses.) Write with these possibilities in mind.

It is often helpful to have someone other than your adviser(s) read some sections of the thesis, particularly the introduction and conclusion chapters. It may also be appropriate to ask other members of staff to read some sections of the thesis which they may find relevant or of interest, as they may be able to make valuable contributions. In either case, only give them revised versions, so that they do not waste time correcting your grammar, spelling, poor construction or presentation.

How much detail?

The short answer is: rather more than for a scientific paper. Once your thesis has been assessed and your friends have read the first three pages, the only further readers are likely to be people who are seriously doing research in just that area. For example, a future research student might be pursuing the same research and be interested to find out exactly what you did. ("Why doesn't the widget that Bloggs built for her project work any more? Where's the circuit diagram? I'll look up her thesis." "Blow's subroutine doesn't converge in my parameter space! I'll have to look up his thesis." "How did that group in Sydney manage to get that technique to work? I'll order a microfilm of that thesis they cited in their paper.") For important parts of apparatus, you should include workshop drawings, circuit diagrams and computer programs, usually as appendices. (By the way, the intelligible annotation of programs is about as frequent as porcine aviation, but it is far more desirable. You wrote that line of code for a reason: at the end of the line explain what the reason is.) You have probably read the theses of previous students in the lab where you are now working, so you probably know the advantages of a clearly explained, explicit thesis and/or the disadvantages of a vague one.

Make it clear what is yours

If you use a result, observation or generalisation that is not your own, you must usually state where in the scientific literature that result is reported. The only exceptions are cases where every researcher in the field already knows it: dynamics equations need not be followed by a citation of Newton, circuit analysis does not need a reference to Kirchoff. The importance of this practice in science is that it allows the reader to verify your starting position. Physics in particular is said to be a vertical science: results are built upon results which in turn are built upon results etc. Good referencing allows us to check the foundations of your additions to the structure of knowledge in the discipline, or at least to trace them back to a level which we judge to be reliable. Good referencing also tells the reader which parts of the thesis are descriptions of previous knowledge and which parts are your additions to that knowledge. In a thesis, written for the general reader who has little familiarity with the literature of the field, this should be especially clear. It may seem tempting to leave out a reference in the hope that a reader will think that a nice idea or an nice bit of analysis is yours. I advise against this gamble. The reader will probably think: "What a nice idea---I wonder if it's original?". The reader can probably find out via the net or the library.

If you are writing in the passive voice, you must be more careful about attribution than if you are writing in the active voice. "The sample was prepared by heating yttrium..." does not make it clear whether you did this or whether Acme Yttrium did it. "I prepared the sample..." is clear.

Style

The text must be clear. Good grammar and thoughtful writing will make the thesis easier to read. Scientific writing has to be a little formal---more formal than this text. Native English speakers should remember that scientific English is an international language. Slang and informal writing will be harder for a non-native speaker to understand.

Short, simple phrases and words are often better than long ones. Some politicians use "at this point in time" instead of "now" precisely because it takes longer to convey the same meaning. They do not care about elegance or efficient communication. You should. On the other hand, there will be times when you need a complicated sentence because the idea is complicated. If your primary statement requires several qualifications, each of these may need a subordinate clause: "When [qualification], and where [proviso], and if [condition] then [statement]". Some lengthy technical words will also be necessary in many theses, particularly in fields like biochemistry. Do not sacrifice accuracy for the sake of brevity. "Black is white" is simple and catchy. An advertising copy writer would love it. "Objects of very different albedo may be illuminated differently so as to produce similar reflected spectra" is longer and uses less common words, but, compared to the former example, it has the advantage of being true. The longer example would be fine in a physics thesis because English speaking physicists will not have trouble with the words. (A physicist who did not know all of those words would probably be glad to remedy the lacuna either from the context or by consulting a dictionary.)

Sometimes it is easier to present information and arguments as a series of numbered points, rather than as one or more long and awkward paragraphs. A list of points is usually easier to write. You should be careful not to use this presentation too much: your thesis must be a connected, convincing argument, not just a list of facts and observations.

One important stylistic choice is between the active voice and passive voice. The active voice ("I measured the frequency...") is simpler, and it makes clear what you did and what was done by others. The passive voice ("The frequency was measured...") makes it easier to write ungrammatical or awkward sentences. If you use the passive voice, be especially wary of dangling participles. For example, the sentence "After considering all of these possible materials, plutonium was selected" implicitly attributes consciousness to plutonium. This choice is a question of taste: I prefer the active because it is clearer, more logical and makes attribution simple. The only arguments I have ever heard for avoiding the active voice in a thesis are (i) many theses are written in the passive voice, and (ii) some very polite people find the use of "I" immodest. Use the first person singular, not plural, when reporting work that you did yourself: the editorial 'we' may suggest that you had help beyond that listed in your acknowledgments, or it may suggest that you are trying to share any blame. On the other hand, retain plural verbs for "data": "data" is the plural of "datum", and lots of scientists like to preserve the distinction. Just say to yourself "one datum is ..", "these data are.." several times. An excellent and widely used reference for English grammar and style is A Dictionary of Modern English Usage by H.W. Fowler.

Presentation

There is no need for a thesis to be a masterpiece of desk-top publishing. Your time can be more productively spent improving the content than the appearance.

In many cases, a reasonably neat diagram can be drawn by hand faster than with a graphics package, and you can scan it if you want an electronic version. Either is usually satisfactory. A one bit (i.e. black and white), moderate resolution scan of a hand-drawn sketch will be bigger than a line drawing generated on a graphics package, but not huge. While talking about the size of files, we should mention that photographs look pretty but take up a lot of memory. There's another important difference, too. The photographer thought about the camera angle and the focus etc. The person who drew the schematic diagram thought about what components ought to be depicted and the way in which the components of the system interacted with each other. So the numerically small information content of the line drawing may be much more useful information than that in a photograph.

Another note about figures and photographs. In the digital version of your thesis, do not save ordinary photographs or other illustrations as bitmaps, because these take up a lot of memory and are therefore very slow to transfer. Nearly all graphics packages allow you to save in compressed format as .jpg (for photos) or .gif (for diagrams) files. Further, you can save space/speed things up by reducing the number of colours. In vector graphics (as used for drawings), compression is usually unnecessary.

In general, students spend too much time on diagrams---time that could have been spent on examining the arguments, making the explanations clearer, thinking more about the significance and checking for errors in the algebra. The reason, of course, is that drawing is easier than thinking.

I do not think that there is a strong correlation (either way) between length and quality. There is no need to leave big gaps to make the thesis thicker. Readers will not appreciate large amounts of vague or unnecessary text.

Approaching the end

A deadline is very useful in some ways. You must hand in the thesis, even if you think that you need one more draft of that chapter, or someone else's comments on this section, or some other refinement. If you do not have a deadline, or if you are thinking about postponing it, please take note of this: A thesis is a very large work. It cannot be made perfect in a finite time. There will inevitably be things in it that you could have done better. There will be inevitably be some typos. Indeed, by some law related to Murphy's, you will discover one when you first flip open the bound copy. No matter how much you reflect and how many times you proof read it, there will be some things that could be improved. There is no point hoping that the examiners will not notice: many examiners feel obliged to find some examples of improvements (if not outright errors) just to show how thoroughly they have read it. So set yourself a deadline and stick to it. Make it as good as you can in that time, and then hand it in! (In retrospect, there was an advantage in writing a thesis in the days before word processors, spelling checkers and typing programs. Students often paid a typist to produce the final draft and could only afford to do that once.)
How many copies?

Talk to your adviser about this. As well as those for the examiners, the university libraries and yourself, you should make some distribution copies. These copies should be sent to other researchers who are working in your field so that:
they can discover what marvellous work you have been doing before it appears in journals;
they can look up the fine details of methods and results that will or have been published more briefly elsewhere;
they can realise what an excellent researcher you are. This realisation could be useful if a post- doctoral position were available in their labs. soon after your submission, or if they were reviewers of your research/post-doctoral proposal. Even having your name in their bookcases might be an advantage.
Whatever the University's policy on single or double-sided copies, the distribution copies could be double-sided paper, or digital, so that forests and postage accounts are not excessively depleted by the exercise. Your adviser could help you to make up a list of interested and/or potentially useful people for such a mailing list. Your adviser might also help by funding the copies and postage if they are not covered by your scholarship. A CD with your thesis will be cheaper than a paper copy. You don't have to burn them all yourself: companies make multiple copies for several dollars a copy.

The following comment comes from Marilyn Ball of the Australian National University in Canberra: "When I finished writing my thesis, a postdoc wisely told me to give a copy to my parents. I would never have thought of doing that as I just couldn't imagine what they would do with it. I'm very glad to have taken that advice as my parents really appreciated receiving a copy and proudly displayed it for years. (My mother never finished high school and my father worked with trucks - he fixed 'em, built 'em, drove 'em, sold 'em and junked 'em. Nevertheless, they enjoyed having a copy of my thesis.)"

Personal

In the ideal situation, you will be able to spend a large part---perhaps a majority---of your time writing your thesis. This may be bad for your physical and mental health.

Typing
Set up your chair and computer properly. The Health Service, professional keyboard users or perhaps even the school safety officer will be able to supply charts showing recommended relative heights, healthy postures and also exercises that you should do if you spend a lot of time at the keyboard. These last are worthwhile insurance: you do not want the extra hassle of back or neck pain. Try to intersperse long sessions of typing with other tasks, such as reading, drawing, calculating, thinking or doing research.
If you do not touch type, you should learn to do so for the sake of your neck as well as for productivity. There are several good software packages that teach touch typing interactively. If you use one for say 30 minutes a day for a couple of weeks, you will be able to touch type. By the time you finish the thesis, you will be able to touch type quickly and accurately and your six hour investment will have paid for itself. Be careful not to use the typing exercises as a displacement activity.

Exercise
Do not give up exercise for the interim. Lack of exercise makes you feel bad, and you do not need anything else making you feel bad while writing a thesis. 30-60 minutes of exercise per day is probably not time lost from your thesis: I find that if I do not get regular exercise, I sleep less soundly and longer. How about walking to work and home again? (Walk part of the way if your home is distant.) Many people opine that a walk helps them think, or clears the head. You may find that an occasional stroll improves your productivity.
Food
Do not forget to eat, and make an effort to eat healthy food. You should not lose fitness or risk illness at this critical time. Exercise is good for keeping you appetite at a healthy level. I know that you have little time for cooking, but keep a supply of fresh fruit, vegetables and bread. It takes less time to make a sandwich than to go to the local fast food outlet, and you will feel better afterwards.
Drugs
Thesis writers have a long tradition of using coffee as a stimulant and alcohol or marijuana as relaxants. (Use of alcohol and coffee is legal, use of marijuana is not.) Used in moderation, they do not seem to have ill effects on the quality of thesis produced. Excesses, however, are obviously counter-productive: several espressi and you will be buzzing too much to sit down and work; several drinks at night will slow you down next day.
Others
Other people will be sympathetic, but do not take them for granted. Spouses, lovers, family and friends should not be undervalued. Spend some time with them and, when you do, have a good time. Do not spend your time together complaining about your thesis: they already resent the thesis because it is keeping you away from them. If you can find another student writing a thesis, then you may find it therapeutic to complain to each other about advisers and difficulties. S/he need not be in the same discipline as you are.
Coda

Keep going---you're nearly there! Most PhDs will admit that there were times when we thought about reasons for not finishing. But it would be crazy to give up at the writing stage, after years of work on the research, and it would be something to regret for a long time.

Writing a thesis is tough work. One anonymous post doctoral researcher told me: "You should tell everyone that it's going to be unpleasant, that it will mess up their lives, that they will have to give up their friends and their social lives for a while. It's a tough period for almost every student." She's right: it is certainly hard work, it will probably be stressful and you will have to adapt your rhythm to it. It is also an important rite of passage and the satisfaction you will feel afterwards is wonderful. On behalf of scholars everywhere, I wish you good luck!

A suggested thesis structure

The list of contents and chapter headings below is appropriate for some theses. In some cases, one or two of them may be irrelevant. Results and Discussion are usually combined in several chapters of a thesis. Think about the plan of chapters and decide what is best to report your work. Then make a list, in point form, of what will go in each chapter. Try to make this rather detailed, so that you end up with a list of points that corresponds to subsections or even to the paragraphs of your thesis. At this stage, think hard about the logic of the presentation: within chapters, it is often possible to present the ideas in different order, and not all arrangements will be equally easy to follow. If you make a plan of each chapter and section before you sit down to write, the result will probably be clearer and easier to read. It will also be easier to write.

Copyright waiver
Your institution may have a form for this (UNSW does). In any case, this standard page gives the university library the right to publish the work, possibly by microfilm or other medium. (At UNSW, the Postgraduate Student Office will give you a thesis pack with various guide-lines and rules about thesis format. Make sure that you consult that for its formal requirements, as well as this rather informal guide.)
Declaration
Check the wording required by your institution, and whether there is a standard form. Many universities require something like: "I hereby declare that this submission is my own work and that, to the best of my knowledge and belief, it contains no material previously published or written by another person nor material which to a substantial extent has been accepted for the award of any other degree or diploma of the university or other institute of higher learning, except where due acknowledgment has been made in the text. (signature/name/date)"
Title page
This may vary among institutions, but as an example: Title/author/"A thesis submitted for the degree of Doctor of Philosophy in the Faculty of Science/The University of New South Wales"/date.
Abstract
Of all your thesis, this part will be the most widely published and most read because it will be published in Dissertation Abstracts International. It is best written towards the end, but not at the very last minute because you will probably need several drafts. It should be a distillation of the thesis: a concise description of the problem(s) addressed, your method of solving it/them, your results and conclusions. An abstract must be self-contained. Usually they do not contain references. When a reference is necessary, its details should be included in the text of the abstract. Check the word limit. Remember: even though it appears at the beginning, an abstract is not an introduction. It is a résumé of your thesis.
Acknowledgments
Most thesis authors put in a page of thanks to those who have helped them in matters scientific, and also indirectly by providing such essentials as food, education, genes, money, help, advice, friendship etc. If any of your work is collaborative, you should make it quite clear who did which sections.
Table of contents
The introduction starts on page 1, the earlier pages should have roman numerals. It helps to have the subheadings of each chapter, as well as the chapter titles. Remember that the thesis may be used as a reference in the lab, so it helps to be able to find things easily.
Introduction
What is the topic and why is it important? State the problem(s) as simply as you can. Remember that you have been working on this project for a few years, so you will be very close to it. Try to step back mentally and take a broader view of the problem. How does it fit into the broader world of your discipline?
Especially in the introduction, do not overestimate the reader's familiarity with your topic. You are writing for researchers in the general area, but not all of them need be specialists in your particular topic. It may help to imagine such a person---think of some researcher whom you might have met at a conference for your subject, but who was working in a different area. S/he is intelligent, has the same general background, but knows little of the literature or tricks that apply to your particular topic.

The introduction should be interesting. If you bore the reader here, then you are unlikely to revive his/her interest in the materials and methods section. For the first paragraph or two, tradition permits prose that is less dry than the scientific norm. If want to wax lyrical about your topic, here is the place to do it. Try to make the reader want to read the heavy bundle that has arrived uninvited on his/her desk. Go to the library and read several thesis introductions. Did any make you want to read on? Which ones were boring?

This section might go through several drafts to make it read well and logically, while keeping it short. For this section, I think that it is a good idea to ask someone who is not a specialist to read it and to comment. Is it an adequate introduction? Is it easy to follow? There is an argument for writing this section---or least making a major revision of it---towards the end of the thesis writing. Your introduction should tell where the thesis is going, and this may become clearer during the writing.

Literature review
Where did the problem come from? What is already known about this problem? What other methods have been tried to solve it?
Ideally, you will already have much of the hard work done, if you have been keeping up with the literature as you vowed to do three years ago, and if you have made notes about important papers over the years. If you have summarised those papers, then you have some good starting points for the review.

If you didn't keep your literature notes up to date, you can still do something useful: pass on the following advice to any beginning PhD students in your lab and tell them how useful this would have been to you. When you start reading about a topic, you should open a spread sheet file, or at least a word processor file, for your literature review. Of course you write down the title, authors, year, volume and pages. But you also write a summary (anything from a couple of sentences to a couple of pages, depending on the relevance). In other columns of the spread sheet, you can add key words (your own and theirs) and comments about its importance, relevance to you and its quality.

How many papers? How relevant do they have to be before you include them? Well, that is a matter of judgement. On the order of a hundred is reasonable, but it will depend on the field. You are the world expert on the (narrow) topic of your thesis: you must demonstrate this.

A political point: make sure that you do not omit relevant papers by researchers who are like to be your examiners, or by potential employers to whom you might be sending the thesis in the next year or two.

Middle chapters

In some theses, the middle chapters are the journal articles of which the student was major author. There are several disadvantages to this format.
One is that a thesis is both allowed and expected to have more detail than a journal article. For journal articles, one usually has to reduce the number of figures. In many cases, all of the interesting and relevant data can go in the thesis, and not just those which appeared in the journal. The degree of experimental detail is usually greater in a thesis. Relatively often a researcher requests a thesis in order to obtain more detail about how a study was performed.

Another disadvantage is that your journal articles may have some common material in the introduction and the "Materials and Methods" sections.

The exact structure in the middle chapters will vary among theses. In some theses, it is necessary to establish some theory, to describe the experimental techniques, then to report what was done on several different problems or different stages of the problem, and then finally to present a model or a new theory based on the new work. For such a thesis, the chapter headings might be: Theory, Materials and Methods, {first problem}, {second problem}, {third problem}, {proposed theory/model} and then the conclusion chapter. For other theses, it might be appropriate to discuss different techniques in different chapters, rather than to have a single Materials and Methods chapter.

Here follow some comments on the elements Materials and Methods, Theory, Results and discussion which may or may not correspond to thesis chapters.

Materials and Methods
This varies enormously from thesis to thesis, and may be absent in theoretical theses. It should be possible for a competent researcher to reproduce exactly what you have done by following your description. There is a good chance that this test will be applied: sometime after you have left, another researcher will want to do a similar experiment either with your gear, or on a new set-up in a foreign country. Please write for the benefit of that researcher.
In some theses, particularly multi-disciplinary or developmental ones, there may be more than one such chapter. In this case, the different disciplines should be indicated in the chapter titles.

Theory
When you are reporting theoretical work that is not original, you will usually need to include sufficient material to allow the reader to understand the arguments used and their physical bases. Sometimes you will be able to present the theory ab initio, but you should not reproduce two pages of algebra that the reader could find in a standard text. Do not include theory that you are not going to relate to the work you have done.
When writing this section, concentrate at least as much on the physical arguments as on the equations. What do the equations mean? What are the important cases?

When you are reporting your own theoretical work, you must include rather more detail, but you should consider moving lengthy derivations to appendices. Think too about the order and style of presentation: the order in which you did the work may not be the clearest presentation.

Suspense is not necessary in reporting science: you should tell the reader where you are going before you start.

Results and discussion
The results and discussion are very often combined in theses. This is sensible because of the length of a thesis: you may have several chapters of results and, if you wait till they are all presented before you begin discussion, the reader may have difficulty remembering what you are talking about. The division of Results and Discussion material into chapters is usually best done according to subject matter.
Make sure that you have described the conditions which obtained for each set of results. What was held constant? What were the other relevant parameters? Make sure too that you have used appropriate statistical analyses. Where applicable, show measurement errors and standard errors on the graphs. Use appropriate statistical tests.

Take care plotting graphs. The origin and intercepts are often important so, unless the ranges of your data make it impractical, the zeros of one or both scales should usually appear on the graph. You should show error bars on the data, unless the errors are very small. For single measurements, the bars should be your best estimate of the experimental errors in each coordinate. For multiple measurements these should include the standard error in the data. The errors in different data are often different, so, where this is the case, regressions and fits should be weighted (i.e. they should minimize the sum of squares of the differences weighted inversely as the size of the errors.) (A common failing in many simple software packages that draw graphs and do regressions is that they do not treat errors adequately. UNSW student Mike Johnston has written a plotting routine that plots data with error bars and performs weighted least square regressions. It is at http://www.phys.unsw.edu.au/3rdyearlab/graphing/graph.html). You can just 'paste' your data into the input and it generates a .ps file of the graph.

In most cases, your results need discussion. What do they mean? How do they fit into the existing body of knowledge? Are they consistent with current theories? Do they give new insights? Do they suggest new theories or mechanisms?

Try to distance yourself from your usual perspective and look at your work. Do not just ask yourself what it means in terms of the orthodoxy of your own research group, but also how other people in the field might see it. Does it have any implications that do not relate to the questions that you set out to answer?

Final chapter, references and appendices

Conclusions and suggestions for further work
Your abstract should include your conclusions in very brief form, because it must also include some other material. A summary of conclusions is usually longer than the final section of the abstract, and you have the space to be more explicit and more careful with qualifications. You might find it helpful to put your conclusions in point form.
It is often the case with scientific investigations that more questions than answers are produced. Does your work suggest any interesting further avenues? Are there ways in which your work could be improved by future workers? What are the practical implications of your work?

This chapter should usually be reasonably short---a few pages perhaps. As with the introduction, I think that it is a good idea to ask someone who is not a specialist to read this section and to comment.

References (See also under literature review)
It is tempting to omit the titles of the articles cited, and the university allows this, but think of all the times when you have seen a reference in a paper and gone to look it up only to find that it was not helpful after all.
Should you reference web sites and, if so, how? If you cite a journal article or book, the reader can go to a library and check that the cited document and check whether or not it says what you say it did. A web site may disappear, and it may have been updated or changed completely. So references to the web are usually less satisfactory. Nevertheless, there are some very useful and authoritative sources. So, if the rules of your institution permit it, it may be appropriate to cite web sites. (Be cautious, and don't overuse such citations. In particular, don't use a web citation where you could reasonably use a "hard" citation. Remember that your examiners are likely to be older and more conservative.) You should give the URL and also the date you downloaded it. If there is a date on the site itself (last updated on .....) you should included that, too.

Appendices
If there is material that should be in the thesis but which would break up the flow or bore the reader unbearably, include it as an appendix. Some things which are typically included in appendices are: important and original computer programs, data files that are too large to be represented simply in the results chapters, pictures or diagrams of results which are not important enough to keep in the main text.
Some sites with related material
Writing and publishing a scientific paper
How to survive a thesis defence
Research resources and links supplied by Deakin University
"Final year projects": a guide from Mike Hart at King Alfred's College, Winchester, UK
Postgraduate Student Resources supplied by University of Canberra
A useful aid to surviving meetings with management
The National Association of Graduate - Professional Students (USA)
Some relevant texts
Stevens, K. and Asmar, C (1999) 'Doing postgraduate research in Australia'. Melbourne University Press, Melbourne ISBN 0 522 84880 X.
Phillips, E.M and Pugh, D.S. (1994) 'How to get a PhD : a handbook for students and their supervisors'. Open University Press, Buckingham, England
Tufte, E.R. (1983) 'The visual display of quantitative information'. Graphics Press, Cheshire, Conn.
Tufte, E.R. (1990) 'Envisioning information' Graphics Press, Cheshire, Conn.
Distribution
If you have found these documents useful, please feel free to pass the address or a hard copy to any other thesis writers or graduate student organisations. Please do not sell them, or use any of the contents without acknowledgement.

Suggestions, thanks and caveats

This document will be updated occasionally. If you have suggestions for inclusions, amendments or other improvements, please send them. Do so after you have submitted the thesis---do not use this invitation as a displacement activity. I thank Marilyn Ball, Gary Bryant, Bill Whiten and J. Douglas, whose suggestions have been incorporated in this version. Substantial contributions will be acknowledged in future versions. I also take this opportunity to thank my own thesis advisers, Stjepan Marcelja and Jacob Israelachvili, for their help and friendship, and to thank the graduate students to whom I have had the pleasure to be an adviser, a colleague and a friend. Opinions expressed in these notes are mine and do not necessarily reflect the policy of the University of New South Wales or of the School of Physics.

A FAQ and some observations about the web

Why and how did I write this document? The need for it was evident so, as one of my PhD students approached the end of his project, I made notes of everything that I said to him about thesis writing. These notes became the plan for the first draft of this document, which has been extended several times since then. I am surprised that it has hundreds of readers each day. However, this is an important message about the web. It takes time and thought to make a good resource but, if you do, it can benefit a lot of people. When this document was first posted, the web was relatively new and feedback showed that people were often surprised to find what they sought. Now there is a tendency to take the web for granted: one is almost disappointed not to find what one is seeking. However, the web is only as good as the collective effort of all of us. The readers of this document will be scholars, experts and educators: among the many contributions you will make to knowledge and your communities, there may be contributions that should be made freely available, all over the world. Keep this observation about the web in the back of your mind for later, when you are not writing a thesis.

© 1996. Modified 2/11/06 Joe Wolfe / J.Wolfe@unsw.edu.au, phone 61- 2-9385 4954 (UT + 10, +11 Oct-Mar).

Mixed Method and Mixed Model Research

Mixed research is research in which quantitative and qualitative techniques are mixed in a single study. It is the third major research paradigm, adding an attractive alternative (when it is appropriate) to quantitative and qualitative research.

Proponents of mixed research typically adhere to the compatibility thesis as well as to the philosophy of pragmatism.
· The compatibility thesis is the idea that quantitative and qualitative methods are compatible, that is, they can both be used in a single research study.
· The philosophy of pragmatism says that researchers should use the approach or mixture of approaches that works the best in a real world situation. In short, what works is what is useful and should be used, regardless of any philosophical assumptions, paradigmatic assumptions, or any other type of assumptions. (Pragmatism was started by the great American philosophers Charles Sanders Peirce, William James, and John Dewey).

Today, proponents of mixed research attempt to use what is called the fundamental principle of mixed research.
· According to this fundamental principle, the researcher should use a mixture or combination of methods that has complementary strengths and nonoverlapping weaknesses.
· To aid you in applying this fundamental principle, we have provided tables that show the strengths and weaknesses of quantitative research and qualitative research. Here they are for your convenience:

[Tables omitted]


Here is a list of the strengths and weaknesses of mixed research. Looking at the strengths, you will see where you want to go in planning a mixed research study.

The Research Continuum
Research can be viewed as falling along a research continuum with “monomethod” research placed on the far left side, “fully mixed” research placed on the far right side, and “partially mixed” located in the center. You should be able to take any given research study and place it somewhere on the continuum.

[Table omitted]


Types of Mixed Research Methods
There are two major types of mixed research: they are mixed model research and mixed method research.

Mixed Model Research
In mixed model research quantitative and qualitative approaches are mixed within or across the stages of the research process.

Here are the two mixed model research subtypes: within-stage and across-stage mixed model research.
1. In within-stage mixed model research, quantitative and qualitative approaches are mixed within one or more of the stages of research.
· An example of within-stage mixed model research would be where you used a questionnaire during data collection that included both open-ended (i.e., qualitative) questions and closed-ended (i.e., quantitative) questions.

2. In across-stage mixed model research, quantitative and qualitative
approaches are mixed across at least two of the stages of research. Across-stage mixed model research designs are easily seen by examining designs 2 through 7 in Figure 14.2 (shown below):

[Figure omitted]

· Here is an example of across-stage mixed model research: A researcher wants to explore (qualitative objective) why people take on-line college courses. The researcher conducts open-ended interviews (qualitative data collection) asking them why they take on-line courses, and then the researcher quantifies the results by counting the number of times each type of response occurs (quantitative data analysis); the researcher also reports the responses as percentages and examines the relationships between sets of categories or variables through the use of contingency tables. Note that this is design 2 shown above in Figure 14.2.

Mixed Method Research
In mixed method research, a qualitative phase and a quantitative phase are included in the overall research study. It’s like including a quantitative mini-study and a qualitative mini-study in one overall research study.

Mixed method research designs are classified according to two major dimensions:
1. Time order (i.e., concurrent versus sequential) and
2. Paradigm emphasis (i.e., equal status versus dominant status).

Below, in Figure 14.3, you can see the specific mixed method designs that result from crossing time order and paradigm emphasis. It is a 2-by-2 matrix, and it includes nine specific mixed method designs. In order to understand the designs, you need to first understand the notation that is used.
· QUAL and qual both stand for qualitative research.
· QUAN and quan both stand for quantitative research.
· Capital letters denote priority or increased weight.
· Lowercase letters denote lower priority or weight.
· A plus sign (+) indicated the concurrent collection of data.
· An arrow (→) represents a sequential collection of data.
· For example: qual→QUAN is a dominant status, sequential design where, the overall study is primarily quantitative but it is preceded by a qualitative phase. Perhaps a researcher does an open-ended survey to find some important categories or variables that students say are important reasons for dropping out of on-line courses. Then in the quantitative phase the researcher does a quantitative study of predictors of dropping out, using quantitative statistical methods. In other words the quantitative phase was primary and the qualitative phase was supportive (and occurred first).

· In order to use Figure 14.3, you need to ask yourself two questions:
1. Do you want to operate largely within one dominant paradigm or not (i.e., do you want to use a dominant status design or an equal status design?), and
2. Do you want to conduct the phases concurrently (i.e., at roughly the same time) or sequentially (i.e., one before the other)?
· Your answers to these two questions will lead you to one of the designs in Figure 14.3.
· Your goal is to pragmatically design a study that fits your particular needs and circumstances.

[Figure omitted]


· It is important to understand that you are not limited to the mixed method or mixed model designs provided in this chapter. Our designs are provided to get you started. You should feel free to mix and match the designs into a design that best fits your needs. This includes designing studies that are a mix of mixed model and mixed method designs. You goal, always, is to answer you research question(s) and then to design a study that will help you to do that.


Stages of Mixed Research Process
There are eight stages in the mixed research process, as shown in Figure 14.4 (in the text, and here for your convenience).

[Figure omitted]


It is important to note that although the steps in mixed research are numbered, researchers often follow these steps in different orders, depending on what particular needs and concerns arise or emerge during a particular research study.
· For example, interpretation and validation of the data should be done throughout the data collection process.

I will very briefly comment on each of the eight (nonlinear) steps:

(1) Determine whether a mixed design is appropriate
· Do you believe that you can best answer your research question(s) through the use of mixed research?
· Do you believe that mixed research will offer you the best design for the amount and kind of evidence that you hope to obtain as you conduct your research study?


(2) Determine the rationale for using a mixed design
· The five most important rationales or purposes for mixed research are shown below in Table 14.4:

[Table omitted]


· You can see in Table 14.4, that mixed research can help researchers to a lot of important things as they attempt to understand the world.


(3) Select the mixed method or mixed model research design
· We have already shown you, in this lecture, the basic mixed model designs and the basic mixed method designs.
· Remember that you can also build more unique and/or more complex designs than the ones we have shown as you plan a study that will help you to answer your research question(s).


(4) Collect the data
· Keep in mind the six major methods of data collection that we discussed in chapter 6: tests, questionnaires, interviews, focus groups, observation, and secondary or already existing data (such as personal and official documents, physical data, and archived research data).

(5) Analyze the data
· You can use the quantitative data analysis techniques (Chapters 15 and 16) and qualitative data analysis techniques (Chapter 17).
· You might want to use the technique of quantitizing (i.e., converting qualitative data into quantitative data).
· You might want to use the technique of qualitizing (i.e., converting quantitative data into qualitative data).
· For more information on data analysis in mixed research, I highly recommend the following:

Onwuegbuzie, A.J., & Teddlie, C. (2003). A framework for analzing data in mixed methods research. In A. Tashakkori & C. Teddlie (Eds.), Handbook of mixed methods in social and behavioral research (pp. 351-383). Thousand Oaks, CA: Sage.

(6) Validate the data
· Data validation is something that should be done throughout your research study because if your data are not trustworthy then you study is not trustworthy. In Chapter 8 we discussed validity strategies used in quantitative research (pp. 228-248) and validity strategies used in qualitative research (pp. 249-256).
· You should consider using quantitative and qualitative validity strategies in your study, and you should mix these in a way that best works for your mixed research study.


(7) Interpret the data
· Data interpretation begins as soon as you enter the field or collect the first datum (datum is the singular of data), and data interpretation continues throughout your research study.
· Remember that data interpretation and data validation go hand-in-hand; that is, you want to make sure that you continually use strategies that will provide valid data and help you to make defensible interpretations of your data.
· A couple of strategies to use during data interpretation are reflexivity (i.e., which involves self-awareness and critical self-reflection by the researcher on his or her potential biases and predispositions as these may affect the research process and conclusions), and negative-case sampling (i.e., attempting to locate and examine cases that disconfirm your expectations and tentative explanations).


(8) Write the research report.
Writing the report also can be started during data collection rather than waiting until the end.
Remember that mixing MUST take place somewhere in mixed research if it is to truly be mixed research, and your report should also reflect mixing; that is, as you discuss your results you must relate the quantitative and qualitative parts of your research study to make sense of the overall study and to capitalize on the strengths of mixed research.

In conclusion, mixed research is the newest research paradigm in educational research. It offers much promise, and we expect to see much more methodological work and discussion about mixed research in the future as more researchers and book authors become aware of this important approach to empirical research.