How to write a thesis?

This guide to writing a thesis gives some simple and practical advice: how to start, how to organize, how to divide the enormous task into pieces less frightening and how to attack these pieces. It also gives some clues practices to survive the ordeal. A structure is suggested and a guide to what we should put in each section is proposed. This guide was written primarily for doctoral students in physics, and most of the specific examples given are from this discipline. Nevertheless, the return of readers said it had been used and appreciated by students in various fields: in letters as in science.

Getting Started

At the start, writing a thesis seems to be a long and difficult task. This is because it is a long and difficult task. Fortunately, it will seem less daunting after one or two chapters. Towards the end, you'll even find fun --- a fun based on the fulfillment and satisfaction in improving your scientific writing, and of course, that you approach the end. Like many jobs, writing a thesis seems more difficult before it began. Come take the first step.


First, do a thesis outline: several pages containing chapter titles, headings, titles of figures and tables (to indicate what results will arise and where) and perhaps other notes and comments you come to mind. You will find a section devoted to the structure of a thesis at the end of this text. Once you have a list of chapters, and under each heading, a reasonably comprehensive list of things to relate or explain, you've already broken one of the most difficult barriers. When you sit down to type, your goal is more intimidating to a thesis --- --- but something simpler. Your new goal is simply a paragraph or section about one of your topics. It is easiest to start with an easy section: this puts you in the bath and gives you confidence. In a scientific thesis, the chapter easier to write is often that the materials and methods --- this is simply to note --- carefully, formally and in a logical order --- what you've done . How To Plan chapter? You could try the method I use for a scientific article, a method that I learned from my master thesis: assemble all the figures you use in the chapter and put them in the order you would use if you go explain to someone what they mean. You can also prepare as if you explain them orally to a colleague --- after all you probably will present several papers based on your thesis. Once you've found the most logical order, note the keywords for your explanation. These keywords provide a skeleton for your chapter plan.
Once you have established a plan, discuss it with your master thesis. This step is important: first, it will be helpful suggestions. Secondly, the meeting noted that it should expect to see happen chapters, which will make high priority demands on his time. Once you and your adviser agree on a logical structure, it should give him a copy of the plan to help navigate the flood of chapters as they arise in the disorder. If you have a second master thesis, discuss the plan with him too, so that you can make comments.


It is encouraging and helpful to start a filing system. Open in your word processing software, a different file for each chapter and for references. You can put notes in these files, as well as the formal text. In writing a paragraph in chapter X, you may come to think of referring to it in Chapter Y. So you write a note in it to remember it the right time. Or do you think of something interesting or within another chapter. Plus you have accumulated such notes, plus the drafting of the latter will be facilitated. Make a backup of these files and repeat these daily backups (depending on the reliability of your computer and the age of your hard drive). Never keep the backup drive near the computer in case the hypothetical thief who loves your computer will also think he needs some discs. It's a good idea to make a backup copy of geographically separated from your computer. For example, you could send a copy of each chapter, by email, a colleague at another institution. You can also send a copy to yourself and leave the server, but be careful not to exceed the memory limit given to you.
You should also have a filing system for your papers: a collection of files, one for each chapter. It will do you well psychologically (ah, I already started) and also solve the mess on your desk. Your files must not only contain the results and pages of calculations, but also all sorts of old notes, references, calibration curves, the addresses of suppliers, specifications, speculations, letters from colleagues, almost anything that seems relevant a chapter, put it in the workbook. When you wrote a chapter, put it all inside. Then put all the files in a box or a large workbook. Look at the size of this box from time to time --- ah, the thesis is taking shape!
If you have data that exist only on paper, copy them and keep the copy in a different place. Also makes a copy of your lab book. This is just for a second reason: after the thesis, the lab book will remain in the lab, but you could yourself be led to need it someday. In addition, scientific ethics requires that the lab keeps the original data for at least ten years, and it is more likely that a copy is found if there are two.
When you're organized, you should contact the academic bureaucracy. Members of the thesis committee must be appointed, etc. invited. The forms required by your department and university administration must be completed. Advice from a recently graduated PhD student will also be useful.

A schedule

I strongly recommend a meeting with your adviser to determine a schedule for your newsroom: a list of dates when you give him the first and second streams from each chapter. This planning structure your time and provide intermediate targets. If you intend "to have finished the whole thesis before" date so remote {}, you can cheat more easily. By cons, if you promised to your master thesis you give him a first draft of Chapter 3 Wednesday, attention will focus on the immediate task.

Iterative solution

Whenever you sit down to write, it is very important to write something. So write something, even if it is not pretty. It would be nice if the lucid and precise prose leapt from the keyboard without effort, but this is rarely the case. Instead, most of us find it easier to improve an existing text file as the text itself, ab initio. So be a first draft (as rough as it is) for you, and then buff it several times before giving it to your master thesis. The word processing software are excellent for that: in a first draft, you do not have to start early, you can leave gaps, you can put little notes to yourself, and then you can turn readable text and polish all later. Your adviser will expect to read each chapter in draft form. It will reward you with suggestions and comments. Do not be depressed if a chapter especially --- the first --- you give back covered in red ink. Your master thesis will want your thesis to be as good as possible because his reputation as well as yours will be in. The scientific writing is a difficult art, and it takes time to learn. Consequently, there will be many ways in which your first draft can be improved. Take a positive attitude all the doodles with which your adviser decorates your text: each comment gives you an opportunity to improve your document.
While writing your thesis, your scientific writing style will certainly improve. Even those who write well in other styles dramatically improve their scientific writing between the first draft of the chapter written in the first and last draft of the chapter written last. The process of writing a thesis is like a course on scientific writing, and in that sense each chapter is like an assignment in which you learn, without receiving a grade. Remember, this is the final and definitive account: the greater the number of comments that made your master thesis, the better. That said, do not give him the first draft!
Before submitting a text to your master thesis, do go through a spell-checking software so he can concentrate on the important points. If you have any failures grammatical characteristics, or a tendency to add phrases informal, pay attention to it before making your work.

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

Your thesis is a research report. This report concerns a problem or a series of problems in a specialty. It should describe what was known before, what you did to solve the (s) problem (s), what your results mean, the new issues raised by your research, and how these new problems can be solved. A thesis is very different from a duty to students: the reader of a duty is usually the one who asked the question. He already knows the answer (or answers), not to mention the original documentation, assumptions and theories and the strengths and weaknesses of them. Readers of a thesis do not know "response." In the case of a doctoral thesis, the university requires an original contribution to the current state of knowledge. Obviously, your thesis is aimed primarily at members of your jury. This will be experts from the scientific field in which you conducted your research. But the world expert on your subject, you, never forget. Your readers probably will not spend their last three or four years on this subject. You do not write for yourself, but for them, and your presentation must be sufficiently clear to enable them to understand.
Your thesis will also be used as a scientific report and will be accessed by future researchers in your laboratory and elsewhere who want to know in detail what you did. Theses are occasionally consulted by people belonging to other institutions, and the library sends copies from the microfilm version (yes, again) to those who request them. Increasingly, theses are stored in a fully digital form (figures as well as the text in digital form). One consequence of this is that your thesis can be easily accessed by researchers worldwide. Write this in mind.
It is often helpful to ask someone other than your adviser to read some sections of the thesis, particularly the introduction and conclusion. Also ask other teachers or researchers to read the sections of the thesis they can find relevant or of interest, because they may be able to make valuable contributions. In either case, do they submit the revised version, so they do not waste time correcting your grammar, style or presentation.

How much detail?

More than a scientific article. Once connected to your thesis and your friends have read the first three pages, it is likely that the only people likely to read your work in the future are those who conduct research on this topic. For example, a future research student may continue the same research and be interested to find out exactly what you did. ("Why does he has built over the thing Dupont? Where is the design of the circuit? I will look at his thesis." Durand's algorithm does not converge in my parameter space. I find his thesis "." How this group has managed to get this technique? I ask for a copy of the thesis they cited in their article. ") For important components of the equipment, you should include shop drawings, drawings circuit that you have made and the software code you have written, often in the form of annexes. (When there is, the intelligible annotation of code is rare, but desirable. You wrote this line of code for one reason: at the end of the line explain why.) You've already read the theses of students earlier in the lab where you are now, so you already know the benefits of a thesis that clearly explains and / or disadvantages of a wave theory.

Indicate clearly what is yours

If you use an outcome, an observation or a generalization that is not yours, you must specify where in the literature this result is reported. The only exceptions are cases where every researcher in the field already know: the dynamics equations can be written without mentioning Newton, analysis of the circuit can do without a reference to Kirchoff. The reason for this our obligation as a writer in science is that it allows the reader to check the foundation's scientific research. Physics, they say, is a vertical science: results are built from the results, which in turn are constructed from other results. References complete and accurate we can verify the basics of each addition to the structure of science, or at least be traced to a level that we believe are reliable. References also tell the reader what parts of the thesis are descriptions of what we already knew, and which parts are your additions to this knowledge. In a thesis written for a reader not necessarily familiar with the literature of your field, it should be especially clear. It may be tempting to omit a reference to the readers to believe that a good idea or a nice piece of analysis is yours. I strongly advise against this risk. The reader will probably think: I wonder if this idea is original? "The reader can learn by visiting the website or the library, or even a phone call. I prefer the active voice for a thesis, but if you write in the passive voice, be especially careful to identify who did what. "The sample was prepared by heating yttrium ..." does not make clear if you did this or if the supplier of yttrium did. "I prepared the sample." is clear.


The style of your thesis is important, The text must be clear. A well-written thesis will be easier to read, which will bring some benefits. Scientific writing should be a little formal --- certainly more formal than the text you are reading. Remember that those who read French are a minority, and slang and informal language will be more difficult for a non-francophone.
Short words and simple sentences are generally better. However, sometimes you need a complicated sentence because the idea is complicated. If your fundamental statement requires several qualifications, each of them may need a sentence: "When [qualification], and that [stipulation], and if [condition] then [statement]." Some lengthy technical words will also be needed in many theses, particularly in areas such as biochemistry. Do not sacrifice truth for brevity. "Black is white" is simple and striking. It will appeal to a writer of advertising. "Objects which are very different albedo may be illuminated differently so as to produce similar spectra of reflection" is more complicated and less common uses of words, but in comparison with the previous example, has the advantage of being true. The second example would be fine in a physics thesis because a French physicist has no problem with those words. (A physicist who would not know all these words would be happy to correct its deficiencies by either the context or by consulting a dictionary.)
Sometimes it is easier to present information and discussion through a series of numbered points, rather than one or a few paragraphs long and awkward. A list of points is often easier to write. You should however be careful not too make use of this presentation: your thesis must be a persuasive discussion, and 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 makes it immediately clear what you have done and what has been done by others. Passive voice ("the frequency was measured ...") makes it easier sentences ambiguous, incorrect or clumsy. This choice is a matter of taste: I prefer the active because it is clearer, more logical and makes it simple attribution. The advantages cited for the passive voice are (i) many theses are written in the passive voice, and (ii) people are very polite use of "I" immodest.


There is no need to turn your dissertation into a masterpiece of editing. Your time and energy can be more effectively spent to improve the content and appearance. Often, a clear and acceptable diagram can be drawn by hand faster than with a graphics program. If you put the thesis in electronic form, you can scan the version made by hand in black and white, she will not know a huge file.
In general, students spend too much time on the charts --- time that could have been spent in criticizing the talks, would make it clearer explanations, looking for errors and thinking more about the significance of the results. The reason is that drawing is easier than thinking.
Readers are not impressed by a thesis that is too long. They will not be happy having to read a lot vague or unnecessary text.

The end.

A deadline is very useful for several reasons. You must make the argument, even if you think you should make another draft of this chapter, or change the order of this section, or any other refinement. If you do not have a deadline, or if you think of postponing, I beg you to accept that a thesis is a work as big as you can not make it perfect in a finite duration. It there will inevitably be things that you could have done better. There will inevitably be some mistakes. Thanks to a law similar to that of Murphy, you will discover a copy when you open the newly arrived linker. No matter how you think and how many mistakes you find, there will be some things that could be improved. Needless to hope that reviewers will not notice: many reviewers feel compelled to find examples of improvements (if errors) just to show they have read. So be a deadline and stick to it. Make the argument as good as you can in this time, and then submit it! (In retrospect, I see an advantage for theses written before the process-texts. We paid a typist to type it and there was no money to do it more than once.)

How many copies?

Discuss this with your adviser. For board members, libraries, university and for you of course. But you should make copies to distribute. These copies should be sent to other researchers working in the field so that:
  • they can discover what a wonderful job you did before it appears in the newspapers;
  • they can see the details of methods and results are published more briefly elsewhere;
  • they may realize that you are an excellent researcher.
These achievements could be useful if a site appears in their labs, or are the readers of your proposal for a contract for postdoctoral research. Simply having your name on their shelves can be an advantage. The following commentary is by Marilyn Ball of the Australian National University in Canberra: "When I finished my thesis, a postdoc wisely advised me to give a copy to my parents. I never thought about it, because I could not imagine what they would do. I'm still glad I took his advice, because my parents really appreciated receiving a copy and have proudly shown for years. ( My mother never finished high school and my father worked to build and repair trucks. However, they were delighted to have a copy of my thesis.)


In an ideal situation, you can spend a large part --- perhaps most --- of your time writing your thesis, which can be bad for your physical and mental health.
Position your chair and computer properly. The health department of the university or professional typists can give you advice or a poster with relative heights recommended the look healthy and also exercises you should do if you spend much time at the keyboard. This is important: you do not want to be bothered by back pain or neck. Try interrupting long sessions at the keyboard with other tasks. If you do not type by touch, you should learn as well as for productivity for the health of your neck. There are several good software packages that teach touch typing interactively. If you use one for say 30 minutes a day for two weeks, you'll be able to type by touch. From here until the end of the thesis, the write speed so you won repay more than ten hours invested with the software. Be careful, however, not to use the typing exercises to avoid working on the thesis.
Do not drop exercise during drafting. Lack of exercise makes you feel bad skin, and you do not need another reason to feel bad by writing a thesis. 30-60 minutes of exercise a day is not lost time of writing I find that if I do not exercise regularly, I sleep less and lasts longer. Go to college at the foot or, if it is far out of the subway stations around before. Many people notice that a walk helps them think, or clears the head. You might find that rides occasionally improve your productivity.
Remember to eat, and made an effort to eat healthy food. You should not lose shape or risk a disease at this critical time. Exercise is good to give you an appetite. Obviously you have little time to cook, but keep a stock of fruit, vegetables and bread. It takes less time to make a sandwich or a salad to go that fast food restaurant, and you will feel better after.
Products psycho
PhD students have a long tradition of using coffee as a stimulant and alcohol or marijuana as relaxants. Used in moderation does not seem to have negative effects on quality of thesis produced. However, excess obviously reduce productivity: double and multiple coffees you fly too much to sit down and write several digestive evening you slowed down the next day.
Others sympathize with you, but do not forget them. Husbands, lovers, family and friends should not be underestimated. Spend time with them and, in doing so, have fun. Do not spend your time together to complain about your thesis: they already offended by the idea because it separates you from them. If you can find another PhD student, not necessarily in the same discipline, it might be therapeutic to complain to one another about the masters thesis and difficulties.


Writing a thesis is hard work. A postdoc anonymous said: "You should tell them it will be unpleasant, it changes their lives, they will drop their friends and their social lives for some time. It lasts for a period almost every student. " She is right: it's certainly hard work, it will probably stressful and you need to adjust your rhythm to it. It is also an important rite of passage. From scholars everywhere, I wish you 'shit' to the thesis!

A suggested structure

The table of contents and chapter titles below is a suggestion. In some cases, one or two of these categories may be irrelevant to your thesis. The chapters of results and discussion are often filled in several chapters. Think in terms of chapters and decide what is best for your work report. Then make a list of topics that show what you put in each chapter. Try this list in detail, so you end up with a list of points that correspond to subdivisions or even paragraphs of your thesis. At this stage, think carefully about the logic of the presentation: In each chapter, it is often possible to present ideas in different orders, and these different versions will not be equally easy to follow. If you make a plan for each chapter and section before you start writing, the result will probably be clearer and easier to read. It will also be easier to write.
Keynote speakers Pages
In the original version of this document, I give a list of pages Listing agents according to the rules and traditions of the English universities. For a French view, I suggest you ask the office for doctoral studies in your institution if they have rules, and also look at examples of recent theses that have not been subjected to bureaucratic problems. Usually, a French view has a title page that shows:
    the name of the institution Thesis type specialty the name of the doctoral the title of the thesis the date of the defense the names of jury members, with the president and (s) Rapporteur (s) indicated.
It is followed by a page (optional) thanks, a table of contents, and then the chapters of the thesis. The summary in French and English, is at the end. I discuss the summary before proceeding to other parts of the thesis in normal order.
Your whole argument, that part will be the one the public the greatest. It is better to write to the end but not at the last minute because you'll need to do more jets. In addition, the summary is so important that it is really worth asking a friend or colleague to edit your English version. The summary should be a distillation of the thesis: a concise description of the problem (s) address (es), your method (s) resolve, your results and conclusions. A summary should be independent. Usually it does not contain references. When a reference is necessary, its details should be included in the abstract text. Check the length limit with your university.
Most doctoral students are in a page, their thanks to those who helped directly in scientific subjects, and also to those who helped them indirectly by providing such essential qualities as food, education, genetic , financial support, assistance, advice, friendship etc.. If any part of your work is a collaboration, you must clearly indicate which parts that did work.
Table of Contents
The introductory chapter begins on page 1. If you want to number the previous pages, do it with Roman numerals. It is very useful to the entry sub-divisions of each chapter, as well as chapter titles. Remember that the thesis can be used as a reference in the lab, so it is important to be able to easily find what you are looking for.
What is it about and why is it important? Explain (s) problem (s) as simply as you can. Remember that you worked on this project for several years, so you know very well. Try to mentally back and take a broader view of the problem. What is the ratio of your problem with the major themes of your discipline? Especially in the introduction, do not overestimate the reader's familiarity with your subject. You are writing for experts in your discipline, but they are not experts in your particular subject. It is useful to imagine a particular person --- think of a colleague you may have met at a conference to your discipline, but who worked in a different domain. She's smart, she has the same general education, but she knows little literature or techniques specific to your particular subject.
The introduction should be interesting. If the reader finds it tiring, it is unlikely that you will revive his interest in the chapter on materials and methods. In the first paragraph or two, tradition permits a language that is less dry than the scientific standard. I hope that inspires you about you always, and this is the place to enjoy it. Try to inspire the reader want to read the tome that arrived on his desk. Go to the library and read the introductions of several theses. There were some that he gave you the desire to read more? There were others that he was boring?
This chapter will probably go through several streams, so it can be simple, logical, short and interesting time. For this chapter, I find it helpful to ask someone who is not a specialist to read and to make criticisms. Is this a sufficient introduction? Is it easy to follow? There are arguments for writing this chapter --- or at least to make a major revision --- towards the end of writing the thesis. Your introduction should tell where the argument goes, what will become clearer during the writing. In addition, the chapter needs anything you will learn about scientific writing.
Review of Literature
How is this problem? What we know of already knew about? What are other methods we tried to solve it? In principle, you have already done a lot of hard work, if you've watched the logs on as you have promised to do three years ago. If you have already summarized the important papers for you, you have a good starting point for review.
How many articles to cite? At what point should they be relevant to be included? It's a tough question. A hundred is reasonable for many theories, but it depends on the field. You're the world expert on the subject (narrow) of your thesis: you must show this in your journal.
A point roughly politics: do not forget to quote the relevant articles published by members of your board, or the lab heads who can send you a copy of the thesis in the next year.

The middle chapters

In some theories, the middle chapters are simply sections of the newspaper whose candidate was the lead author. There are several reasons to avoid this format. One reason is that in a thesis, it is allowed and expected that the level of detail is deeper than that of an article in a scientific journal. For publications, it is usually necessary to reduce the number of curves. In many cases, all data and interesting cases may be included in the thesis, not just those already published in the newspaper. The experimental detail is more explicit in a thesis. Quite often, a researcher requests a copy of a thesis to be of benefit to detail the manner in which the study was done.
Another reason is that there will probably matter in common among the sections' Materials and methods "of your publications.
The exact structure of the middle chapters vary PhD thesis. Often in a thesis, it is necessary to establish a theoretical background, describe the experimental techniques, and then report what has been done on several different problems or different stages of the problem, and then finally to present a model or new theory based on the new job. For such a thesis, the chapter titles are: Theory, Materials and Methods, first problem {} {} second issue, third issue {} {} proposed theory or model and then the concluding chapter. For other theses, it would be more logical to discuss the different techniques in different chapters, rather than having a single chapter materials and methods. Here are some comments on items of materials and methods, theory, results and discussion, which will not always correspond to each chapter.
Materials and Methods
This chapter varies greatly PhD thesis, and may be absent in a theoretical thesis. The principle is simple: it should be possible for a competent researcher, following your description to reproduce exactly what you did. It is likely that this event will start one day after you're gone, another researcher wants to make a similar experience, or in the same lab and with your device, or independently elsewhere. Please write this chapter to make easy the task of the researcher. In some theories, especially theories or multidisciplinary applied research, there may be several such chapters. In this case, the different disciplines should be included in the chapter titles.
When you report a theoretical analysis which is not original, you should bring enough material to allow the reader to understand the development and its physical bases. Sometimes you will present the theory ab initio, but you should not reproduce two pages of algebra that the reader might find in a standard text. Do not put the theory which has no connection with the work you've done. In drafting the theory section, do at least as well as the discussions physical equations. What are these equations? What are the important cases?
When you bring your own theoretical work, you need more detail, but consider placing long derivations in the appendices. Remember also to the more logical order and style of presentation: the order in which you have done the work is not always as clear for the presentation.
The suspense is not necessary to bring science: you should tell the reader where you'll drive before you start. (And of course you can tell him or after you have brought.)
Results and discussion
Sections of results and discussion are often placed in several chapters of a thesis. This practice is recommended because of the length of a thesis: you may have several chapters of results and, if you wait until they are all presented before the discussion, the reader may have forgotten about you're talking about. Be careful that you described the conditions under which you obtained each set of results. What parameters were kept constant? What were the values ​​of other relevant parameters? Show measurement errors and use appropriate statistical tests for comparisons and analysis you do.
Be careful also with your curves. The origin and intercepts are often important, therefore, unless the scale of your axes and the range of your data makes it impossible, the zeros of the two axes (or axis) should appear on the curve. You should show error bars on each data unless the errors are very small. For a single measurement, the bar should be your evaluation of experimental errors in each coordinate. For multiple measurements, the bars should include the standard error. ? (Standard error?) In a set of data, errors can be quite variable, so when this happens, your regressions must be done by minimizing the sum of squares of the ratio of the difference for each point to its error. That is a flaw in a lot of software for manipulating data and curves. Mike Johnston, a student at UNSW, wrote software that curves with such regressions. It is .
In most cases, your results need discussion. What do they mean? What contribution do they do to human knowledge? Are they in agreement with current theories? They give new perspectives? They suggest new theories or mechanisms?
Try to get away from your usual perspective and look at your work and far. Do not just ask what he means by the orthodoxy of your own research group, but as other people in the field --- or even non-specialists --- can see it. Does it have implications that are unrelated to the issues on which you started?

Final chapter, references and appendices

Your findings will appear in your resume but in very brief form, because it must also include other material. A summary of conclusions may be longer than the conclusion of the summary, so you have the space to be more explicit and more careful with qualifications. You can find it useful to put your findings into a list of points. It is often the case that scientific inquiry delivers more questions than answers. Your work suggests there additional questions interesting? In what way can we improve a study like yours in the future? What are the practical implications of your work?
This chapter should be reasonably short --- maybe a few pages. As with the introduction, I think it's a good idea to ask a layman to read and to make criticisms.
References (see also Literature Review)
You'll probably be tempted to omit the titles of articles cited, and the university allows this practice, but think of all the times when you tried to find a reference, simply by looking at its title, it was not useful after everything.
If there is material that should be in the thesis but which would break the thread or annoy the reader unbearably, put it in an appendix. Typically, the appendices include: important and original software, data files that are too large to be placed in the results chapters, graphs or tables of results data that are not sufficiently important to keep in the main text.


Post a Comment