| Home · Topics · Link Farm |
Thesis FormulationThis page contains frequently asked questions on how to formulate your thesis. These may be questions that are frequently asked or questions that we hope would be frequently asked. |
Navigation |
There are no hard-and-fast rules how to formulate your thesis. Scientific writing is the topic of several good books. I recommend particularly The Craft of Scientific Writing, by Michael Alley. This section covers only a few topics that come up often.
Use the book style for theses.
Use the mathtime package in order to improve
legibility on PDFs and to avoid the ugly mix of Times for the main
font and Computer Modern Roman for the mathematics.
I'm a sucker for Postscript, but I recognize that PDF is the document format of the future. I've had texts badly mauled in conference proceedings only because I submitted them as Postscript files. People who had submitted PDFs had no problems. So if you have a choice, use PDF.
Use them. There is nothing so dreary as a thesis with lots of spelling mistakes, especially since spell checkers make most spelling mistakes unnecessary. I am inclined to view too many spelling mistakes as signs of sloppiness on the part of the student, and this may well reflect on your grade.
Generally, you should write your thesis in the language in which you feel most comfortable for technical writing. If you write in a language that is not your mother tongue, do not expect your colleagues to do your proofreading for you. It is OK to expect the occasional typo or wrong formulation to be caught, but you will have to come to grips with basic grammar and usage yourself. The section on reference books may have some useful hints.
If your thesis is the outgrowth of a paper, you should write the thesis in the language of the paper, which will generally mean English. Conversely, if you start your thesis with the goal of extracting a paper from it, you should at least consider writing the thesis in English, because the paper will most likely be in English, too. This way, you avoid unnecessary translation to and from the paper.
You choose. But whatever you choose, stay consistent. Good dictionary are essential here. Here are some examples of different use:
| American English | British English |
|---|---|
| -ize (prioritize) | -ise (prioritise) |
| -or (flavor, behavior) | -our (flavour, behaviour) |
| Apartment | Flat |
| Sidewalk | Pavement |
| Subway | Underground |
For papers, if there are two or more authors, ‘we’ is always the correct choice. This section is about theses, however, and they have only one author.
Use ‘we’ to indicate a dialogue with the reader (as in “the author(s) and the reader together”), not to give the impression of plurality (“the authors”). An example of bad usage of ‘we’ would be: “We believe that…”. Good usage would be “If we accept the hypothesis that …, we are forced to conclude …”.
Use ‘I’ only if your personal opinion is important. (Hint: it usually isn’t.)
The general rule is to avoid the apostrophe in favor of the long forms. So in general, use ‘is not’, ‘has not’, etc. Sometimes the long form makes a sentence awkward, in which case you can use the short form to improve the sentence. That is a matter of personal style.
That is, should you write in a formal style or should you use colloquialisms?
Until you know exactly what you are doing, it is best to stick with a formal style and not to use colloquialisms. Formal writing is more polite; colloquialisms might make the reader feel as uncomfortable as Terry Jones in Monty Python's “Nudge, Nudge” sketch.
| Use | Don't Use |
|---|---|
| The data agrees with Eq. (1) to within 5%. | The data shows that Eq. (1) is a winner. |
| We had to make some modifications. | We needed to tweak the software a bit. |
Stay away from jokes. They are very hard to get right. Something that is tremendously funny when you write it could turn stale in a year or two. Worse, within several years, somebody may not even be able to understand the joke anymore. Your sense of humor might not agree with your reader’s. There are just too many things that can go wrong, so again the advice is to stay away from them.
Note that you can write plain English that is formal, yet not stuffy. You don't have to say “usage is subject to terminological inexactitude” when you mean “the term has been used in a wrong way”. In fact, it is good style to avoid polysyllabic when simpler words suggest themselves. Winston Churchill once said, “[b]roadly speaking, the short words are the best, and the old words best of all”, and that is still good advice.
There is much confusion about how to write numbers. For example, which of the following sentences is correct?
Only sentences 2 and 7 are correct. The general rule is: spell out small numbers if they are used as adjectives. Write them as numbers when they are used as nouns (i.e., when talking about numbers as numbers). (This formulation is from Knuth, Larrabee, Roberts, Mathematical Writing, MAA, 1989.) Also, don't start a sentence with a number and don't use adjacent numbers.
According to these rules, sentence 1 is wrong because it uses the small number ‘2’ as an adjective; sentence 3 starts with a number; sentence 4 has the same problem as sentence 1; sentence 5 should be rewritten, “…Example 2 four times”; and sentence 6 uses ‘four’ as a noun.
The following rules are lifted from Knuth, Larrabee, Roberts, Mathematical Writing, MAA, 1989. Each rule is stated in a sentence that violates the rule, so do not mistake the sentences below for examples of good writing. They are anything but that.
The last rule should be used with caution, however, since the rule against splitting infinitives is by no means as strong as it once was. The Wikipedia (agreeing mostly with Fowler here) has this warning:
On a historical level, it is possible that years of attacks against split infinitives by prescriptive grammarians have cowed some people into needless reluctance to split other compound verb forms. For example, people will contort sentences to avoid placing an adverb in its usual position between the auxiliary verb and the participle, leading to constructions such as, "The argument originally had been used…" instead of "The argument had originally been used", which is more natural for most speakers.
Most of the following are lifted from Knuth, Larrabee, Roberts, Mathematical Writing, MAA, 1989.
| Right | Wrong |
|---|---|
| implement | impliment |
| complement | compliment |
| occurrence | occurence |
| dependent | dependant |
| preceding | preceeding |
| its (belonging to it) | is’s (it is) |
There exists considerable confusion when to use ‘which’ and when to use ‘that’. Which of the following sentences are correct?
Sentence 4 is correct. Sentence 3 is certainly wrong: ‘which’ should not be used here. Sentence 2 is correct, both in its meaning and in its usage of ‘that’. Whether you consider Sentence 1 to be correct depends on whether you believe that commas are always unnecessary or not. That sentence could be paraphrased as “Commas are always unnecessary; therefore, don't use them”. This is probably not what the writer intended—because he used a comma!
The Chicago Manual of Style devotes Sections 5.55–5.63 to the correct usages of ‘which’ and ‘that’. In order to profit from these sections, you have to know what pronouns and antecedents are. But you need to know that anyway, even if you don't have to be able to give a formal definition.
As a rule of thumb, ‘which’ is correct when it introduces a subclause that is separated from the main clause by a comma (Sentence 1) or when used in a question or question-like way (Sentence 4).
Good reference books on language usage are essential for good writing, no matter whether you are writing in your mother tongue or not. You will need a good dictionary, a thesaurus, and possibly a manual of style. A dictionary lists words and their definitions, a thesaurus can help disambiguate between similar words and helps find synonyms and antonyms, and a style guide helps finding formulations that are generally thought of as “good usage”.
You can use online references, provided they come from reputable sources, such as the Oxford English Dictionary (OED) or the Duden. You should emphatically not use Leo (see how there is no link?), except when you know the correct word, but it just eludes you. Leo will present you with a list of possible translations of a word, but doesn't give you the means to disambiguate between them. For this, you need a good thesaurus.
Online references are generally not reputable when it comes to citing them in a printed work such as a thesis. This is because the content may change or the URL may have become invalid by the time a reader follows the reference. This also effectively excludes definitions from the Wikipedia from being cited in printed works (but not in electronic publications, such as this page).
The canonical reference books for the German language are the Duden and Wahrig books, although there are other good references, too. (Warning: since most of my writing is in English, I have not yet used either the Duden or Wahrig dictionaries. I have only used the Duden for occasional spell-checking; today, I rely more on automatic spell checkers.)
The definitive reference for English is the Oxford English Dictionary. The print edition of the entire dictionary has twenty volumes! The name suggests a predominance of British English, but the dictionary covers words from all English-speaking countries. Some people also like Merriam-Webster's Dictionary, but be aware that words may be missing due to editorial prudery. (For example, the f-word is not in Merriam-Webster's, but it is in the OED. In my opinion, the attempt to remove words that the dictionary editor does not like smacks of Newspeak and I therefore do not recommend M-W's.)
Plagiarism is the act of passing somebody else's work off as your own. It is the ultimate in academic dishonesty. If you plagiarize someone in your thesis, and get found out, you will face severe charges. At the very least, you will fail your thesis. At worst, you could be expelled from the University.
You can avoid potential charges of plagiarism by correctly citing others' work. You do not need to cite commonly known facts, such as the sum of integers from 1 to n. Everything else that is not commonly known in computer science and that is not your own original work must be quoted.
Note that putting someone's ideas in your own words does not magically make their ideas your own. No matter whether you rephrase something or quote it verbatim, a quote needs a reference. A reference tells the reader where to find the source from which you got the quote. LaTeX manuals tell you how to use BiBTeX, a literature database and reference formatting program for LaTeX; the Wikipedia Manual of Style can give good hints on how to put references into online sources.
There are several typographic niceties that deal with
verbatim citations; they are handled in good style guides, such as
the Chicago
Manual of Style. LaTeX usually does the right thing when asked to
typeset a quote or quotation.
As explained above in the section on Reference Books, online articles should not be cited. This is because online material can change, move or disappear completely without your control. For example, it is possible that a Wikipedia article changes between the time you write the citation and the time the reader decides to follow up on it. This is not likely with a printed edition of the Encyclopædia Britannica. Cite online material only if it is not published elsewhere.
Self-plagiarism is when you try to pass off your own work as new work done by you. It can happen, for example, when you submit the same text to two different conferences simultaneously, or when you submit a marginally altered version of an accepted paper to another conference. Each conference has a different rule on whether previously published work is acceptable or not; see the conference's Web page. If in doubt, contact the conference's program chair.
To avoid the accusation of self-plagiarism, cite yourself.
You will undoubtedly create your own literature database or even databases. The question then arises, how much information about my reference do I put into the database? The temptation is not to include any fields that you know will not be output by the bibliography style that you are currently using. Resist that temptation! Always put in as much information as you can get, provided you can get it easily. Apart from the rather obvious information about author or authors, title and proceedings name, that includes things like conference name, page numbers, ISBNs, publication dates, conference locations, publisher information, DOIs etc. You will probably need them at some point in the future.