Put your examiner in a good mood and gain style marks by avoiding common errors when writing your report...
- Remember that the report should be a balanced, concise summary of the advances achieved, with supporting detail that
provides their experimental or theoretical justification, together with an indication of the relevance and context. It
should not assume significant prior knowledge: if you wish to avoid repetition of material readily available elsewhere,
simply replace it with an appropriate reference.
- The report should not be a diary: your job in writing it is to weave a useful and readable document from the available material,
and you don't need to (and shouldn't) make it represent the balance or order of the work it describes.
It is generally inappropriate to write up unproductive or unsuccessful activities, unless there is a new and important
conclusion to draw from them; and the weight given to different activities or topics need not reflect the effort required
or time spent in pursuing them.
- Mathematics in particular should not require much prior knowledge, so ensure that ALL variables and parameters are explained
properly at their first occurrence.
- The report should be in the style of a scientific paper... so look at a scientific paper. See, e.g.,
American Journal of Physics,
Physical Review Letters.
This should also indicate an appropriate structure for your report, and the format and content of the abstract.
- Note in particular the formats for
It is rare for students to get the formatting right when it has to be applied manually, as in MS Word, so it is well worth learning
to use a TeX/LaTeX package which, with a suitable choice of style file, will
look after things for you - and make a fine job of your equations too.
- references and their citation in text: references should generally be listed in the order in which they are cited,
and all references should be cited
- figure captions: again, these should be cited within the main text, and there should be no uncited figures. Captions
should generally be sufficiently informative for the reader to understand roughly what they're depicting without
reference to the main text; detailed discussion however should be in the main text of the paper. The source should
be referenced for any figures taken from other works
- tables: these should be numbered and captioned exactly as for figures
- equations: these should be centred and numbered, with the number at the end of the line
- symbols: there are clear typesetting conventions to avoid ambiguity - see physics.nist.gov/cuu/pdf/typefaces.pdf
- Note the tense(s) used in scientific papers: these are reports of work carried out, so are generally written in the past
tense, with occasional use of the present. Use of the future tense is only rarely appropriate.
- Note also the usual style adopted. Language is quite formal and impersonal, with far greater use of the passive mode than
is common elsewhere. Colloquialisms, exaggeration and litotes should never be used, and rhetorical questions are extremely rare.
Scientific papers are intended for an international audience, and need to convey complex meanings with concise precision.
- All text should form complete sentences within paragraphs. In particular, pieces of apparatus and their arrangement should be
described - never listed. The use of bullet-points is extremely rare.
- Common errors of language (including a few hobby horses):
- allowing spelling checkers to insert capital letters (and even paragraph indentation) after equations in mid-sentence
(typically 'Where...'). Remember also that spelling checkers merely ensure that each word exists; they cannot judge whether
the word is correct.
- misplaced or dangling modifiers - eg "After being weighed, I transferred the sample to a beaker":
this implies that the writer was weighed
- incorrect prepositions:
- errors from colloquialisms and homophones:
|try and obtain||try to obtain|
|should of done||should have done|
- mistaking past participle with archaic adjective: 'has been proven' (should be 'has been proved')
- confusing the singular and plural forms of nouns of Latin origin:
|datum (or point, value)||data|
- confusing 'may have' with 'might have': may implies that the situation is still possible; if it is not, one should use might.
|Without evidence to the contrary, she might still be alive.||sorrowful reflection|
|Without evidence to the contrary, she may still be alive.||grounds for hope|
- for both amusement and illumination, consider William Safire's Rules for Writers from his On Language column
in the New York Times Magazine. A few favourites - with some evolution - are:
Many more compilations have been made by, amongst others, the
|Avoid annoying alliteration.||Don't overuse exclamation marks!!|
|Remember to never split an infinitive.||Verbs has to agree with their subjects.|
|Never use a preposition to end a sentence with.||About those sentence fragments.|
|And don't start a sentence with a conjunction.||Make each pronoun agree with their antecedent.|
|Writing carefully, dangling participles must be avoided.||Don't verb nouns.|
|Don't use no double negatives.||Last but not least, avoid cliches like the plague; seek viable alternatives.|
- Common errors of content:
- confusing the terms error, accuracy and precision (the distinction can be subtle). While, colloquially,
our schoolboy reference to errors persists, an error is really something that is incorrect; having realized our
error, we should have repeated the experiment correctly, so it should not appear in our account. What we probably meant
was either the precision or the accuracy: the precision indicates the uncertainty in the measurement or
calculation, and the accuracy the degree to which the result was correct or valid. Roughly, the precision of a measurement
may be improved by repetition and averaging, but its accuracy - if determined by calibration 'errors', for example - will
Surely you're joking, Mr Feynman!, illustrates this with the tale of determining the length of the Chinese Emperor's nose.
Unable to meet the Emperor, researchers instead conducted a survey, and asked citizens across the country how long they thought
the Emperor's nose to be. The result - obtained by averaging the many answers - was extremely precise, but hardly accurate.
Your analysis should generally consider both the precision of your results, and possible sources of inaccuracy.
When you have determined these, remember to quote them with a reasonable precision themselves: more than one significant
figure is rarely justified for the 'error' on a figure.
- failing to write a proper introduction and summary. The introduction should outline, without assuming detailed knowledge, the
nature and interest of the problem, and summarize the direction of the report. The conclusion should draw the paper
together, and not merely summarize what has already been stated earlier.
- Many students have given up working on the quality of their written English. Don't! It is still possible - and indeed not
difficult - to make huge improvements even at this stage, and the number of 'rules' that need to be understood is actually quite small.
- Buy a textbook on writing English/English grammar and punctuation/technical writing
- Buy a reference book such as Fowler's Modern English Usage
or Gowers' Complete Plain Words.
(Note that both these books were originally written as guidelines to improve official writing.)
- Look at the style guides of newspapers and broadcasters, such as the
and The Times or
- Thumb through popular treatises on the English language, such as Bill Bryson's
or Lynne Truss's Eats, Shoots and Leaves.
- If you don't already, read some well-written novels - maybe the book of a film that you liked - or factual tales of
great explorers and so on. Before publication, books such as these will have been edited, corrected and generally
massaged into a reasonable style. Remember, of course, that they're not the scientific style - but they should
encourage correct use of grammar and punctuation.
- Practice - for you're almost certain to have to do this for real some time in your future career - reducing the number of
words you require to say something: take your first draft, and try to cut the word count by removing unnecessary
words or phrases or finding more elegant constructions. We tend to pad out speech - and make ourselves sound more
grand - by adopting wordy or somewhat pompous constructions, which can be greatly simplified. For example, 'In
order to' usually loses nothing if the 'In order' is deleted. Most people can halve the number of words
used (from their first draft) without changing the meaning of the text, and can probably achieve a further factor
of 1.5-2 at the expense of only minor omissions. You may find the result unduly terse, but the exercise is an
excellent one in forcing you to consider the meanings of, and relationships between, the words that you write.