A personal view of scientific writing or The mistakes I have made!

Download 20.71 Kb.
Size20.71 Kb.

A personal view of scientific writing or The mistakes I have made!

  • John Kirby
  • Postgraduate tutor
  • “This presentation was given by Prof John Kirby’s it is his personal overview on scientific writing.
  • He gives his own methods of writing, outlines the mechanics of publication, and offers some of the mistakes he has made!”

What do scientist write?

  • Abstracts
  • Research papers
  • Reviews
  • Book chapters
  • Books
  • Grant applications
  • Theses (and assessment reports)
  • In the course of your career as a scientists, you will probably be writing much of the above, each of which is different and needs to be written in a different style. Each output is different, appealing to a similar audience, but the way in which you write in will vary.

Clarity is essential

  • All documents need to read swiftly
  • No room for ambiguity
  • English may not be the readers 1st language
  • There are no real rules, but essentially someone has to understand what you have written, so clarity is very important. It is important that information is assimilated rapidly, so documents must be able to be read swiftly, with key points standing out. There should be no room for ambiguity, and it must also be realised that the reader may not have English as their first language.

Rules for clarity

  • Everyone will acquire his or her own style
  • However, there are some general rules:
    • Use short sentences which express single concepts
    • Use short paragraphs
    • Avoid clever clauses and parentheses
    • Use good grammar and punctuation
    • If in doubt, keep it simple
  • Everyone ultimately develops their own style and the way they use the English language, but there are some general rules. Short, simple sentences expressing a single concept, should be used, as should shorter paragraphs. Use simple language, good grammar and punctuation.

A useful tips

  • You will annoy your friends, but please try to read your own work out loud.
  • If you find it hard to speak then
  • then something is wrong with the text
  • When correcting text try little and often rather than long boring sessions
  • Get a friend to read your work
  • A useful tip is to read aloud what you have written – if it doesn’t flow easily , there is something wrong with the text. When correcting your document, try reading short pieces, rather than long sessions of proof-reading, as you are more likely to skim over your text and miss errors. Using people around you to read through your document often helps.

Don’t worry too much!

  • Many international journals now employ copy editors and proof readers who pick up most errors before publication
  • Often this will convert your English into American English!
  • Many journals now employ proof-readers and copy editors who will pick up most errors before publication.

Common errors

  • Keep track of singular and plural forms
    • Remember data is the plural of datum!
    • Hence, “these data suggest…”
  • Keep track of tense
    • Most experiments and procedures will be described in the past tense
  • There are a lot of common grammatical errors, which won’t be mentioned in much detail here, but in the reference section of the library are a number of publications which are quite specific in their approach to scientific writing, telling you how to write a paper, how to use English grammar and basically give exhaustive lists of the type of common errors made. It is useful to keep note of singular and plural forms, and also of the use of tense in a sentence.

More on tense

  • A good way to separate what you have shown from what others have reported is to mix tenses in your writing
    • This is common in a discussion section
  • For example:
  • The protein was non-functional after modification of the terminal residue. This result is consistent with those from other groups (references) and indicates….
  • A good way of separating what you have done from what others have reported is to use mixed tenses in your writing, as shown in the example above, using the past tense to describe what you have done, then present tense to discuss what others are discussing or showing.

“Instructions to authors”

  • Read these before you start writing!
  • All journals have a house style
    • Examples:
      • The BMJ insists all papers are written in first person
        • I demonstrated that….
      • Most pure science journals tend to require third person
        • These data demonstrate that….
  • One of the key things to do, when writing a scientific paper, is to read the instructions for authors before you actually start writing the paper. Decide what you are doing and where you are aiming the paper, then look at the instructions, as all journals have their own house style. This can vary considerably, depending on the publication, as shown above.
  • This is the BMJ, which gives one of the most comprehensive lists of instructions to authors, even giving directions on which dictionary to use. It also gives a detailed description on the use of English grammar.


  • Dear John Kirby
  • I looked at your manuscript more closely and at first glance it seems
  • to be rather long. The limit of articles is 6,500 words as stated in the
  • instructions.
  • Therefore, I would like to know the exact word count of your paper and
  • if it is too high to shorten the manuscript to meet the guidelines.
  • Sincerely yours
  • Read the instructions! If your work does not meet the criteria, you will have to rewrite – many journals now have around a 3500 word limit, and as many now take online submissions, it is very easy for the editors to do a word count.

Writing a scientific paper

  • First question
    • Have I got sufficient data to support my conclusion?
  • Have a look at a typical journal in your field
    • What do the results sections look like?
    • In my field they seem to contain about 2 tables and 6 figures
  • How do you know when you can write a paper? The first thing is to check that you have collected sufficient data to support your conclusions. The best thing to do is to look at typical journals in your subject, either online or in the library – look at the results sections, as these are actually the whole of the paper – all the rest (hypothesis, etc.) simply support the results. This will give you an idea of the scale and output you need for your results section.

We are not butterfly hunters!

  • Think about what you need to support output and what you actually need to write a paper – you can have a hypothesis and work out the minimal data set to support that will take you through to your conclusions. Rather than randomly collect data, you can be clinical and focussed to generate a results section.

The next step

  • When you have decided what you are trying to communicate set up a mock results section
  • Label several sheets of blank paper:
    • Table 1, 2 etc
    • Figure 1, 2 etc
  • Roughly sketch what data will go on what page
  • Shuffle the pages into a logical order
  • Does it seem complete
    • Yes? Write the paper!
    • No? What else do you need to do?
  • When you have decided that you have enough data and how you are going to address your question, it may be an idea to set up a mock results section, sketching out as outlined above. It will become obvious if there are any gaps, so you can go back to think what you need to do to make the story.

Choose your journal

  • Look carefully at a selection
    • Which is most appropriate?
  • Talk to your supervisor(s)
    • No point going for Nature unless everyone agrees it is worthwhile
  • Consider the impact factor
    • Not all journals are equal!
    • The impact factor is a measure of how often an average article in a journal is cited
  • Going back to thinking about which journal to choose - your work is not suitable for every journal and every journal is not suitable for your work, so do make your selection carefully. Talk to your supervisor and don’t be unrealistic. Not all journals are equal – often citing of articles indicates the impact factor.

Writing the paper

  • Read the instructions to authors
  • What sections should the text be divided into?
  • Often:
    • Title
    • Abstract
    • Introduction
    • Methodology
    • Results
    • Discussion
    • References
    • Figure legends
  • The sections that the text may be divided into are often, but not always, as shown – read the instructions to authors.
  • The title page format should have authors and corresponding authors, and a running title. Keywords should be used, maybe any abbreviations listed, and any commercial interest which may conflict with the data.
  • The abstract is something which quite often is the only thing people read, and therefore needs to be clear and cleverly worded, an advert for your paper. The introduction
          • What do you do first?

This is what I do

  • On a 1000 mile journey, the hardest thing is the first step.
    • Make the first step easy!
  • The methodology is often easiest to write as is simply descriptive.
    • Order this in the same way as you will present your results
  • First steps are personal preference, but this is generally how I start.
  • To start with methodology, simply write what you did to generate results, maybe adding statistical analysis, or Ethics Committee approval, but generally following the order of your results.

The next step

  • I usually write the results text next
  • This is also descriptive as you simply describe your data (figures and tables)
    • “These data show that something is higher/faster/larger than something else (p<0.001)”.
  • A common error is to add discussion and interpretation to this section
    • This leaves nothing for the discussion section!
  • Next, add your results – just the description, as there is generally a separate section for discussion.

The home straight

  • I usually then write the introduction
    • Details why you did you did the study (not what you found)
  • Then the discussion interprets your results and places into context with the literature.
    • End with a nice ‘take home’ message in the final paragraph
  • The introduction should describe why you did the study, with enough details to show the nature of the study, any problems encountered and how you have dealt with them, and current literature. You do not need to say what you have found, unless specifically instructed to do so, as you will already include this in the abstract and discussion sections.

Crossing the line

  • Figure legends should be ‘stand alone’
  • The title should be clear and attract attention
    • You need to lure readers to your paper amongst all the others
  • Similarly, the abstract should be very clear with simple messages, clear results and snappy conclusions
  • Figure legends should be ‘stand alone’ i.e. figure and legend must be interpreted without the necessity of searching through the written text.
  • The title should be clear and modest, and be able to attract attention without being unrealistic.
  • Similarly, the abstract should be clear and supportive of the results, with a snappy conclusion.


  • Use Endnote (or similar) to output the references in the correct format
  • But, which references do you cite?
    • High impact factor journals
    • Avoid over citation of yourself
    • Write what you know and then reference the text or you will need to stop every few words to find a paper in the heap on your floor!
  • Endnote is ideal in most cases to generate the references, in the correct format for the specific journal, and with suitable bibliographies. Cite important primary papers where necessary, avoiding over citation of yourself. Write what you know, then reference the text at that point, otherwise you may find progress slow.

A knotty problem

  • Who will be included as authors (and in what order)
  • This can cause some dispute!
  • Some journals have a clear policy
  • Some supervisors or research groups also have a policy (ask)
  • All authors carry full responsibility for the content
  • You may find it difficult to organise who will be included as authors, and in which order they are placed – some journals have a clear policy, as do some research groups or supervisors – you will have to ask.
  • Remember that the author carries full responsibility for the content.

The mechanics of publication

  • Submission
    • You may need to learn how to use an on-line system like ‘Manuscript Central’
  • The decision
    • Rejection. Learn from the referees comments and try again
    • Revision. This is common. Answer the referees questions carefully (maybe generate some more data) and you’ll be OK
    • Immediate acceptance. This is rare!
  • It may be useful to learn how to use an on-line system of submission, such as ‘Manuscript Central’, which a lot of journals use.
  • Submission can be harder than you think!
  • The online submission is not always simple to use – once you have filled in the forms, uploaded your documents, written letters to the editor and sorted out what format your graphics should be, it can take some considerable time.
  • What does the referee think?
  • The online submission is not always simple to use – once you have filled in the forms, uploaded your documents, written letters to the editor and sorted out what format your graphics should be, it can take some considerable time.

The mechanics of publication

  • Submission
    • You may need to learn how to use an on-line system like ‘Manuscript Central’
  • The decision
    • Rejection. Learn from the referees comments and try again
    • Revision. This is common. Answer the referees questions carefully (maybe generate some more data) and you’ll be OK
    • Immediate acceptance. This is rare!
  • Comments to the author are made hopefully to aid with any modifications necessary, and to learn from.

Errors and glitches

  • Check proofs VERY CAREFULLY!
  • Make necessary corrections clearly
  • Answer all the editorial questions
    • You must not re-write the text at this stage
  • Don’t hurry this process or errors will get through!
  • Proofs are sent through once your paper has been accepted. Proofread your document carefully, making any corrections absolutely clear. Do not re-write any of the text at this stage, and answer the editorial questions in full. Take your time with the process, or mistakes will be made – ask your co-authors to read it through.
  • JBC; 77 citations and nobody has ever commented…
  • Mistakes do get published…..
  • It is very easy to make a foolish mistake …. so don’t do it!

Conference abstracts

  • Problems
  • You often need to describe work in progress months before the meeting
  • Acceptance is highly competitive
  • and
  • You want to be accepted as your travel grant depends on giving a presentation!
  • Conference abstracts’ acceptance is highly competitive – often they have to be submitted months in advance of the meeting, and often before any work on your subject is completed. Your travel grant may also depend upon giving a presentation – non acceptance means no airline ticket!

More on abstracts

  • Follow the rules
    • Strict word or (even character) counts etc
  • Make whatever you hope to present sound as good and positive as possible
  • Avoid empty statements like:
    • These data will be discussed
    • This work is still in progress
  • The abstracts have to be assimilated in a very short time, so follow the rules – there are strict limits on word or even character counts.
  • Don’t overstate your conclusions, but sound as positive and as convincing as possible. Don't make empty statements – you need to have a definitive conclusion to the abstract.

Even more….

  • Choose a punchy title
  • Write a brief introduction. Maybe only 2 sentences
  • Very briefly describe the methods
  • Show ‘solid’ data (with statistics if needed)
  • The conclusion should show how you have answered your original question.
  • Choose a positive, punchy title, and keep your introduction short. Don’t describe your methods in any great length and give data with significant values. The abstract should give a convincing positive message, with the original question answered in your conclusion.

Download 20.71 Kb.

Share with your friends:

The database is protected by copyright ©www.sckool.org 2022
send message

    Main page