• It is very important to decide who the authors will be, in what order, and what responsibilities each will have.
• Ideally this should be decided before the experiments start. If not, meet now with your supervisor.
• Normally if this is your experiment, your name comes first, the most important place in the list.
• The second most important place in the list is the last. Usually the lab head occupies this slot.
• The remaining authors are listed in the middle, ordered in terms of their contribution to this paper. This should not include colleagues who have only provided advice and feedback.

• Reassemble a subset of your data into a good simple story. Resist the urge to include every observation and idea.
• Draft the figures and tables that best explain. Some readers begin by scanning the figures. The figures, with the legends, should provide a self explanatory overview of your data.
• Prioritize and order them.
• Exclude things that detract from the main story line.

To select which ones to use, you can work backwards.

• Often one begins with an example of the data in as raw form as possible and ends with a figure which compares a parameter across subjects and conditions.
• Decide first what this last figure will look like. 
• Then decide what figures you need to explain how you got from the raw data to this summary figure.

Choosing the best example.

• Select the data which most clearly illustrates the result you want to emphasize.
• Sometimes the best is not representative. If so, say so, and support your argument with appropriate statistics.

The important thing is to explain things in the right order.

• What that right order is varies; e.g. from simple to complex, some logical order (if A true then B or C is the next thing to test).
• If possible, start and end with your most interesting data, and put the less interesting stuff in the middle.
• The order in which you did the experiments is often not the best order for the paper

• First state what the point of the figure is.
• Then explain each and every element of the figure.

• Often the data looks better in your memory than on paper. That is why it is important to keep the real data there in front of you.
• Take the reader slowly through each part of the figure. Do this in point form first.
• Take the points and group them into paragraphs.
• Order the points within each paragraph.

The results section is the most important part of a paper.

• Your conclusions may soon be superseded, by yourself or other scientists. Science moves on.
• However if the experiment was done right and is thus an honest reflection of what exists, your data should remain true.
• Present the data as fully as possible, including stuff that at the moment does not quite make sense.
• On the other hand, you have to cope with the reality that space is limited.

There is nothing duller than a long listing of data with no explanation as to why it might be interesting.

• Motivate the reader to read on. Provide brief rationales.

• Include enough detail so the someone can repeat the dish
• The results are very dependent on the particular combination ingredients and procedures.
• It is important for the reader interpret the results in light of the context in which they were obtained.
• Explain why you selected particular ingredients over others.

1. Summarize what you found.
2. Explain how your finding relate to those of others.
3. Explain the implications of your findings.

1. What your question is.
2. Why the reader should find this question interesting.
3. What is currently known on this question.
   • Cite 1) the classic papers 2) the most pertinent to the point you are making and 3) the latest papers. If you are not sure what these are, look at a few recent papers and see who they cite.
4. What you found.

Notice that you will have a much better idea of what these are, after having written the results and discussion.

The introduction is a very important part of the paper because reader will continue on to the other parts only if you have convinced the reader here that this would be worth while. Often it is the only part that is read carefully. Take the time to make it something your reader will not only enjoy reading but remember. To do this you must come up with a compelling story line for why you did this study.

• More scientists are using computer searches to keep up with the literature. These computer searches include only the abstract. The abstract may be to only thing many people read.
• Your reviewer's first impression of your paper will be based on the quality of the abstract.
• Keep to the stated word limit. Abstracts are published in electronic form that have a strict word limit. PubMed only includes the first 400 words.
• An abstract should include
  • a very short summary of what the question (goal, purpose, hypothesis) is (present tense)
  • explain key procedures (past tense)
  • what you found, the key 1 or 2 things (past tense)
  • why it may be important/significant (present tense).
• Avoid
  • hype and speculation
  • abbreviations
  • extensive statistics

• Important because this, the title, the abstract will by used by computer searches to scan the literature.
• Make sure you use terms that are common to data bases.

• Besides the above, the title should be short (about 10 words), interesting, & describe what you found. Remember that your potential readers will be scanning through long lists of titles. Your title should be designed to attract their attention!
• Consider the title carefully. It will become a permanent part of your CV.

• Include only those that pertain to the question at hand.
• But make sure that you include them all. This is important because these authors may be reviewing you paper.
• Look through the reference of related papers and include any that are relevant. Reread them make sure that you correctly state what they found!

• Say why you think readers of this particular journal might be interested in your results (cite previous related papers published in this journal).
• Suggest possible referees (respected scientists in the field who you believe will be positive, but not your colleagues).
• If there is someone with whom you strongly disagree, perhaps suggest that this person not be selected as a referee. Phone the editor if you wish this to remain private.

• Who else helped and what did they do?
• How was the study funded?

• Who will correct the galley proofs and send out reprints?

• Be prepared for lots of revisions. 10 to 15 is not uncommon.

• Often each has different ideas on how the paper should be written. The first author should make the final decision.
• The first author should send a the draft to each author asking them to write in changes. All the authors should then meet and iron out the changes page for page.
• If you circulate copies electronically, make sure everyone uses the revision feature. Otherwise you will have trouble spotting all the changes.

Copyright © 1995
Tutis Vilis
Department of Physiology and Pharmacology
University of Western Ontario
London Ontario Canada

Created 28 Sept 1995
Last updated 10 January 2007
Comments welcome: tutis.vilis@schulich.uwo.ca