How to Write Papers

By Jamie Bartram, with inspiration from Fran DiGiano

Choosing a Journal
Reflect on which journal you want to target ideally before writing and definitely early on. You may want the highest impact factor but read your target journal’s editorial policy (e.g. On types of papers accepted and their target subject matters); look at a few recent numbers to see what they are carrying and be honest with yourself about the quality, originality and relevance of your work.

Take careful note of your target journal’s requirements and take a look at some articles in it for ideas on style and typical subtitles. Read the journal’s style guidelines and follow them.

Knowing the maximum word length for your journal/paper category will assist in drafting. For example, there is little sense in writing a 3000 word introduction for a paper with a word limit of 5000 words.

General Writing Tips
I believe that the most important ‘style’ issue is clarity; the objective is to communicate effectively with readers, not to baffle them with jargon and complex multi-clause sentences. When reviewing a draft of my own work I try to think, “Could I express that more simply or clearly?” And if the answer is yes, then to try to do so. I am not always successful!

Most journals work to the typical modern academic style: introduction, methods, results, discussion, and conclusion sections. The discussion and conclusion sections may be separate or combined.

  • Introduction
    The introduction provides context to the reader – typically explaining the issue and/or physical context of the work. The key purpose is to ‘position’ the reader so that, as they move into the paper, they understand your knowledge starting point, including the question you were trying to explore and anything surrounding the study that is of relevance to it. For a reviewer of your paper its where they determine whether you have an understanding of the relevant literature and issues. Typically begin with a broad description of the problem and end with a description of objectives. I would counsel that you aim to ensure that your introduction does three things: beginning with a broad description of the problem it guides the reader through a sequence of evidence that builds up to and makes self-evident your research question; and sets the scene such that the reader understands the general context of what follows. The latter may include issues of overall approach and method.
  • Methods
    The usual rule of thumb with a method is that it should provide the level of detail that would enable another person with an appropriate training/education to replicate it correctly – and implicitly and in doing so, to get similar results. Where a method is described in detail elsewhere in the public domain this should be referred to, but any refinements or variants need to be stated.
  • Results
    Results should present results but not discuss them. Information in figures and tables should not be linearly repeated in words rather words should complement figures and tables. It is often tempting to start to discuss results and this is best kept for the discussion. In some disciplines it is more common that results and discussion sections are merged. When this is done it is important that he reader can clearly distinguish the findings of the work from their discussion and comparison with the work of others
  • Discussion
    A discussion typically focuses on: whether you are confident in the results; and on interpreting them in the context of published literature. Sometimes during drafting a discussion evolves into a whole new introduction and when this happens it is helpful to reflect on what lives in the introduction and what lives in the discussion. There should not be unnecessary duplication between them.
  • Conclusion
    Sometimes discussion and conclusion are merged. Whether merged or separated, the conclusion section should be crisp and clear, not a summary of results. Conclusions are also where the loop back to the study objective and introduction is closed. The purpose is typically to succinctly identify what new is known because of the study and to state its significance and implications in the context of the wider literature and situation.
  • Abstract
    Finally, take care with preparing your abstract. It will be read by many more people than will read the paper as a whole. Even if your journal does not require submission in structured form it may be useful to draft your abstract in this way.

People are often reluctant (and incorrectly so) to publish a negative finding. I was once taught that “a negative result is a positive conclusion,” and it is true! If you have high confidence in a negative finding then this is valuable information.

Reviewing and Editing
When you think you are finished, if possible put the manuscript to one side for a while and then review it critically yourself. Delete superfluous material and duplication, ensure that statements are sound and remain sound taken out of context, and ensure that journal guidelines and general writing guidelines have been followed. Problems frequently identified on review of a first complete draft include:

  • Recognizing that there is material not tightly or clearly related to the flow of logic. In most cases this is a valuable opportunity to shorten the text and tighten the reasoning.
  • Shifting or imprecise terminology. Introduce the terms you use and give preference to established definitions and uses. Use the minimum number of necessary terms and do not change the way you use them as you progress through the paper. Experience confirms great scope for ambiguity even among terms like coliforms, total coliforms, faecal coliforms and thermotolerant coliforms when these are inadequately distinguished in a single paper.
  • Repeating table content in text, failure to refer to a table or figure; and poor comprehensibility of tables or figures in conjunction with their title.
  • Superfluous words – often either adjectives or introductory phrases (‘It is widely recognized that …’ and similar). A rule of thumb is that if words can be deleted without detracting from your meaning then those words are better deleted. Less is more!
  • Ambiguity. Among the most frequent is use of the term ‘negative’ in describing an association when what is meant is adverse (ie a negative association is a statistical relationship which can be a good or bad thing for, for example, human health)
  • Mixing tenses
  • Mixing persons. Use the third person.
  • Informal terminology and casual abbreviations. Writing for a scientific journal requires formality
  • Exaggerated or advocacy-driven terminology. Exercise neutrality and objectivity.
  • Misuse of units, rounding, and poor labeling of figures and axes.
  • Use of secondary rather than primary references.
  • And while it may sound obvious, just using the wrong words. If you say that you ‘describe’ something then describe it, do not give a list; if you say you summarize then do not provide detail; do not say ‘detailed’ when you mean listed and so on. Similarly index and indicator are not synonyms; and neither are estimate, determine and measure. There are many examples.

It can be useful, when you have a finished draft, to ask an informal reviewer to comment on it. They would normally be someone you know well enough to ask them for such assistance, familiar enough with paper writing to critically comment on it as a paper and familiar enough with the subject matter and/or method to comment knowledgeably. It is inappropriate to use such a reviewer on a crude draft. For example you should have checked your manuscript for the above-mentioned common problems and against general style guidance beforehand. Otherwise, you risk getting editorial rather than substantive feedback when you are seeking the latter. Such a person is normally thanked in the acknowledgements section.

The act of submitting a manuscript is becoming increasingly onerous with more attachments and supporting documents required. Persist and do what is required. If there is opportunity to provide a covering letter to the editor then it is an opportunity for you to make the case that the journal should carry your paper. Quality, originality and relevance are likely to appear in the editor’s list of concerns. Recall, too, that editors are busy people for whom editing is a demanding and unpaid sideline. So be clear, precise, and compelling!