Introduction

Experts do not always agree

Experts often disagree about what amounts to good writing in science writing, and the judge of what amounts to good and bad writing therefore to some extent lies with the readership and is determined by what they expect.  Still, there are certain guidelines that must be adhered to, conventions that should be followed and trends that evolve over time and need to be observed. These changes and trends have evolved ever more rapidly over the past two decades, and many of the rules that applied 20-30 years ago have now been undone or modified or even reversed in response to “the demands of target audiences, editors and peer-reviewers, especially given the growing number of articles submitted for publication and the strong competition among researchers in pursuit of international recognition and reward” according to Marta. An example of this phenomenon is the issue of active versus passive writing.  While the passive writing used to be the preferred voice in the past, the active is often preferred today, not least in American English scientific writing.

Good writing

Purpose must be clear

Good writing needs to be very clear about its purpose and objectives. This involves that the author of the text is conscious of how the rhetorical purpose shapes text structure at the level of subgenre, move and steps. It also involves writing the text in conformity with given formats and style requirements, both the general ones as well as the research-design specific ones and the particular requirements of the journal to which the paper is submitted.

Terminology must be accurate and consistent

Accuracy is another important feature of good scientific writing. The wording needs to be grammatically correct, concise and precise. Using terms consistently and in the sense they are usually used commands attention to the nuances of meaning and knowledge of current, accepted terminology. For example, in a randomised trial, adverse event is preferred over harmful event. The terminology you use to designate core concepts should be consistent and appropriate, so if, for example, you designate the object of your study participants, do not suddenly call them patients or cases. 

The logic and flow of the text must be clear so that the report of what you did in your research comes across as a strong, compelling story with a clear message based on a logical strain of thought. New findings must be reported and interpreted in the context of findings already published and must be congruent with accepted institutional or regulatory values. The structure of your text should be logical with appropriate headings and subheadings, paragraphs and data displays.

A well-balanced mixture of text and visuals (e.g. figures and tables) should be chosen,in line with the relevant instructions for authors. Finally, data presentation should be of high quality and data should be presented clearly, using tables and figures, as appropriate, and data duplication should be avoided.

The plain language movement

 Both in the UK and in the US, recent years have seen a move towards plain writing in academic, medical and other professional texts. This movement towards plainer language is an attempt to demonstrate the benefits of writing clearly and concisely and in a reader-focused style that avoids wordiness, clichés and jargon. Plain writing furthers target audience comprehension and takes the readers’ knowledge of the subject into account. In short, the plain language movement may be called a recipe for using a logical organisation of your text; common everyday words except for necessary technical texts; we and other personal pronouns; the active voice; and short sentences.

Many scientists believe that they have to write only for colleagues or experts in their own field. Yet, as scientists we often have to communicate the results of our research to a wider audience. This may be difficult as it may imply using the layman’s less efficient, effective and precise vocabulary instead of the more precise technical terminology. Indeed, it may be quite difficult to describe complex phenomena in simple terms. In such situations, resources like the Plain Language Medical Dictionary. 

In their “Instructions to authors”, most journals now encourage the use of plain language. For example, the British Medical Journal (BMJ) recommends that authors choose plain and clear language over academic prose and that they write in the active rather than the passive voice

An advantage of writing in clear, plain language is, of course, that this approach may expand the readership. There is little doubt that more medical and scientific breakthroughs will see the day if more readers are enlightened and inspired by clear, understandable articles. Moreover, scientists who reach a wider audience might be more successful in persuading policy makers to fund their research.

Make an outline

Becoming more product-oriented, the next step is to outline the structure of your paper. For this purpose, you may use the general IMRAD structure or some of the research-design-specific guidelines available from the EQUATOR Network as discussed in the text Reporting guidelines. Requirements for contents and style. Simply break the writing into bite-sized pieces, using the available checklists and move structures.

Outlining is an effective way of communicating your ideas and exchanging thoughts. So at this early stage of your writing, you should seek feedback from peers and supervisors to get their input and to save important time by avoiding spending time on unnecessary writing. This is also the best stage at which to decide to which journal you will later submit your paper.

Outlining is an effective way of communicating your ideas and exchanging thoughts. So at this early stage of your writing, you should seek feedback from peers and supervisors to get their input and to save important time by avoiding spending time on unnecessary writing. This is also the best stage at which to decide to which journal you will later submit your paper.

Decide where to start!

The real process of product-oriented writing begins after you have received enough feedback on your early ideas and outline and have decided on the target journal. You may want to copy your outline into a separate file and then expand on each of the points, elaborating on the details and adding data. It is important that you start with the work you are most familiar with. So if you generated the data, start with methods or results; or start with figures and tables and organise them as if they tell a story! Continue with results, then methods, then discussion/introduction. If you are at a very abstract level, start with concepts and get your definitions in place.

When you create the first draft, do not give in to the temptation of any editing. Searching for a better word or a better phrase invariably slows you down; do not halt to improve your choice of words and sentence structure at this stage. Simply write as much as you can; pour your ideas into the paper and leave revision and editing for later. Writing a paper is not a linear process, but rather an iterative one where you jump back and forth between the sections. Still, for practical reasons, I will now briefly describe each part of the paper, offering some general advice, which will be explained in much greater detail along with text samples in other texts.

Title

The title is the most-frequently-read part of any article, so it evidently must be informative rather than indicative, i.e. it must specify rather than indicate the results or findings of your research. Generally, rather than just indicating an effect, for example, Effect of X therapy on patients with…, you should specify the nature of that effect (reduces), e.g. X therapy reduces Y in patients with …. For practical purposes, start with a working title; you can always change it later! Beware not to make too long titles and avoid being humorous.

Abstract

The abstract is the most read part of an article; indeed, often the only part of a paper that is being read. So, it evidently needs to be crisp and to the point. You therefore need to make several drafts, and it is generally recommended to write the Abstract as the last piece of the paper. There is an abundance of valuable information to be found on the Internet on the process of abstract writing and the structure and contents of abstracts.

Introduction

The best way to structure your Introduction is to follow the three-move approach known as the CARS model. You first hook your reader with a topic sentence and a controlling idea, giving the reader the background for your research, essentially telling them what the problem is, what is already known and why the issue you raise is important. This is referred to as establishing your territory, which must be done meticulously and precisely. You may want to think of this move as one that presents the context in which you place your research, and hence prepares the ground for your later discussion of how you contribute to current research. Next, you establish the niche, telling your audience why the proposed work is needed, i.e. what is missing or wrong. This move outlines the scope of your research problem and enters the scientific dialogue. The final move, occupying the niche, is where you state the purpose of your research in a nutshell or where you raise a hypothesis. Essentially, Introductions zoom in from the general to the specific.

In some disciplines, like biomedicine, you may announce principal findings as part of move three. In other traditions, for example in qualitative papers, you may indicate the structure of the article. Finally, some of the more prestigious journals, like Nature, have their own author resources; see for example the advice given in NatureEducation for the function and structure of the Introduction.

Even if the purpose is essentially the same, viz. to inform the reader about the topic and to persuade the reader that it is worth taking the time to read your paper, the move structures in Introductions differ slightly between qualitative and quantitative studies, different disciplines and different genres. See, for example, a neat overview of the purpose and structure of review paper introductions.

The Introduction should not be long.  Indeed, in most journals, this is a very brief section of about 250 to 600 words, but it may be the most difficult section due to its importance. General problems in introductions are that they get too long and fail to include relevant studies; another problem is that Introductions include too much background or material that should rather be in the methods, results and discussion sections. 

Materials and methods

This section is the “what and how” of the paper. As far as contents of this section is concerned, there is useful general advice in the ICMJE guideline and specific checklists for the most often used research design in specific guidelines, e.g. CONSORT for randomised trials; STROBE for cohort, case-control and cross-sectional studies; STARD for diagnostic accuracy studies; as well as various evaluation tools for qualitative studies, including the SRQR. There is usually very little leeway as to what goes into this section and not, so you are strongly advised to be very observant of the guidelines and to include only that which is necessary to allow evaluation and replication.

Writing the Materials and Methods sections is often a meticulous and time-consuming task requiring extreme accuracy and clarity. You should therefore ask for as much feedback from your peers and colleagues as possible when drafting this section.

Results 

Like with the Materials and Methods section, you can find much valuable advice on what goes into the Results section in existing guidelines. Whether you follow the guidelines or not, a useful approach to writing this section is first to make a set of figures and tables and then arrange them to tell a story. After you have done this, you write the text.

Tables, figures and graphs play an important role in conveying complex information. For general tips on how to design effective tables and figures, you may get some good advice on the editage Insights and Bates College websites. But do remember that medical journals often have their own specific demands for the design of graphics, and strict adherence to these demands is usually a must. Consulting the Author Resources of the target journal early during the drafting of your graphics and following their guidelines strictly will save you much time.

In the traditional paper, the Results section excludes data interpretation, which is left for the Discussion section. The Results section is often the shortest, and most guidelines emphasize the importance of not repeating in words what can be read from the graphics. Indeed, it is advisable to present only those results that go the core of your aim statement and that are essential for your reader to appreciate the significance of your findings.

Discussion  

Research traditions differ and so do the structures of Discussion sections. Moreover, as every paper has its unique results and findings, the Discussion section differs in its length, shape and structure from discipline to discipline. Still, there are certain general tendencies and like with the Introduction, we may use these tendencies, rooted in convention, as a starting point to establish the Discussion’s move structure.

The Discussion is a mirror image of the Introduction. Where the Introduction’s first move is to move from the general background to the specific research topic you are exploring, the Discussion’s two first moves zoom out from a summary of your specific findings to the broader research context. And where the Introduction’s last move is typically to announce the purpose of your research, the last move in the Discussion paves the ground for additional research.

The Discussion is a mirror image of the Introduction. Where the Introduction’s first move is to move from the general background to the specific research topic you are exploring, the Discussion’s two first moves zoom out from a summary of your specific findings to the broader research context. And where the Introduction’s last move is typically to announce the purpose of your research, the last move in the Discussion paves the ground for additional research.

Thus, in the first move in the Discussion, you mention the major findings of your study, you explain the meaning and importance of your findings and you consider possible explanations for your findings. Relating your research to extant knowledge, the second move compares and contrasts your findings with those of other studies; you explain any unexpected findings and discrepancies between what you found and what others found; and you mention the limitations, strengths and weaknesses of your study. In the final, third move, the closing move, you may summarise your findings and your contribution to existing research, suggest possible applications and make recommendations for future research. It goes without saying that any specific journal requirements for a particular structure, e.g. that of the British Medical Journal, take precedence over these conventions.

One of the more frequent mistakes when writing up Discussion sections is to assume that the importance of the findings is obvious to others. Explaining the results and their significance to the research community in the Discussion is just as important as stating the aim of the research in the Introduction. Other common pitfalls include repeating some of the introduction, excessive repetition of results, discussion of issues that may be interesting, but are irrelevant to the aim or the results, presentation of new data and, indeed, failure to properly discuss the results.

Refine and proofread your paper

Once you have finished the draft version of your paper, it is time to refine it. This can be done at both a macrostructure and a microstructure level. At the macro-level, revision includes the revision of the organisation, the contents and the flow of the text. In this part of the revision, try to avoid editing your paper for sentence structure and grammar; focus only on the flow of your ideas and logic of your presentation. At the micro-level, revision includes individual words, sentence structure, grammar, punctuation and spelling, among others.

Ideally, you should limit yourself to working on small sections of about five pages at a time. After these short sections, your eyes get used to your writing and your efficiency in spotting problems decreases.

Academic writing 

When writing up your manuscript for publication, it is important not only to be product-oriented in the sense discussed above, but also to adopt the proper, formal tone that characterises written academic English. In this very short introduction to written academic English, I will focus on formality, impersonal style and conciseness, but you may find much other valuable advice and writing tips on formal writing on the Internet.

Formal writing 

At the level of choice of words, you should choose the more formal variant where synonyms exist (e.g. we got a lot of advice is rather informal, whereas we received much advice is more formal). Most words, irrespective of which word class they are (verbs, nouns, adjectives, etc.), have (near) synonyms. And in academic writing, the formal variant is the one you should go for. In case of verbs, the informal variants often consist of a verb plus a preposition (e.g. The doctor looked at the patient’s history to find out which tests had previously been done. Adapting this sentence to the more formal style is simply a question of finding equivalent single verbs conveying the same meaning as the verb + the preposition (e.g. The doctor examined the patient’s history to establish which tests had previously been done).

Impersonal writing 

Formality exists at many levels of text; and one aspect of formal style is the focus on the issue rather than the agent or the doer. Consider the following example of an informal text: When I think about the situation in surgery wards with many staff leaving, it’s hard not to worry about how many doctors will be around to treat patients in the future.

Now consider the more formal variant: If we consider the situation in surgery wards, with increasingly low staff retention rates, there are concerns about the capacity of hospitals to maintain adequate doctor-to-patient ratios. Here you note that the first-person singular pronoun I has been replaced by the slightly more formal first-person plural pronoun we; the contraction of it and is to it’s has been abandoned in favour of there are, and the text has also become more nominal , i.e. it uses nouns with action meaning, e.g. low staff retention rates and doctor-to-patient ratios) and less verb-driven (staff leaving and how many doctors will be around to treat patients).

Conciseness

Another feature of formal, academic writing is that it is concise. The goal of concise writing is to use the most effective words and sentences. Concise writing does not always have the fewest words, but it always uses the strongest ones; and in concise writing you seek to avoid weak or unnecessary words that can be deleted or replaced. It is very much about brevity; and there is much good advice on how to achieve brevity at Purdue University Online and Webster Grammar Online, among others. 

In English academic writing, having a clear paragraph structure is important to the overall perception of the conciseness of a text. A paragraph should raise one topic only and should do so in an opening or topic sentence with a controlling idea. The body of the text should have a clear, logical structure (coherence) and any transitions (cohesion) between sentences must be clearly marked. Every paragraph should ideally end with a closing, summarising sentence.

At the sentence level, conciseness may be obtained by following the seven steps in the so-called Paramedic method: 1) Circle the prepositions (of, in, about, for, onto, into); 2) Draw a box around the “is” and “has” verb forms; 3) Ask, “Where is the action?; 4) Change the “action” into a simple verb; 5) Move the ‘doer’ into the subject position to make an active sentence, or – better in medical discourse – maintain the passive voice but show indicate the ‘doer’ with a prepositional phrase starting with by; 6) Eliminate any unnecessary slow wind-ups (e.g. according to; in the first section); and 7) Eliminate any redundancies (e.g. repetition in the sentence). In the example, adapted from Purdue Online Writing Lab the sentence In this sentence is a demonstration of the use of good style in the writing of a paper is changed into This sentence (the ‘doer’) demonstrates (active verb) good style in writing which is shorter and much more concise. You may find a neat explanation of the Paramedic method on YouTube.

Now, let us look at a more complex example, using the Paramedic method to collapse two sentences into one: The relationship between iodine intake and thyroid volume has been described in a number of previous studies performed in areas with iodine deficiency. Most of these studies are cross-sectional and they found an inverse relationship between low iodine intake and enlarged thyroid volume, as well as a lower frequency of thyroid enlargement after iodine fortification. These two sentences can be reduced into one using the Paramedic method: Most prior studies (doer) of iodine deficiency areas used (simple verb) a cross-sectional design and reported (simple verb) an inverse relation between a low iodine intake and an enlarged thyroid volume, as well as a lower post-iodine fortification thyroid enlargement frequency.