Writing (a paper)
Last updated: 17 August 2025
Words matter. They can kill or lead people to die for a cause. Cultures are based on ideas developed and expressed using words. Writing is the foundation of most human societies. When people ignore, wilfully misuse, or ignore the true meaning of words, chaos and lawlessness follow—so, learning to use words well, whether speaking or writing, is vital for success.
This page focuses on writing for publication in scientific journals and textbooks, such as Clinical Rehabilitation. Still, the main messages apply to all writing, such as applying for research grants, proposing a service development, or writing to patients about your assessment and advice.
The advice is based on my own experience over about 50 years (I once edited the medical school gazette and even helped with a school newspaper). It is also based on 27 years as an editor of a journal, over which time I read at least 11200 submitted papers, probably many more. When giving talks on the topic, I usually use the subtitle, “How to avoid immediate rejection of your paper, because the lack of care or interest that many authors demonstrate suggests they do not want their paper to be accepted.
The two main messages I give are, first, know what you want to say, and say no more than is needed to get your message across and, second, at all points, consider your reader – the editor, the reviewer, and, if you are lucky, a reader who wants to read your paper.
Table of Contents
Introduction
I failed the Use of English examination at school (I passed on the second attempt), yet I now edit a journal and write extensively. Writing is a skill that must be learned; it is not (for most people) a naturally present ability.
Like everything in rehabilitation (and in life), with effort and practice, you may improve. But, like most car drivers, who typically assume they have above-average driving abilities, many authors believe they can write well.
Feedback is crucial for learning. I was taught well when about 21 years old, learning from being given excellent (but quite strong) feedback with advice. My first article, for a newspaper, had every word, sentence, and paragraph changed by my wife! She continued to give critical, constructive comments, and eventually, I improved.
Writing is hard work, and few people realise how hard it is. Some of my most read or cited papers went through 10-30 different drafts, and some took several years. Writing should force you to review and reconsider your ideas.
My experience as an editor (and as a reader of many work-related documents) is that most people write badly, but, like most car drivers, they believe they have above-average ability. Indeed, I also think I am good – until I see the reviews of my submitted papers.
I will highlight the significant weaknesses I notice when reading. I will mainly review the problems observed in papers submitted to Clinical Rehabilitation and other documents I read.
Think of your reader.
The most important single piece of advice is to think of your reader. At all times, consider who will be reading what you produce. This includes, obviously, the actual content, but many other essential matters such as layout, font size, line spacing, clarity of figures and tables, removing tracked changes and comments, etc.
Questions to ask yourself are:
- who will be reading this?
- Ans. Usually, a wide variety of people.
- what will they be looking for?
- Ans. Often quite different things.
- will they find what they want quickly?
- Ans. Only if you structure everything logically can they do so.
- does my paper have a clear, unifying main message?
- Ans. It will only do so if you remain focused on the point; your article is not a nineteenth-century Russian novel.
Put yourself in the position of potential readers. There are several ways to emulate the reader:
- ask a friend or colleague unfamiliar with the research or the article’s content to read it and give feedback. They will represent the reader.
- read the paper out aloud to yourself. You will soon pick up mistakes and poor writing. (I picked up this idea from Twitter.)
- leave the paper completely untouched for one week before submission. Then read it anew; you will notice many areas needing improvement
Stick to your message.
When writing any document, the most critical matters to consider are whether you have a single, vital message and whether everything in the piece is crucial to conveying your message. Your central idea is your theme, and everything within the document must relate to the overall message. If something does not relate to your article’s core message, remove it because it will confuse the reader.
Therefore, you must always:
- have a central message for the document, something you can encapsulate in no more than 50 words.
- Often this will evolve as you write, because your understanding changes, and this changes the message.
- ensure that everything you write relates to that key, thematic message.
At every point, you must consider how this sentence, paragraph, page, or chapter relates to my theme. Is what it is saying relevant to the larger message? Because writing usually leads to changes in the message, you must review the whole document when you have finished, to remove any irrelevant detours.
One advantage of writing on a computer screen is that you can only see a limited part of the document. When you come back to read and edit it, ask yourself if you can understand the sentence or paragraph without having to go back and discover how it relates to your central idea.
To help yourself, imagine that your paper has hit the headlines (none of mine ever have!), and you are on national television. You have probably 60 seconds. The interviewer asks, “What is your paper on XXX about?“. In two sentences, you need to say whatever you think is the paper’s most important message—no more.
Use or provide a structure.
Scientific journals have evolved a standard structure, often abbreviated to IMRAD:
- Introduction,
- Methods,
- Results,
- And
- Discussion
For a standard scientific paper, the structure is generally appropriate. Some journals suggest additional structures. If the structure suits your paper, use it.
Also, consider structuring the content of each section so that the reader can quickly find the detail they want. For example, an introduction can pose a question you are considering, review what is already certain, aim to identify the gap you are interested in, and specify your hypotheses or questions.
The precise structure of the methods and results sections depends on your design. For most standard designs, such as randomised trials and systematic reviews, there are now well-established guidelines on how to structure the paper. These are easily found on the EQUATOR website. You should always use the guidelines if they exist.
When writing other types of paper, search for guidance on structure. For example, reports on complex rehabilitation protocols are best described using the TIDiER framework. If you cannot find a recommended structure, then give significant thought to how you structure the paper.
Ultimately, you must remember that:
- The reader needs to know where she has been, where she is, and where she is going at all times.
- Structure provides the reader with a mental map, making them feel confident and enabling them to find what they need quickly.
Moreover, all aspects of writing must have structure. Letters are formed into words, carrying meaning. Groups of words form sentences structured (sometimes!) by grammatical rules. Sentences are grouped into paragraphs, etc. Each level of the structure holds a meta-message: this word means ….; this sentence means ….; this paragraph covers …; and so on.
The greatest weakness of most authors is an inability to use paragraphs appropriately, coupled with an inability to structure a group of paragraphs. A paragraph concerns a single idea.
For example, a single paragraph in the methods section of a paper submitted to Clinical Rehabilitation might contain, in this order, sentences about the random allocation of patients; recruitment of patients; collecting baseline data; the treatments given; blinding (masking) of the people collecting data; and finishing with a sentence about trial registration. (This example is real.) A reader will struggle to find the information.
This weakness is exacerbated by a tendency to use subheadings to give structure. Often the subheading contains material that is unrelated to the subheading. Sometimes the heading is followed by one very long paragraph. The subheadings are rarely in a logical order. They force people to put items in an illogical place. Subheadings should be like chapters in a book, indicating a fresh start on something different. As you are not writing a book, you will rarely need subheadings.
Any section covering more than one topic must have a logical framework to assist the author and help the reader. It ensures that the author covers all vital issues, assuming the framework is complete; it ensures the reader can quickly find what she wants.
Standard sections
Introduction:
This should usually cover the problem being addressed, a summary of what is known, identifying a gap, and showing how this study will fill that gap.
Method:
This should usually start with bureaucratic matters (registration of the study, ethics, etc). Then, for most clinical studies, the easiest framework follows a patient over time: how will or did a patient enter, flow through, and leave the project, and what happened at each point?
Results:
This presents data, and a different logical structure is needed. In most clinical papers, the data typically includes patient flow in numbers, baseline data, important outcome data, secondary data, and other analyses.
Discussion:
This can be more fluid. It requires the author to take an overview and consider their message. One general structure is an overview of results, placing results in a broader context, considering limitations and cautions, and finishing by discussing the implications for theoretical understanding, clinical practice, or future research.
No abbreviations or jargon. NONE!
One day, at a hospital executive board meeting, the conversation about a paper had been progressing for about five minutes when I asked what an abbreviation (acronym), which was reasonably central to the report, stood for. No-one knew! But this was the first time anyone else had been prepared to ask. And the document’s authors must have presumed that everyone would know, as there was no explanation.
Please realise that:
- all abbreviations and all jargon reduce readability and risk profound misunderstanding
- Articles are judged on the standard of writing and clarity, not on word count
Abbreviations hinder understanding and readability. See here, here, and here. Please do not use them. The only exceptions I allow in clinical research papers are for standard statistical measures and processes. The excuses given by the authors include the following:
- abbreviations save space (probably untrue, and anyway, writing clearly and succinctly will save more space)
- everyone knows what it means (rarely true)
- it makes reading quicker (undoubtedly false)
- other journals use them (when was following the crowd ever proof of being correct?)
- a paper you published in 2018 had them (probably true; some authors just do not read my requests, and I give up!).
Acronyms are made from the initial letters of a phrase, such as NASA for the National Aeronautics and Space Administration. (Did you know that?) Acronyms are used for large studies. They may be helpful when searching for all papers related to the study or when searching for the website. Often, the acronym carries no information; it is chosen as it sounds neat. An acronym may be used provided it is explicitly expanded in the abstract and when first used in the text.
Jargon is more insidious. Jargon is a form of exclusive, in-house language that serves as much to exclude others as it does to improve communication. Whoever talks about their ‘upper limb’ except when acting like a professional? Who would say to a friend, “My lower limb hurts”?
Both profession-specific and rehabilitation jargon exist, and both should be avoided. Rehabilitation is a multiprofessional activity, and most jargon is limited to one or two professions, so communication across the whole field of rehabilitation is impaired. Moreover, patients and people outside healthcare may read articles, especially now that Open Access is increasing and Google indexes scientific articles and recommends them to everyone. Hence, you must use plain words and explain any unusual terms.
Instead, write clear prose. Avoid over-complicated sentences. Keep it short and never use three words when one suffices. If you are still determining whether something is necessary, remove it. Remember, “If anything at all, perfection is finally attained not when there is no longer anything to add, but when there is no longer anything to take away”. Antoine de Saint-Exupery. (Thank you, Dr James Cauraugh.)
Respect your readers
To show your respect for the reader,
- present your paper carefully to avoid a reputation for a slapdash approach
- read journal guidance and articles to learn how submissions are expected to look in the chosen journal.
When submitting a scientific article for publication, you should respect the editor and reviewers. Start by addressing any letter to the journal editor you are offering the article to; many authors recycle a letter written to another journal without changing it. Continue by ensuring that your submitted article is well written, well presented, follows journal guidance on layout, etc., and is free of obvious errors.
Some journals employ someone to check all this before the editor sees an article. Most do not. It shows a lack of respect to present an article that does not adhere to the journal’s guidance, and, as an editor, I am very wary of authors who submit papers full of apparent errors. If the author cannot be bothered to check the quality and accuracy of the submitted article, why should I trust the validity of their reported methods and results? At the end of the submission process, there is always a chance to review and check the whole paper. Use it.
Even more importantly, the review process does not and cannot check for errors, mistakes (or fraud), and all the named authors have complete responsibility for everything written and presented. It behoves anyone designated as an author to satisfy him- or herself that the paper is high quality and accurate. If any malfeasance is detected, all authors are liable. Beware.
Learning to write.
In seven words, you should:
- Practice, practice, practice, and
- Read, read, read.
Learning to write well is essential. The pen, or now the word processor, is indeed mightier than the sword. People will die willingly to defend or promulgate an idea, and ideas ultimately come from someone’s writing. Moreover, whoever writes the history of events determines the future perception of what happened and decisions and actions. ‘Fake News’ is more convincing if accompanied by an image of a printed newspaper or document.
More prosaically, whoever writes the minutes of a meeting often determines what people believe was said and decided.
It distresses me that most people seem unable to write. Teaching someone to write well is not central to learning about research. Yet, researchers depend upon writing applications to obtain grants, writing papers that are published and cited, and writing textbooks. A researcher’s status and progression ultimately depend on publications, which must be written.
Only a tiny proportion of grant applications succeed, and only a small proportion of papers submitted to higher-impact journals are accepted. Writing well will significantly improve the chances of success in both spheres. As a reviewer, I judge that someone who can write clearly can also think clearly, and that poor writing may indicate poor research skills.
Learning to write well requires hard work, but it pays. Most importantly, the effort needed to write clearly greatly improves your ability to analyse data and generate hypotheses.
As in every other matter, practice improves ability, and education is much more effective when the person receives feedback on performance. Failure to gain a grant or have a paper accepted is not effective feedback. There is a significant delay, and the connection between the quality of writing and feedback could be more robust, as there are many other reasons for failure. Poor writing will rarely be given as a reason, even though it may be the primary reason. Effective feedback needs to be timely, to point out where improvements are needed and why, and then to suggest ways to improve, with explanations.
There are other ways to learn, some quite enjoyable. When reading a book, article, or other document, consider how easy it is to read and understand, as well as how much it attracts and maintains your attention. Then analyse what it is that helps.
A good paper.
I have summarised the written qualities of a good paper in the figure below. It must also be based on research or evidence, be ethically sound, and adhere to more general standards. The figure can be downloaded.
Further resources.
There are many resources available, and I will list a selection.
- The Equator Network, an excellent general resource, has many guidelines on how to report studies using specific designs, and a specific section on writing for publication.
- The Equator Network also jointly produced Good Reports.
- I wrote detailed guidance for Clinical Rehabilitation, which can be downloaded here.
- The difficulties arising from abbreviations are discussed in the following papers:
- Other papers on writing for publication include:
- Computer applications such as Grammarly, which I use, and Microsoft CoPilot help improve your writing, but every change must be:
- checked to ensure it maintains your ideas correctly, and
- used to improve your personal writing style.
Conclusion
Writing is a crucial skill for any professional. Sadly, few people are good at it, and healthcare organisations rarely provide education and training in writing. Although I may not be excellent, I have learned a great deal over the last 60 years, particularly after failing the Use of English examination at school. I am proof that hard work, excellent personal, critical and constructive feedback, and continual practice with reflection on failures does pay off. You too can learn. Active reading, which means reflecting on how the piece you are reading is constructed, how the words are used, and how it attracts you and maintains your interest, will help you. Asking for constructive feedback from friends and colleagues – and even family members – is a powerful method for improving your skills. Learning from rejection of submitted articles is essential, remembering that 80% to 90% of all articles submitted to most journals are rejected.