A Short Guide on Scientific Writing
Scientific writing is an art and skill that must be learned by repetitive practice of reading, writing, and revising. A well written manuscript not only has better chances of being published, but also goes a long way in increasing the potential impact of your work within the scientific community.
The task of converting scientific results into a peer-reviewed publication can be daunting, both for native and no-native English speakers alike. The following style guide contains a list of simple recommendations and conventions that may help you avoid some of the more common pitfalls when writing a scientific text.
Why Scientific Writing?
Before embarking on your first manuscript, you need to understand the purpose of scientific writing and your target audience. Communicating one's research results to fellow scientists is a vital component of the scientific process, and the way scientists do this is through peer reviewed articles, conference presentations, and reports. By publishing your results, others can build upon your work (just like you have built on the work of others) and it can become part of the body of scientific knowledge.
Plagiarism and Scientific Misconduct
The mounting pressure on scientists to publish (or perish) may tempt some individuals to "recycle" some of their own texts or "borrow" from others. Irrespective of whether this occurs intentionally or by accident, repetitions (either literal or in a slightly altered form) of already existing work is classified as "plagiarism" and considered a serious form of academic misconduct. An increasing number of journals employ automated systems to detect plagiarism and you may inadvertently end up on some internal black list. In order to avoid being accused of plagiarism, never literally copy any phrases or paragraphs from a published source (article, book, website, or report) and if you must paraphrase others, always provide the proper reference to the original source from where you derived your insights and knowledge.
- Write in your own words and if you paraphrase others or if your work builds on existing ideas, always provide the proper reference
- Do not literally copy entire phrases or paragraphs from existing texts
Structuring Your Content
The IMRaD Structure
Most journals require that you follow the so-called IMRaD structure. IMRaD stands for: Introduction, Methods, Results, and Discussion. This structure is meant to reflect the scientific process of discovery, although in reality this process is of course by no means as linear as suggested by the well structured logical flow in a manuscript. Each IMRaD element will now be discussed in a little more detail:
The introduction to your manuscript sets the stage for your research and broadly serves three purposes:
- To introduce the background of your study (through a short literature review)
- To name the research question, the aim of your study, or the hypothesis
- To outline the approach taken to tackle the research question
The literature review alllows you to demonstrate that you are aware of all the relevant research in your field and to place your own work within this context. You can exlpain what motivated your research and why your topic is relevant and deserves to be in the public domain.
An important part of the introduction is the last paragraph, where you introduce the research question with a phrase like “The aim of the present study is to determine …/test the hypothesis of … ”. This paragraph ends with a sentence like: “We approach this problem through … (modelling/field sampling/a particular technique or method/ etc)”. Do not explain the entire method, merely provide the necessary keywords.
By adhering to this convention, a reader can quickly skip to the last paragraph of your introduction and decide whether your paper is relevant to them or not.
- Provide a clear and comprehensive yet concise literature review and motivation
- Clearly formulate what you are doing (your research questions) and why (motivation and relevance)
- Clearly indicate which part of the research questions have already been addressed by others and how you intend to complement/change the existing knowledge
- Do not deviate from the IMRAD structure unless specifically requested by the target journal
- Do not make this section too long as your reader may lose interest otherwise
In this section you describe the study area, the laboratory or numerical procedures used for the analyses, in short, every part of the protocol you followed from collecting the samples to obtaining the final data point. This serves two purposes:
- If you are using a new method or if your results turn out to be surprising or controversial, a meticulous description of your method will allow others to repeat your steps exactly and thereby either confirm or disproof your findings
- By demonstrating that you have followed best practises in your methodology, you can protect yourself from unwarranted criticism
- Only describe the method you used
- Do not discuss other methods or compare your method to others (this should go into the Discussion)
The Results Section
Here you report the findings of your work. You should maintain a neutral and factual tone and avoid any kind of exaggeration or qualitative adjectives like “great” or “fantastic” etc. Do not bury your results in the main text but present them prominently in figures and tables, such that someone only browsing through your manuscript can quickly see what you found.
- Stay factual, concise, and precise in your formulations
- Only describe relevant and/or new findings
- Only describe results from your present work (not from previous work or other authors)
- Use figures and tables to present your main findings
- Do not include any personal (subjective) information (e.g., about how you felt during the field work, etc)
- Do not repeat the methods
- Do not discuss the merits of your results (this is for the discussion)
- Do not bury your results in the main text but display them prominently
Start this section with a short (2-3 sentences max) summary of your main findings from the results section. This will set the scene and allow readers who only skipped through the results to understand what you are talking about here. In this section you critically discuss your results and should:
- Mention possible limitations, flaws, uncertainties of your study design/method (e.g., you could not collect all the data you needed or the accuracy of your machine/model was limiting)
- Identify your sources of errors and quantify them
- Place your findings within the wider scientific context (e.g., how do your results compare to previous studies? Do your results contradict/confirm previous findings and what are the consequences?)
- Suggest possible remedies for the identified limitations/errors and provide suggestions for future work
- Be honest and find a healthy level of self-criticism, i.e., do not demolish your own study, but identify the real and important weaknesses
- Clearly state what your results mean (for your field, the public, or other stakeholders)
- State whether you have answered your research question
- Avoid presenting any new results here, unless it is necessary to prove a point (when you compare your result to previous work for example)
The Conclusion (optional)
Some journals allow/require a separate section with conclusions. If you need to provide such a section, use this to provide a concise (2-3 sentence) summary of your work and explain how (if) you have answered your initial research question and what this means in the broader context. Do not repeat points from the discussion but rather move the relevant points from the discussion to the conclusion.
Other Structural Elements
The title should reflect the manuscript's content. It should be concise and enticing but remain factual. Avoid redundant phrases such as “A study of …” or starting with an article in general.
This is arguably one of the most important parts of your manuscript. Not only will the abstract be read by the journal editor during the initial assessment of your manuscript, but the abstract is also typically freely available on the web, even if the remainder of the article is behind a pay wall.
The abstract should be able to stand on its own (i.e., avoid citations and explain all abbreviations and acronyms) and contain the same IMRaD structure as the remaining manuscript, only condensed to about 250 words. Briefly (one sentence each): summarise the context, outline the research aim or question, describe the research approach (methods and materials), and identify the main finding(s) and conclusion(s) – in this order. Inform the reader about the highlights of your work and entice them to continue reading.
Your reference list contains all the references you cited in the main text. It should neither be too large (unless you are writing a review) nor too short. Find a healthy middle-ground by citing only the most relevant work (e.g., the publication where a new method was first introduced and the most important revisions - as a rule of thumb, try not to exceed 3 references per topic). Using phrases like “and references therein” can considerably reduce your reference list.
Different journals require not only different styles for citing (e.g.; , or Smith et al. (2010), or as a superscript1, etc.) but also for formatting your reference list (e.g., "Smith, MC" vs "M.C. Smith", or with and without the article title, etc.). Consult the target journal's Author Instructions to ensure that your references meet the journal’s style requirements. Referencing software like Endnote provide a wide range of styles that can be used to automatically format citations and references to the correct style.
Tables and Figures
"A picture says more than a 1000 words" also applies to scientific manuscripts, so use figures not only to present the most important results of your work, but also whenever a verbal description would be too cumbersome (maps of the study area, model schematics, flow charts, photos of experimental set ups, etc.).
Ensure that all figure axes all labelled and that the font size is large enough for those labels to be legible. Avoid clutter in figures and maximize contrast by avoiding similar colours or gradations of grey whenever possible.
Figures and tables should be able to stand alone, i.e., the legend should be sufficiently detailed to understand what one is looking at without having to read the main text.
Do not repeat the figure or table legend in the main text and avoid verbose references to figures and tables. E.g., instead of “As the results shown in Figure 1 clearly demonstrate, the grain size decreases with increasing depth” simply use “The grain size decreases with increasing depth (Figure 1).”
An appendix contains materials that would obstruct the logical flow or even obscure the take-home message if they were included in the main body of the manuscript. Elements such as long tables, supplementary/supporting results, computer code, or large maps are likely candidates for an appendix. Some journals even limit the length of the Method section and require you to place the more detailed description of your method into an appendix. All appendices should be given an appropriate title and numbered consecutively.
- Ensure the title is concise and well chosen to reflect the content of the manuscript
- Spend some time on your abstract to make sure that it is well written and provides a good summary of your work
- Place materials that would distract from the main message in an appendix
- Do not include redundant phrases in the title
- Avoid references or too many acronyms in the abstract
- Do not repeat (parts or all of) a figure/table legend in the main text
While the emphasis in scientifc writing is typically on the substance and content, the element of style should not be neglected, as even the best of scientific breakthroughs may go unnoticed if the manuscript is incapable of getting the message across or keeping the reader interested long enough. The following tips and conventions are meant to improve the overall clarity and readability of your manuscript.
General Writing Style
Particularly in the quantitative (natural) sciences, the main focus of scientific writing is be to present the results in an as clear and concise manner as possible. Also bear in mind that your text will be read by scientists with varying degrees of English proficiency, so as a general rule, try to keep your sentence structure simple, avoid convoluted constructions with many subclauses, and reduce verbosity and other unnecessary ballast. This not only makes the writing easier for you, but also keeps others interested and allows them to grasp the essence of your work more quickly.
Provide Logical Links
You can greatly increase the readability of your manuscript, if you provide logical links between sentences. This can be done with what is referred to as conjunctive adverbs:
- To illustrate a point: as shown by, e.g., for instance, in particular, namely, such as, etc.
- To add to a point: in addition, furthermore, moreover, next, again, etc.
- To compare points: also, likewise, similarly, commensurate with, irrespective of, etc.
- Counter points: although, while, nevertheless, whereas, in contrast, however, unlike, etc.
- Logical connection: accordingly, as a result, therefore, thus, so, hence, etc.
- Temporal connection: meanwhile, subsequently, when, while, after, etc.
- Spatial relation: next to, west of, in direct proximity to, etc.
- Summarize a point or argument: in conclusion, in summary, therefore, to conclude, etc.
Do not start every sentence with such a connector but use them wherever they increase coherence and logical flow.
Lists are useful stylistic elements but can quickly become incomphrehensible if used improperly. In our editing work, we have come across lists with over 20 elements that were presented like: “We tested xxx, then we tested yyy, which was followed by zzz, etc.”, which makes for very unpleasant reading indeed. As a rule of thumb, try to limit your lists to a maximum of 5 elements and employ the approriate punctuation to present the list in a clear and pleasant manner, e.g., “We tested 5 different methods: (i) xxx, (ii) yyy, (iii) zzz, …”.
Proper punctuation is critical to avoid ambiguity in lists. Many journals recommend the use of the serial (or Oxford) comma, i.e., to set a comma after the penultimate item in your list, before the conjunction. For instance, compare "I dedicate this book to my parents, Peter Smith and Lisa Walker." with "I dedicate this book to my parents, Peter Smith, and Lisa Walker". The first sentence implies that you are dedicating your book to your parents, who are called Peter Smith and Lisa Walker, whereas the second version (with a comma before the "and") implies that you are dedicating your book to three different entities.
It is equally important to know when not to set a comma. Take for instance: "The effects of gingivitis, smoking and other drug intake were evaluated in a model of lung desease." Here, only 2 factors were evaluated: "gingivitis" and the combined effects of "smoking and other drug intake". If "smoking" and "other drug intake" were evaluated as separate factors, you would need to separate them with the serial comma. Being consistent in your use of the serial comma can go a long way in promoting clarity in your writing.
Ambiguity and Word Use
Avoid ambiguity whenever possible. For instance: “The method was tested in an experiment. It consisted of 3 steps” is ambiguous as “It” could refer to either “method” or “experiment”. Better: “The method was tested in an experiment that consisted of 3 steps.”
Unlike prose, scientific texts must be precise and avoid even the slightest ambiguities. Say you have introduced a particular label, acronym, or terminology, then do not replace that label with a synonym, simply because you are trying to avoid repeating the same word. In scientific writing, repetition is good and necessary as it prevents ambiguity while helping the reader remember the meaning of the introduced label throughout the entire length of the manuscript.
Avoid Colloquialisms, Contractions, and Clichés
- a lot of
- look at
- go up/down
- more or less
- perform/carry out/conduct
- such as
- don’t use
- be more precise
Avoid contractions (use “do not” instead of “don’t”) and vague phrases such as: “The layer thickness differed quite a lot”; instead: “The layer thickness differed by xx “.
Many articles are riddled with vacant clichés like “Our results will greatly advance the field and benefit the scientific community.”. Such phrases are essentially meaningless and should be avoided. Instead, be specific on how your results will advance your field.
- As a matter of fact
- At the present time
- Due to the fact that
- In spite of the fact that
- In fact
Avoid redundant phrases such as: “It is interesting to note that …” or “It can be remarked that …”.
Avoid the unnecessary use of pronouns: Instead of “The flowers that we saw were blue.” use “We saw blue flowers.”.
Use of Tenses
Tenses are important as they allow the reader to identify whether you are talking about your own results and actions (present) or someone else’s (past), or whether you are stating a common truth (present) vs a hypothesis (conditional) or an observation (past).
For instance: “Iron oxides frequently occur as secondary precipitates” implies that you are stating a commonly known truth or accepted fact (hence present tense), while: “Iron oxides frequently occurred as secondary precipitates” implies that you are talking about your own findings.
Or: “These results show that …” implies that you are talking about your own work, while “Smith et al. (2003) found that … . “ and the continued use of past tense in subsequent sentences would suggest that you are (continuing) to talk about findings by Smith et al.
For actions that started in the past but are still ongoing use present perfect. E.g., “Recently, there has been a strong debate about …” or “This method has been effectively used to demonstrate that …” .
Passive vs Active voice
Traditionally, the use of active voice in science was frowned upon as it places the protagonist in the centre (e.g.,"We found ...") while passive voice was considered more objective and less personal. Over the past couple of decades, this paradigm has gradually been changing, as both authors and publishers began to recognise that the use of active voice tends to allow for a more concise and less ambiguous writing style. As a result, the use of active voice is now fairly common and even recommended whenever it reduces amibiguity:
OK to use: “Plant seeds are dispersed by wind.”
Wordy and ambiguous: “Using the survey data, the effects of education on job satisfaction were examined.” (who is using the data?)
Better: “Using the survey data, we examined the effects of education on job satisfaction.”
Many journals now explicitly recommend the use of active over passive voice:
- Behavioral Ecology: "The first-person active voice is preferable to the impersonal passive voice."
- British Medical Journal: "Please write in a clear, direct, and active style....Write in the active [voice] and use the first person where necessary."
- The Journal of Neuroscience: "Overuse of the passive voice is a common problem in writing. Although the passive has its place—for example, in the Methods section—in many instances it makes the manuscript dull by failing to identify the author's role in the research....Use direct, active-voice sentences."
- The Journal of Trauma and Dissociation: "Use the active voice whenever possible: We will ask authors that rely heavily on use of the passive voice to re-write manuscripts in the active voice."
- Nature: "Nature journals like authors to write in the active voice ('we performed the experiment...') as experience has shown that readers find concepts and results to be conveyed more clearly if written directly."
- Ophthalmology: "Active voice is much preferred to passive voice, which should be used sparingly....Passive voice...does not relieve the author of direct responsibility for observations, opinions, or conclusions (e.g., 'The problem of blood flow was investigated...' vs. 'We investigated the problem of blood flow...')."
- Science: "Use active voice when suitable, particularly when necessary for correct syntax (e.g., 'To address this possibility, we constructed a lZap library ...,' not 'To address this possibility, a lZap library was constructed...')."