Actually, don’t start here. Start by reading the “Scientific Writing” section (previous section in Main Topics) before getting into the weeds of “Publishing Manuscripts”.
Side-note: Poorly written papers can sometimes be published. But your job is to write the best possible paper you can.
How to write a Methods section
This is the most straightforward section to write, and is a good place to start writing and build momentum.
Be clear, precise, detailed, and almost boring. Unlike other sections, do not worry too much here about the sentence structure and musical flow of the text. Provide a faithful account of the different experiments, simulations, and techniques used to acquire and process data.
- Mention all the key details needed for someone else to reproduce the data. Do not underestimate small details that you would take for granted having done the experiment a billion times.
- Be quantitative. For example, use “the sample was heated to 80 C for two hours” instead of “the sample was heated”.
- Build on methods sections from previous papers by your group. This is a good starting point, but do not be satisfied. They (or your previous self) may have overlooked key details.
- Make sure you conform to the style of your target journal. Some journals do not include a methods section at all in the main text and defer all technical details to the Supplementary Information.
Useful phrases and words
Was used – was utilized – was employed – equipped with – operated at – we performed ____ using a ____
Analyzed – acquired – obtained – processed
Samples – aliquots – batches – cycles
How to write a Results and Discussion section
Start by describing and observing your results, not by interpreting them. Be patient. You have thought long and hard about this problem. We have not. Help us get there with you.
When writing a paragraph, cycle between these different types of sentence structures:
- Get help from the figures:
“Figure X shows that the apple fell from the tree after an observation time of seven minutes.”
- Use active voice:
“We observed that the apple fell from the tree after a duration of seven minutes.”
- Direct observation:
“The apple fell from the tree after an observation time of seven minutes.”
- Start with a dependent clause:
“After a duration of seven minutes, the apple fell from the tree.”
- Elaborate on an observation:
“This observation shows that …”
- Start with a transition word or expression:
“Subsequently, the apple fell from the tree.”
Whenever you are stuck on how to write the next sentence, fall back on one of these six sentence formats. A sample text discussing some interesting results on apples falling from trees is presented below (please do not take these apples too seriously). The numbers in parentheses indicate each type of sentence structure as described above:
(1) Figure X shows three time-lapse images of an apple falling from a tree. (3) The apple moved vertically downwards and accelerated until it reached the ground. (4) After several minutes, another apple fell from a separate tree branch (Figure Y). (6) Interestingly, the second apple followed a similar trajectory to the first one during its free fall. (5) This observation suggests that the same force was acting on both apples causing their similar behavior.
(4) To examine this hypothesis, we plot the velocity of the apples during free fall. (1) Figure Z shows that the velocities increase linearly with time, indicating a constant acceleration…
Important notes from this sample paragraph:
- Shuffling between the types of sentence structures creates a crisp, flowing text.
- The sentences are more or less the same size. There are no loooooooong sentences that could be broken up for clarity. In real life manuscripts, you will have long sentences. But be compact and clear as much as possible.
- The last sentence in the first paragraph is a general statement that concludes the observations and builds up the story. We do not know everything about the system yet, but we did learn something. Tell us what we learned, and put it in the broader context. You do not have to end every paragraph this way. But it helps to start or finish paragraphs with summary statements, sort of to remind us where we are in the story.
- Do not over-interpret. Do not get ahead of yourself, but help the reader follow your line of thought. A paragraph is not an island; it should flow into the next paragraph.
- In this vein, the first sentence of the second paragraph is a liaison that builds on the previous paragraph and prepares the reader for a more rigorous analysis.
- When you feel stuck, fall back to the figures. “Figure X shows …” is a life-saver.
List of useful transition words and expressions:
Specifically – In particular
Subsequently – Eventually
As a result – In conclusion
Clearly – Notice that – Note that – Interestingly – Evidently
In contrast – By comparison – Similarly – Similar to – In this context
Moreover – Furthermore – In addition – To further confirm/demonstrate – To examine this hypothesis
However – Nevertheless
List of useful verbs:
Shows – highlights – presents – demonstrates – exemplifies – denotes – indicates – exhibits – is characterized by – is composed of – consists of – includes – provides insight into – results in – confirms – validates – corresponds to – is related to
Active voice: We observe – perform – report – demonstrate – analyze – emphasize that – examine – conclude – consider – delineate – elucidate
List of useful nouns:
Approach – strategy – method – technique
Data – observation – result – instrument
How to write a Conclusions section
A typical conclusion can be two to three paragraphs that accomplish the following:
- First paragraph: Recap the results. Do not recycle exact sentences from the Abstract or Results sections. Use fresh wording. Do not re-state every detail from your results. Instead emphasize the one or two main findings.
What do you want the reader to remember from your study?
- Second paragraph: Reiterate the challenges. In addition to mentioning the results, remind the readers how your work fits in and contributes to the research field. Then mention what are the limitations of your study in particular and the research field in general.
What problem would the next study solve?
- Third paragraph: End with big picture. The conclusion should end with an outwards look beyond the narrow focus of the current study. This part can be more ambitious about where you see the research field going, and the broader impact on other areas of research. As a fun exercise, think about this question: if you had $50,000,000, where would you take this research field? (For most of us, this is nothing more than a fun exercise).
What problems would years of research in this field solve?
Notice that the three paragraphs fuse into each other to some extent, which is kind of the point.
How to write an Introduction
A good introduction should: 1) present a brief survey of the literature that eventually lands the reader on the question you are asking, 2) provide the right balance between big picture statements and more technical examples, 3) introduce science concepts that you will use later in the manuscript, and 4) set the stage for the research problem that you will solve. All of that is easier said than done…
Below is a sample introduction on potato chip research for a study that will explore the use of olive oil in oven-baked kale chips. (Please do not take this example too seriously!). The numbers preceding each sentence correspond to the comments below:
“(1) Potato chips have been a regular staple of American snacking for decades.1 (2) In recent years, the potato chip industry has expanded to include a vast range of natural and artificial flavors,2 as well as kettle-cooked products with a distinctive crunchy texture.3,4 (3) Despite these exciting advances, consumer satisfaction with potato chips has been steadily declining, as consumers have opted for healthier snacks.5,6
(4) In this context, oven-baked potato chips have emerged as a promising alternative, mostly due to their significantly reduced saturated fattyacid content. (5) For example, Rootveg et al. showed that oven-baked chips contain 35-50% less saturated fat compared to traditional chips of the same brand.7 (6) Similarly, Spud et al. reported that the saturated fat content in the final product can be further decreased without any drastic impact on chip crunchiness.8 (7) These results suggest a marked change in the nutritional profiles of potato chips on the market and might help shift public sentiment regarding this product.
(8) Another approach has been to apply potato chip processing techniques to other vegetables with different nutritional properties. (9) Interestingly, this method has had long-standing success in other countries where plantain chips and other non-potato items dominate the market.9 (9) More recent examples include lentil chips which contain a decent dose of vegetable-based proteins per serving; possibly appealing to a broader consumer base. (10) However, these alternative products have been criticized for overuse of artificial flavoring and have thus not gained traction in the market.10
(11) In this manuscript, we studied another increasingly appealing snacking option, homemade kale chips. Specifically, we investigated the effect of the use of olive oil on the crunchiness and flavor profile of oven-baked kale chips.”
Here is the breakdown of each sentence:
- Introduce the broader research field.
- Acknowledge the recent awesomeness in this field and cite interesting papers.
- Start setting the stage for the main problem or challenge that your manuscript will tackle.
- Describe previous attempts to solve this problem, firstly using a general statement.
- Elaborate on this broad approach by giving a specific example.
- Give another specific example to elaborate more.
- Recap the main idea in this paragraph and remind us how of how it fits in the broader narrative to solve the problem.
Now for another paragraph:
- Again, describe another broad approach to solving this problem.
- Acknowledge previous success in this field with a nice, flowing interjection statement. This is where a broad knowledge of the research field helps, not simply a narrow focus on your work.
- And again, elaborate using an interesting example.
- Demonstrate how this problem has not been completely solved, and there is still room for improvement.
- Important pivot: Now bring attention to your work. We flowed naturally from the literature survey into a justification of why your work is important and how it will advance this field. Watch out for the two most common traps when closing out an introduction:
- Do not describe experimental details (this is not a methods section).
- Do not summarize the results (this is not an abstract).
As you write your introduction:
- Think in terms of the storyline: The introduction should look something like this: Introduce the research field in broad terms. Describe recent advances and ambitious results. State a problem or gap. Mention broad approaches to solve this problem and specific attempts. Comment on the limitations of those solutions. Pivot to your work.
- Think in terms of the target journal. For example, the same research results might be relevant to both a physics journal and a materials science journal. But an introduction to Chemistry of Materials would focus more on materials properties, synthesis routes, and potential applications, while an introduction to Physical Review would focus more on the physical concepts pertaining to the study. You might write an excellent introduction, but for the wrong journal. Get inspiration from reading the journal description on their website, and browse through some of the abstracts from recent publications.
- Think in terms of what you research advisor is thinking. Make sure you are on the same page. It helps to think deeply about the introduction structure and then discuss it with your advisor before writing. Ask for interesting papers to serve as starting points. Your advisor has written some version of this introduction many times already. You will bring fresh ideas and hopefully more than that, but it helps to build on core publications from your research group. Otherwise, your advisor will make you re-write everything. (Actually, never mind, your advisor will make you re-write everything anyway).
- Think in terms of the papers you are citing. If you know the literature, writing the introduction is almost fun! (Just kidding, it’s not, but I mean it’s not superhorrible). Just as having good figures will guide the results section, citing the proper references guides a good introduction. In fact, a decent chunk of the introduction would be single-statement summaries of some relevant previous studies. In a later article, we provide a detailed guide on choosing which references to cite.
Common mistakes when writing introductions
- Do not write a purpose-less introduction. There are hundreds of papers that you can cite and lines of thought that you can follow. But you want to be purposeful in guiding the reader towards your hypothesis. As an example, let us revisit the potato chips introduction. Notice the reference to “crunchy texture” in statement (2). Similarly statement (6) continues the theme of chip crunchiness. The reason is to direct the reader to this specific property early on, because of its relevance to our study, as becomes evident in statement (11). One could have mentioned and cited any other potato chip quality, such as the chip size, shape, or uniformity. But among the multitude of references and ideas that seem relevant, it is useful to pick the truly relevant ones that create a flowing text and build your scientific story. (Please do not take the potato chip example too seriously).
- Do not be too technical and detailed. This is probably the most common mistake. A non-specialist scientist should enjoy reading the introduction. The purpose is to clearly delineate the line of research progress that has led you to this hypothesis. Trim the excess.
- Do not be too un-technical and un-detailed. This problem is less common than the previous one (scientists like to talk a lot!). But it is still worth mentioning.
- Do not start from the beginning of time. This is not a historical lesson, or even a comprehensive scientific review. Of course, that is why you refer the readers to textbooks and reviews for more details. After opening with a general statement on the research topic, quickly shift the attention to the problem at hand. Do not derive Newton’s laws.
- Do not describe your results or experiments in detail. Instead use the last small paragraph of the introduction to elegantly turn the attention from the published literature to your current work. Do not give away the story yet. Remember that the condensed version of your story is in the abstract, not the introduction.
- Do not exaggerate the research significance. Write what you believe. Your research does not need to have applications in energy storage and drug delivery and cancer treatment and multiverse theory and optophotomagnetoelectronics to be meaningful. By the way, I am not convinced that overselling in that sense actually increases the chance of getting the paper published.
- Do not be dismissive or overlook the contributions of other researchers.
Useful expressions for introductions:
Recent advances – in this context – has gained increasing interest – shows great promise – In principle – in recent years
Despite recent advances – Other challenges include – Although – While previous work has focused on – To date – remains lacking – remains poorly understood – vastly understudied – Further work – To address these challenges – gain an understanding
For example – For instance – One interesting example – One important contribution – Typically
How to choose which References to cite
- Find a good review paper. Not only is this a useful citation for readers, it will also lead you to other papers and research groups you might have not noticed before. Read a couple review papers for inspiration in writing the introduction.
- Search for recent work by prominent research groups (or groups of groups) and their approach to working on this problem. Do not only cite the “big names” from top schools; there is a lot of good work out there.
- Include the following types of papers: 1) Textbooks or review papers, 2) classic papers that shaped the research field, 3) recent hot papers rather indirectly related to your work, 4) specific papers directly related to your work–scrutinize the literature, this is the most important type to cite, 5) papers you borrowed technical details from, 6) some of your research group papers, if relevant, 7) contributions from as many of the prominent research groups in this field as possible.
- Tailor the references to the readership of the target journal. For example, if you are targeting a physics journal, make sure you cite enough papers from physics journals as well, not overwhelmingly from materials science or chemistry journals. Also do not cite a million papers. More is not necessarily good.
- Check which papers were previously cited by your group. This is a good starting point, but go an extra step in searching. Do not simply re-cite the same papers out of habit.
- Oh by the way, read the papers you cite. At least read the part that is relevant to what you are citing. Second-hand citations (i.e. citing papers simply because they were cited by others) is not a good idea.
There is another “strategic” reason why doing a good job with references is very important, if we are being perfectly honest. Your reviewers are potentially experts who have made contributions in this research field. You do not want to disregard their work completely. That is not a good look. It is also not a good way of doing science.
Needless to say, you should always be updated on the literature; not only when it is time to write the paper. Otherwise, you might learn belatedly that your paper has already been published by someone else. Sometimes we are too narrowly focused on our next experiments and forget what’s out there.
Finally, format the references properly. Do not trust endnote (or any other tool) to do a perfect job. If your reference formatting is sloppy or inconsistent, you are giving the impression that your science is the same. Be consistent in choosing sentence vs title case, mentioning first and last page vs only first page, abbreviating journal name or not, mentioning issue number or not, as well as the initials/last name format of the authors. All those incredibly fun details.
You have not earned your PhD until you go through this brutal experience: Meticulously format references section. Re-double-check reference list and correct any typos. Send paper to co-authors for their feedback before submitting. Receive paper back from co-authors. The shock. The horror. The humanity. The reference formatting got confused on someone’s computer and now you have to re-format everything.
Anyway, that is why it is a good idea to take one last look at the references section before submitting the manuscript.
How to write an Abstract
Every. Word. Counts.
- As much as possible, communicate what you learned, not just what you did.
- List the main results. No fluff.
- Make use of your word limit. If the journal allows 200 words for the abstract, use them.
- Remember that the abstract is what most people will read. Present your findings in the best possible way.
Okay, that sounds a little bit too abstract (Woahhh! Ta na na na na), so here is a working example:
- Introduce the research topic. (~1–2 sentences).
“Nanoparticles in solution can either aggregate into larger structures or remain stable depending on the nature and scaling of their interaction forces.”
- Set the problem (1 sentence). Mention the challenge or limitation or gap in this field that your work intends to overcome.
“Predicting the outcome of aggregation remains challenging, particularly for non-spherical particles in extreme solution conditions.”
- Pivot towards your work (1 sentence). Luckily you are here to solve the problem. Cue the trumpets. Cue the triumphant armies. Cue the hours of graduate life in the lab.
“In this context, we investigated the structure and dynamics of nanoparticle aggregation as a function of pH and ionic strength.”
- Describe the main results (remaining sentences). Here, the priority is to communicate what you learned, not what you did. Of course, it is important to also say what you did. But the abstract should summarize findings, not be a laundry list of techniques you have in the department.
“Our microscopy and light scattering data showed that the nanoparticles undergo reaction-limited aggregation with rate constants between 10-5 and 10-3 mol.L-1.s-1.”
is much better than
“We monitored the nanoparticle aggregation using scanning and transmission electron microscopy, as well as dynamic light scattering.”
- Close strongly (1 sentence). If you still have space, remind readers why this work is important.
“These results are important steps towards a predictive understanding of particle transport and aggregation that is relevant to problems in geochemical remediation, crystal growth, biomineralization, and materials science.”
Common mistakes when writing abstract and/or manuscript title
- Consider how your abstract sounds to a scientist working in a different research field. Terminology you take for granted might mean different things for them. Do not make your abstract or title sound mysterious or vague. For example,
“Nanoparticle interactions drive the assembly of superlattice structures”
is better than:
“Particle interactions drive the assembly of superlattice structures”
The word “particles” might be interpreted by a physicist as sub-atomic particles, instead of nano-sized particles, and leaves your message vague.
- Most titles sadly do not describe what the author learned from the study, but rather what they did or what question they tried to answer (minus the actual answer). Dealing with this problem is actually harder than it sounds, but try your best. Here is an oversimplistic example:
“Smoking tobacco increases the probability of lung cancer”
is better than:
“Effect of smoking tobacco on incidence of lung cancer”
The first title actually tells the reader the singular, most important result from the study. It is the take-home message from your months of research. Again, in reality, having a title that elegantly summarizes the main result is harder than it sounds, because 1) studies are often convoluted with multiple incremental advances rather than a clear defining result, 2) studies are often too technical to reduce into a single statement, and 3) you do not want to be sensationalizing the research result without describing the caveats that come with it, which you do not have space for in the title.
Getting rejected is not the end of the world
There are three types of unpleasant feelings when your paper is rejected:
- Feeling that the editor weeded out the paper without giving it a shot, and you believe that was unfair.
- Feeling that the review process took forever only to find out that the paper has been rejected and you are back to square one.
- Feeling that the reviewer did not truly consider or appreciate or comprehend your work to the extent that you find satisfactory.
- Of course, feeling that the work is not as good as you thought.
For the first three cases, that’s too bad. But hey, there are many journals out there. You will get published eventually. Hopefully, your next submission is an improved manuscript. The paper usually ends up being improved by the peer review process, which is kind of the whole idea (especially if you actually received referee reports):
- You address the concerns of the reviewers who provide a different perspective on your work.
- You are encouraged to do an extra experiment or repeat an analysis that ties up the loose ends.
- You re-think the manuscript more clearly and look at it with a new lens. This is actually an underrated part of the peer-review process: the chance for you and your co-authors to be your own critical reviewers.
In many cases, referee reports have clearly convinced me that my manuscript should be radically improved before being published. Often, that is what happens, and I end up being thankful that the earlier version did not make it.
By the way:
The system is not perfect, good papers often get rejected. Less good papers often get published.
Your paper is not perfect. Regardless of what you personally think about the reviewer comments, consider them objectively, and go the extra step to improve your manuscript.
Here are some additional thoughts on responding to referee reports:
- Respond to every reviewer comment, regardless of how you feel about it. You might politely say that you disagree and provide the rationale for why.
- Read all the paper again before re-submitting. Even if you only made minor changes, it is important to see if the paper still flows well and the new comments fit nicely.