Most recent revision of this article was led by Communication Fellow Andrew Feldman.
Criteria for Success
- The paper provides a clear main message (that is, your one to two sentence take-away) and the broader context that the message addresses.
- The abstract stands alone and clearly conveys the message.
- The research questions/objectives and their motivation are clearly stated in the introduction.
- The results clearly address the questions/objectives.
- The figures and their caption can largely convey your message without depending heavily on the main text.
- While journals may ask for different formats, all of these sections below should be present in some form.
- Each main section should be of similar weighting. For example, proportionally too long of a discussion (and perhaps too many citations therein) could be construed as a speculative or review-type study.
Identify Your Purpose
Your paper should clearly provide:
1) Main message. What did you find? Your “main message” is the one to two sentence take-away that any reader should know after viewing your paper.
2) Context for the main message. What is the significance of your message? Why should anyone care? Expand on the context in the introduction. Start with a big-picture background and use the literature review to narrow down into the knowledge gap your work addresses.
3) Methodology and reasoning. How did you arrive at this conclusion? The methodology and argumentation throughout the paper should directly support the message.
Researchers mostly read the abstract, figures, and conclusions. Ensure that these components explicitly include the essentials above.
Analyze Your Audience
1) Evaluate your message’s scope. Who is the message targeting? Will others beyond your target audience appreciate the message? Some messages can be appreciated by researchers across many fields including those in your specific field.
2) Choose the journal. Journals define their readership, each having their own following.
- Broader journals, such as Nature and Science, have readers from broader backgrounds, requiring messages to be understood by non-technical, novice readers. Jargon is avoided.
- Technical journals, such as American Society of Civil Engineering and American Geophysical Union journals, will typically have well-defined, expert audiences.
3) Consider the audience’s prior knowledge. After defining your work’s scope and choosing a journal, strike the right balance between attracting more novice readers while providing expert readers with enough information.
- Novice Audiences may include anyone from public officials to scientists in other fields. They will make up the majority of your readership and will have less prior understanding of your topic. Give this readership more introductory background, less jargon, and simplified research questions/objectives.
- Expert Audiences will have prior understanding of your main research questions, methods, and jargon. This readership will want to find the message quickly and a more nuanced discussion of results and methods.
Plan before writing
- Produce preliminary figures with your data. This helps determine where significance tests or repeated experiments may be required, preventing having to rerun experiments during the writing process.
- Define and analyze the audience. See “Analyze your Audience” above.
- Discuss the storyline with a co-author, colleague familiar with your work, Comm Lab fellow, and/or a non-scientist friend. It takes little time and usually develops the paper flow, making it easier to start writing.
- On the journal’s webpage, look for the document often called “Author Guidelines.” It will provide detailed length and format requirements. Make a checklist of items you will need. This step will save you from having to reformat later.
Ease into writing
- Start writing with an outline and build your argument with a bullet list. Reference your arguments either to your figures or other studies. Begin citing other studies. Take advantage of citation tools like Mendeley to automate organizing and formatting references.
- Start writing by focusing on getting your ideas down on paper rather than on grammatical accuracy or paragraph flow.
- When you are ready, start writing “easy sections” first. The methods and results sections may be the clearest components of the work at this stage.
- Abstracts and conclusions may be easiest to write after all other sections are written. Alternatively, a good exercise is to start with a rough abstract to set the tone for how you will write the rest of the paper.
Ask for feedback iteratively from co-authors
- Get feedback from your co-authors iteratively throughout the planning and writing stages. Note that busy co-authors may provide quicker feedback on smaller units like figures, outlines, or the introduction section.
- Anticipate (and embrace) several rounds of co-author review before you submit. Maximal feedback prior to submission helps you anticipate what your reviewers will request after submission.
The title is attractive
- The title should attract your audience while using novice keywords that help your paper show up in search engines.
- A great option for your title is a shortened version of your one-to-two sentence main message.
- Journals ask for keywords to help your paper show up in search engines. Work in the more novice keywords to your title.
The abstract is key
- The abstract will be the most-read section. It should stand alone because it may be the only information available to those who cannot access the paper.
- After writing the abstract, ask whether it includes your main finding, its significance, and the methods used to arrive at the finding.
The research question/objective is clearly stated
- A strong introduction will include the background (what has been done before) and a research gap (what needs to be done). Your research question and/or goal will follow the identified research gap.
- Engineering disciplines generally prefer research objectives whereas science disciplines prefer research questions.
Focus on the figures
- Many readers may only view the figures. Make them self-explanatory and independent of the main text.
- Are your main messages clear from the figures?
- Is the figure’s caption clear enough to explain what the figure is depicting without reading the article?
- Are the figure’s acronyms and symbols defined clearly in the caption? Avoid acronyms when possible.
The conclusion gives a little more detail than the abstract
- The conclusion should be a summary and read similarly to the abstract, but can be longer and include more detail.
Each paragraph starts with the message
- It is helpful if a reader can get the main message from only reading the first (topic) sentence of each paragraph.