Skip to content

Instantly share code, notes, and snippets.

@vicenterusso
Created March 18, 2023 13:38
Show Gist options
  • Save vicenterusso/0fc7a44f0abb984a1367980a7fb5447c to your computer and use it in GitHub Desktop.
Save vicenterusso/0fc7a44f0abb984a1367980a7fb5447c to your computer and use it in GitHub Desktop.

USER

You are a professional note taker with background experience as a professional book writer and editor. Because of this, all of the notes you take have exceptional grammar and spelling and are very high quality.

I will provide you with text to take notes on, and you will follow the below list of requirements exactly as they are listed before providing a response, checking at every step to ensure that all requirements and suggestions within the following list are met and upheld for each line of notes that you write.

All of your output will be in a Markdown code block You will extract the most important information from the provided text, including but not limited to: Content that could appear in a quiz. Context that provides background information. Examples related to the context or information being presented. Facts or statistics related to the context or information being presented. Useful tips and tricks related to the context or information being discussed. Links to external sites. The output format that you use will utilize the following note-taking structure: Topic title or subject matter Any subtopics or sections to include Formatting: Use headings, bullet points, numbered lists, and other tools to organize the notes Include examples and illustrations where relevant to enhance understanding. Use Markdown tools, Markdown-compatible latex tools, and HTML tags to enhance the notes. You will use syntax highlighting to enhance the notes. Bolster the existing notes you have created so far with any relevant and highly accurate information from your own knowledge, only if you are 100% sure that the knowledge is accurate, relevant, and enhances the context of the notes. If any information appears inaccurate or incomplete, you will supplement it with your own knowledge, while adhering to all of the steps listed in these requirements. Additional Information: Provide a conclusion or conclusion section at the very end that synthesizes the key takeaways from the notes. Do not reply to any of this until I have provided you with text to take notes on.

ASSISTANT

I understand the instructions and will adhere to them. I will await your text to take notes on.

USER

At any given time, Andrea Armani’s lab at the University of Southern California has up to 15 PhD students, a couple of postdocs, nine undergrads, and an occasional high school student, all busy developing new materials for diagnostic and telecommunications devices.

When conducting scientific research, Armani believes it’s important to test a hypothesis—not prove it. She recruits students who are willing to adopt that “testing” mentality, and are excited to explore the unknown. “I want them to push themselves a little bit, push the field a little bit, and not be afraid to fail,” she says. “And, know that even if they fail, they can still learn something from it.”

Armani often coaches students through the process of writing their first scientific paper. Her 10-step formula for writing a scientific paper could be useful to anyone who has concluded a study and feels the dread of the blank page looming.

  1. Write a Vision Statement What is the key message of your paper? Be able to articulate it in one sentence, because it's a sentence you'll come back to a few times throughout the paper. Think of your paper as a press release: what would the subhead be? If you can't articulate the key discovery or accomplishment in a single sentence, then you're not ready to write a paper.

The vision statement should guide your next important decision: where are you submitting? Every journal has a different style and ordering of sections. Making this decision before you write a single word will save you a lot of time later on. Once you choose a journal, check the website for requirements with regards to formatting, length limits, and figures.

  1. Don't Start at the Beginning Logically, it makes sense to start a paper with the abstract, or, at least, the introduction. Don't. You often end up telling a completely different story than the one you thought you were going to tell. If you start with the introduction, by the time everything else is written, you will likely have to rewrite both sections.

  2. Storyboard the Figures Figures are the best place to start, because they form the backbone of your paper. Unlike you, the reader hasn't been living this research for a year or more. So, the first figure should inspire them to want to learn about your discovery.

A classic organizational approach used by writers is "storyboarding" where all figures are laid out on boards. This can be done using software like PowerPoint, Prezi, or Keynote. One approach is to put the vision statement on the first slide, and all of your results on subsequent slides. To start, simply include all data, without concern for order or importance. Subsequent passes can evaluate consolidation of data sets (e.g., forming panel figures) and relative importance (e.g., main text vs. supplement). The figures should be arranged in a logical order to support your hypothesis statement. Notably, this order may or may not be the order in which you took the data. If you're missing data, it should become obvious at this point.

  1. Write the Methods Section Of all the sections, the methods section is simultaneously the easiest and the most important section to write accurately. Any results in your paper should be replicable based on the methods section, so if you've developed an entirely new experimental method, write it out in excruciating detail, including setup, controls, and protocols, also manufacturers and part numbers, if appropriate. If you're building on a previous study, there's no need to repeat all of those details; that's what references are for.

One common mistake when writing a methods section is the inclusion of results. The methods section is simply a record of what you did.

The methods section is one example of where knowing the journal is important. Some journals integrate the methods section in between the introduction and the results; other journals place the methods section at the end of the article. Depending on the location of the methods section, the contents of the results and discussion section may vary slightly.

Science writing graphic 5. Write the Results and Discussion Section In a few journals, results and discussion are separate sections. However, the trend is to merge these two sections. This section should form the bulk of your paper-by storyboarding your figures, you already have an outline!

A good place to start is to write a few paragraphs about each figure, explaining: 1. the result (this should be void of interpretation), 2. the relevance of the result to your hypothesis statement (interpretation is beginning to appear), and 3. the relevance to the field (this is completely your opinion). Whenever possible, you should be quantitative and specific, especially when comparing to prior work. Additionally, any experimental errors should be calculated and error bars should be included on experimental results along with replicate analysis.

You can use this section to help readers understand how your research fits in the context of other ongoing work and explain how your study adds to the body of knowledge. This section should smoothly transition into the conclusion.

  1. Write the Conclusion In the conclusion, summarize everything you have already written. Emphasize the most important findings from your study and restate why they matter. State what you learned and end with the most important thing you want the reader to take away from the paper-again, your vision statement. From the conclusion, a reader should be able to understand the gist of your whole study, including your results and their significance.

  2. Now Write the Introduction The introduction sets the stage for your article. If it was a fictional story, the introduction would be the exposition, where the characters, setting, time period, and main conflict are introduced.

Scientific papers follow a similar formula. The introduction gives a view of your research from 30,000 feet: it defines the problem in the context of a larger field; it reviews what other research groups have done to move forward on the problem (the literature review); and it lays out your hypothesis, which may include your expectations about what the study will contribute to the body of knowledge. The majority of your references will be located in the introduction.

  1. Assemble References The first thing that any new writer should do is pick a good electronic reference manager. There are many free ones available, but often research groups (or PIs) have a favorite one. Editing will be easier if everyone is using the same manager.

References serve multiple roles in a manuscript:

  1. To enable a reader to get more detailed information on a topic that has been previously published. For example: "The device was fabricated using a standard method." You need to reference that method. One common mistake is to reference a paper that doesn't contain the protocol, resulting in readers being sent down a virtual rabbit hole in search of the protocol.

  2. To support statements that are not common knowledge or may be contentious. For example: "Previous work has shown that vanilla is better than chocolate." You need a reference here. Frequently, there are several papers that could be used, and it is up to you to choose.

  3. To recognize others working in the field, such as those who came before you and laid the groundwork for your work as well as more recent discoveries. The selection of these papers is where you need to be particularly conscientious. Don't get in the habit of citing the same couple of papers from the same couple of groups. New papers are published every day-literally. You need to make sure that your references include both foundational papers as well as recent works.

  1. Write the Abstract The abstract is the elevator pitch for your article. Most abstracts are 150–300 words, which translates to approximately 10–20 sentences. Like any good pitch, it should describe the importance of the field, the challenge that your research addresses, how your research solves the challenge, and its potential future impact. It should include any key quantitative metrics. It is important to remember that abstracts are included in search engine results.

  2. The Title Comes Last The title should capture the essence of the paper. If someone was interested in your topic, what phrase or keywords would they type into a search engine? Make sure those words are included in your title.

ASSISTANT

Note-taking on "A 10-Step Formula for Writing a Scientific Paper"

Introduction

  • Andrea Armani's lab at USC has up to 15 PhD students, postdocs, undergrads, and occasional high school students working on developing new materials for diagnostic and telecommunications devices.
  • Armani recruits students willing to test hypotheses and explore the unknown while being open to the possibility of failure.

Armani's 10-Step Formula for Writing a Scientific Paper

  1. Write a Vision Statement
  • The key message of the paper should be articulated in one sentence and guide the decision of where to submit the paper.
  • The chosen journal's formatting, length limits, and figure requirements should also be considered.
  1. Don't Start at the Beginning
  • Starting with the abstract or introduction can lead to telling a different story than intended, requiring rewriting.
  1. Storyboard the Figures
  • Figures are the backbone of the paper and should be organized logically to support the hypothesis statement.
  • "Storyboarding" can be done using software like PowerPoint or Keynote, starting with all data before consolidating data sets and evaluating relative importance.
  1. Write the Methods Section
  • The methods section should include excruciating detail, allowing for replicability of the results.
  • Inaccurate inclusion of results should be avoided, and the section's location may vary depending on the journal's requirements.
  1. Write the Results and Discussion Section
  • Results and discussion can be separate or merged sections, with storyboarding figures providing an outline.
  • The section should include paragraphs explaining the result, relevance to the hypothesis statement, and field, with quantification and comparison to prior work when possible.
  1. Write the Conclusion
  • The conclusion should summarize the paper's important findings and what was learned, with emphasis on the vision statement.
  • It should allow readers to understand the paper's gist, including results and significance.
  1. Now Write the Introduction
  • The introduction defines the problem in the context of a larger field, reviews other research groups' work, and lays out the hypothesis.
  • The majority of references will be located in the introduction.
  1. Assemble References
  • A good electronic reference manager should be used, with references serving to enable readers to get more detailed information, support statements, and recognize others in the field.
  • Foundational and recent works should be cited, and the reference manager should be consistent throughout the group.
  1. Write the Abstract
  • The abstract is the elevator pitch, with approximately 150-300 words describing the importance of the field, the research's challenge, how it solves the challenge, and potential future impact.
  • Key quantitative metrics should be included, and the abstract is included in search engine results.
  1. The Title Comes Last
  • The title should capture the essence of the paper and include words or phrases someone interested in the topic would search for.

Conclusion

  • Scientific paper writing follows a formula that includes a vision statement, storyboarded figures, and detailed methods and results sections, among other elements.
  • Armani's 10-step formula for writing a scientific paper can be a useful guide to anyone who has concluded a study and is facing the blank page.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment