On Writing Scientific Papers

A long time ago, I read some articles on how to write a good scientific paper. This is a synthesis of those notes, presented in a better format than my old “bullet point of quotations” style from years ago. Keep in mind that this research was made for the purpose of writing fiction which feels like an authentic imitation of a scientific paper, and I do not necessarily advise writing an actual scientific paper based purely on what I have presented below.

The content of this post was drawn from:

General Style Advice

Decide where you’re going to submit, check their guidelines (see below), and write an outline which stays aware of those guidelines.

Write in the third-person. Write like Hemingway, and remove unnecessary verbiage. Shorter sentences are better than longer sentences. Soften your language with words like “possibly” and “suggests.”

Be precise in your language. Using the same word over and over again is better than using different words and being less than completely accurate. Be formal, and avoid conjunctions and slang.

Graphic representations, including graphs, are great, but everything should be comprehensible at a glance and you should not use a graph where a single sentence will do (a picture says a thousand words, but make sure there are a thousand words to say!). Color photographs are best, and make sure that the image is clear of distractions (including backgrounds and bizarre clothing). Links to videos are becoming popular, but keep them short (usually less than five minutes total).

Cite everything. Anything that is not 100% your own idea should get a citation. It is better to over-cite than under-cite. Aim for recent citations, to demonstrate that you’re familiar with the current literature. Use indentations when quoting long passages. How to write citations is a whole thing in itself, and you have Google, so I’ll leave you to read up on that on your own.

Target Journal

You will probably be publishing in a peer-reviewed journal, which will have its own guidelines. Make sure that you are following those guidelines.

Reviewers will care most about the following things: (1) the general importance and contemporary relevance of your findings; (2) the quality of your writing; (3) the quality of the study’s procedures; (4) the quality of your citations; and (5) your sample size. Papers are rejected most often for reasons of (1) bad statistics; (2) bad interpretation of results; (3) poorly-described results; (4) bad procedures, especially inappropriate samples; and (5) badly-written text (including typos).

Structure

Use the IMRaD format: Introduction, Methods, Results, and Discussion.

Write your results first, then the introduction and discussion, so that you have a clear place to “aim” at and don’t go off-topic. Write the conclusion next, summarizing everything important that you’ve gone over. The abstract should be written last, once you know what the rest of the paper will look like.

The person who wrote the paper is usually listed as the first author of the paper, but if you’re just a dirty grad student then your prestigious professor will probably be listed first. Other people who made contributions should be listed as co-authors, but get everyone’s permission first.

Abstract

The purpose of the abstract is to summarize the rest of the paper. Ideally, a reader will be able to come away with the abstract knowing everything of importance, and the rest of your paper is there to supply additional details, including context and everything that someone would need in order to perform your study again.

Do not use abbreviations or footnotes.

The abstract should be 100-250 words.

Introduction

Among other things, the introduction explains important background information, including the necessity of your study. You may want to reference previous studies. At the end, state your hypotheses and the purpose of the study in a clear manner.

Remember that only people demonstrate, describe, and think about things: do not say “the evidence showed that…” or “the study concluded that…” but rather “Salt and Wasser described…”

The introduction should be 250-600 words.

Methods

Describe how your study was conducted, with all relevant details. What was your sample size, and how was it collected? How was the data analyzed? Etc, etc. Start with a paragraph-sized overview of everything, then go into the details. Discussion of statistical methods will usually go after everything else. It should be possible for someone to perform your study a second time with only the information provided by your paper, and get the same results.

Make sure to mention at the end that you received approval from the relevant institutional review board.

Results

Show all your results, not just the interesting results or the results that back up your argument. Check out the CONSORT guidelines, STARD checklist, and PRISMA checklist for details if you want to be pass more than the basic sniff test (reminder: this is for fiction, and you should definitely do more than just “check them out” if you’re using this post to write an actual scientific paper for some reason).

Discussion 

In this section, you should put your results in a broader context. Why do your results matter? What do they imply? If your results differ from the results of similar studies, address the discrepancy. Be clear when you are merely speculating.

Conclusion

This is a summary of your findings, but also a place for any final statements. There should be nothing surprising or controversial in your conclusion which has not already been discussed previously. You may want to mention future research opportunities here.

The conclusion should be 40-100 words.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s