How to read an academic science paper

Factsheets
How to cite: Lagut A, Wong M. How to read an academic science paper. Beauty SciComm Group. August 28, 2023. Accessed February 26, 2024. https://beautyscicomm.com/how-to-read-academic-science-paper/

Scientific marketing of cosmetic products is trending, and more people are interested in reading academic papers. Here’s how to do it, and some pointers to keep in mind.

Before you start

Reading scientific papers is a skill like any other – it takes effort and time to develop. Reading a newspaper article can take minutes. In contrast, reading, understanding and critically analysing a scientific paper can take many hours, or even days.

It’s crucial to understand that research papers are written for other researchers, not the general public. The goal of a research article is to convince the reader of the validity of the results and the conclusions. The readers’ job is to look at the data and make their own conclusions – this generally requires at least some advanced science education.

Science isn’t a collection of facts but a developing body of knowledge, and most research papers deal with a grain of sand in an enormous desert. One grain of sand is important, but unlikely to influence the whole, and should be viewed in context.

Before diving into the world of academic literature, textbooks and Wikipedia are usually good starting points. Crash Course and Khan Academy are great for catching up to undergrad level science. Udemy offers a 1-hour course on how to read and interpret research.

Assessing the paper

Before reading the paper, it’s important to understand the context of the paper and how much weight it should hold.

What type of paper is it?

Reviews summarise the existing research on a particular topic, and are a good starting point if you already have an undergraduate-level understanding.

For cosmetic ingredient safety, the Scientific Committee on Consumer Safety (SCCS) and Cosmetic Ingredient Review (CIR) reports are reviews that contextualise academic literature and industry data, and give a better overview than any single paper.

Research articles are the most common type of scientific papers. They present original research that the authors have performed, and put it into the context of the wider field of knowledge.

Who wrote the paper?

Are the authors suitably qualified and experienced in that field of research? Do they have any conflicts of interest that might lead to bias? The H-index of the author can be an indication of their scientific “weight”.

Their institutional affiliations, conflicts of interest and funding sources should be listed in the paper, but it’s important to crosscheck this with their public profiles. Inadequate disclosures can be revealing.

Where was the paper published?

Journals are of varying quality, and many predatory “pay to publish” journals exist.  Preprints have not undergone peer review. Metrics that give an idea of the standing of a journal include impact factor and H-index.

Other things to consider:

  • When was the paper published? If a field is rapidly developing, older papers may be outdated.
  • Is the paper well-cited compared to similar studies?

Reading an academic paper

First, set your research question or decide what is the purpose of reading the paper. Then you can employ this modified SQ4R strategy (or something similar):

Survey

Look through the keywords, titles and figures. Skim through the abstract, introduction and conclusion.

Highlight any terms or concepts you don’t understand and learn about them. Even working scientists have to look up a lot of other sources to properly understand topics they’re not familiar with! Looking up any referenced papers can be useful. Scientific terms can also have very different meanings from how we would use them in everyday life.

Question

Turn the different sections of the paper into questions – who, what, where, when, how, why, so what? Do the answers to these questions interest you? If not, next paper.

Read

Try to find answers to your questions. As you read, make sure you consider:

Is the methodology appropriate?

Think about whether the experiment actually tests the research hypothesis in a fair and rigorous way, and whether there are any limitations. For example, many cosmetic studies don’t use fair comparisons or controls. If the methodology is unusual, is a valid reason given, or is it a sign that standard methods didn’t give the results the researchers wanted?

Are the results presented in an appropriate way?

Look out for ways of presenting the results that make them look more significant e.g. graphs that use inappropriate axes, percentages that don’t reflect a clinically significant result, p-value hacking.

Does the discussion acknowledge limitations and appropriately contextualise the research? Do the conclusions logically follow from the results?

Researchers tend to hype up what their research could mean, but some do it to an unreasonable extent. Remember, extraordinary claims require extraordinary evidence!

Is the study relevant to your research question?

For example, tests on animals or cells don’t necessarily reflect what happens when humans use beauty products.

Respond

Try to answer each of your questions and orally summarise the paper in your own words. This helps you to clarify any foggy concepts to yourself.

Record

Highlight, underline, take notes. Keep the definitions of important terms at the top of the paper so they can easily be found.

Review

Regularly come back to the important points you’ve highlighted to keep them fresh in your mind.

If you are planning to make this a regular habit, use a library manager like Endnote, Mendeley or Zotero. These let you search keywords from your whole library of papers at once and save any highlights.

Common pitfalls when reading scientific papers

Cosmetic research is primarily conducted within the industry, and there is far less incentive to make research publicly available, so the peer-reviewed literature does not represent the full body of knowledge. Public funding is limited, many studies use inappropriate methodologies, and publications are commonly used as a form of marketing.

Peer reviewed papers are evaluated by other scientists in the field before publication. While this is generally regarded as the best system for sharing research, it isn’t foolproof. Academics are not paid to perform peer review and academic appointments depend on frequent publishing (“publish or perish”), so peer review is often not as thorough as it should be – especially for cosmetics where little of the research is publicly available.

Some phrases also used in non-academic life have a slightly different meaning in scientific literature. For example, most papers end by saying “further research is needed” (this practice is so common it even has its own Wikipedia page). This does not necessarily mean that we don’t have enough information in the body of evidence to make everyday decisions. The goal of that phrase is to ensure coherent research development and to justify further funding.

Reviews and article introductions can be inaccurate or incomplete, especially if the authors are not experienced researchers in that field.

While the abstract, introduction, discussion and conclusions are the easiest parts of a paper to understand, these often paint an overly complimentary picture of the importance of the paper and its findings. The method and results reveal the true nature of the study, along with understanding the existing research in that field outside of that paper.

Abstracts are often misleading and leave out important details – remember, their purpose is to entice others to read the full paper! Think of them as an advertisement for the paper.

What to do about the new information?

Applying research results to real life takes a lot of practice. It is necessary to understand the scientific model that the research is built upon. For example, if the participants in a clinical trial have white skin and are in their 20s, the results may or may not apply to other demographics. It would depend on whether these attributes will impact the outcomes, according to the scientific model established by thousands of other studies.

The vast majority of papers deal with tiny advancements in knowledge, and need replication to determine their reliability and applicability to other situations. There are often limitations that the authors do not mention, or may not even have considered.

This is why changes to legislation and other parts of our lives should not be based on just one paper. However, individual papers will contribute to a more complete understanding, with continuing research.

It’s a good idea to include limitations and point out possible misunderstandings when sharing information from a paper. Always include the reference so the audience can look at the source.

References

Spicer A. Explainer: what is an H-index and how is it calculated? The Conversation. Published May 21, 2015. Accessed May 16, 2023.

“Pay to publish” schemes rampant in science journals. Phys.org. Published March 22, 2017. Accessed May 16, 2023.

What is Journal Impact Factor? Elsevier Author Services. Published March 13, 2020. Accessed May 16, 2023. ‌

Pain E. How to (seriously) read a scientific paper. Science. Published March 21, 2016. Accessed May 16, 2023. doi:10.1126/science.caredit.a1600047

Lau S, Gonzalez J, Nolan D. 18.5. P-hacking. In Principles and Techniques of Data Science. Published 2020. Accessed May 16, 2023.

Reardon J, Crawford J, Allen KA, Walsh L. The peer review system is broken. We asked academics how to fix it. The Conversation. Published July 25, 2022. Accessed May 16, 2023.

Barbour V. Publish or perish culture encourages scientists to cut corners. The Conversation. Published September 22, 2015. Accessed May 16, 2023.

Hubbard KE, Dunbar SD, Peasland EL, Poon J, Solly JE. How do readers at different career stages approach reading a scientific research paper? A case study in the biological sciences. Int J Sci Ed, Part B. 2022;12(4):328-344. doi:10.1080/21548455.2022.2078010

Hubbard KE, Dunbar SD. Perceptions of scientific research literature and strategies for reading papers depend on academic career stage. PLoS One. 2017;12(12):e0189753.  doi:10.1371/journal.pone.0189753

Marinova M [@mila.phd]. What a scientist thinks everyone should know about reading scientific articles. Published January 11, 2023. Accessed May 16, 2023.