Deconstructing the Methods Section – The Heart of Replicability

Almost always, when approaching the writing of a scientific manuscript, my starting point is the Methods section. For me, it's the part that flows most naturally. And I'm not alone in this perspective: a survey conducted by Ramírez-Castañeda (2020) on 49 Colombian academics in natural sciences revealed that many of them also consider this section the easiest to write. This is likely because it consists of clear, factual statements about what was done, how subjects were selected, how data were collected, and how these were subsequently analyzed, without much need for interpretation or extrapolation.

Nonetheless, the Methods section follows a rigorous order for presenting these facts to the reader. As with any well-structured text, information should flow from the most general to the most specific. Surprisingly, not all authors follow this logic, which can lead to confusion for the reader and hinder replicability.

Always remember that the primary goal of the Methods section is replicability. Every detail, every step you describe, must allow another researcher with similar resources and training to replicate your study exactly as you did. If someone cannot replicate it based on your description, the Methods section needs more detail.

Below, I share how I organize the Methods sections of the articles I have written about my research on howler monkeys—an approach you can adapt to your own field.

1. The "Where": Study Site Location

I begin by stating the location of the study site. Where was the fieldwork conducted? It's crucial to include precise geographical coordinates for the site. If relevant and visually helpful, you can incorporate a map of the general location, and perhaps a more detailed map of the study site along with the distribution of research subjects within it (see Figure 1, from Van Belle and Di Fiore, 2022, as an example of a compound figure). If applicable to your study, include essential site information in the text: size, dominant vegetation type, climate classification, etc.

Figure 1: Example of a compound figure showing the location of the study site and research subjects.

2. The "When" and "Who": Duration, Subject Selection, and Ethical Considerations

Once the "where" is established, we move on to explaining the study's duration and how subjects were selected. If your research involves repeated observations of subjects (e.g., in animal studies or data collection along transects), it is fundamental to specify the timeline of these repetitions. It is very helpful to include, either in the text or in a table, the total amount of time dedicated to each study group or transect. This allows the reader to assess whether data collection effort might have been biased towards specific groups.

In addition to temporal and specific details, it is essential to address ethical considerations. Do not merely mention ethics committee approval (which is indispensable!), but also any other relevant ethical considerations. For example, if you worked with animals, did you adhere to specific animal welfare guidelines? If you worked with humans, was informed consent obtained, and was confidentiality guaranteed? Specify the approval protocol number from your ethics committee.

Natural sciences research, especially fieldwork, often presents unexpected challenges (extreme weather conditions, equipment failures, unpredictable subject behavior). If these challenges affected the consistency of your data collection or led to protocol modifications, describe them honestly. Explaining these situations and how you managed them (or why you could not mitigate them) adds credibility and transparency to your work. Clarity here is key for readers to understand the context of your methodological decisions.

3. The "How": Data Collection Protocols and Material Details

Up to this point, we have established the temporal and spatial framework of data collection. Now it's time to delve into the specifics of the protocols used. My experience shows that research projects often involve collecting different sets of data, each with its own method and frequency. There are two main ways to organize the presentation of these protocols:

• By Relevance: From the most relevant protocol to your scientific question to the least relevant.

• By Frequency: From the most frequent data collection protocol to the least frequent.

Let's consider an example from my own research: I studied the effect of social and ecological factors influencing the participation of group members (males and females) in vocal bouts in howler monkeys. I collected data on: 1) who participated at 1-minute intervals throughout the vocal bout, 2) the behavior of all group members at 15-minute intervals throughout the observation period (full-day follows), 3) two fecal samples per individual for genetic analyses of paternity and relatedness, and 4) proximity data between pairs of individuals at 15-minute intervals throughout the day.

In the first scenario (by relevance), we would start with the data most pertinent to the research question, the participation in vocal bouts, followed by group behavioral scans and proximity data, and finally the collection of fecal samples. In the second scenario (by frequency), we might start with behavioral and proximity data (collected every 15 min), then participation in vocal bouts (every 1 min), and finally fecal samples (two per individual). Both approaches are valid; choose the one that best suits your study and your writing style, and that ensures maximum clarity for the reader.

A crucial point: if you collected more types of data than you actually used in the specific study you are reporting on, it is best not to mention the protocols for collecting those unused data. Doing so will only confuse the reader, who might think there was a reason this information was presented and thus assume they missed something.

When mentioning specialized equipment, lab kits, reagents, or software, be sure to provide complete information: manufacturer's name, city, and country of origin. This allows other researchers to acquire exactly the same materials or programs. For example: 'DNA samples were extracted using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany)'.

4. Subsections and Consistency: Organizing Information

It is very common and advisable for authors to use subheadings to organize the Methods section. Common examples include: "Study Sites and Subjects", "Data Collection", "Data Analysis", "Genetic Analyses" (or any other specialized analysis), and "Statistical Analyses".

When using these subsections, make sure to divide the information logically and consistently. For example, if you have subheadings for "Data Collection" and "Data Analysis," do not include explanations of the criteria used to include data in the analyses within the "Data Collection" section, as this is technically part of data processing and analysis. This is a common mistake. If the information flows best by grouping data collection and analysis, be sure to use a subheading that indicates this (e.g., "Data Collection and Analyses").

The Data Analysis section must explain in an orderly fashion how raw data were processed and transformed into the variables that will be included in the statistical models. If you coin a specific term for any of these variables (e.g., "percentage of mutual vocalization"), use this same term consistently throughout the document. Switching it up by using synonyms (e.g., "pair vocalization") merely for variation is not recommended, as readers might interpret this as a different variable that was not explained.

Furthermore, excessive use of unconventional abbreviations is not recommended. As a researcher and author, you will have internalized these abbreviations and are fully accustomed to their meaning, but for the reader, this might be the first time they encounter your abbreviation, requiring them to memorize its meaning and hindering their full understanding. Therefore, when not strictly necessary, it is best to avoid introducing new abbreviations.

5. The "How it's Tested": Statistical Analyses

Finally, in the Statistical Analyses section, it is imperative to include the software used for the analyses, along with appropriate citations. Statistical models can be very complex, so clearly state:

• What was assessed with each model.

• What the response variable and predictor variables were.

• Whether there were any interactions between the predictor variables.

• Whether the model also included random variables.

It can be very constructive to include a table that schematizes these models. If not feasible in the main manuscript, consider adding it to supplementary material, perhaps even in the form of R scripts (if you use R) so that other students and colleagues can visualize how the models were constructed.

Additionally, it is highly constructive to include a table with the descriptive statistics of the predictor variables used in your models. This practice offers the reader all relevant information concisely and easily digestible, avoiding unnecessary text length. As an example, Figure 2 shows a table extracted from one of my publications (Van Belle & Estrada 2020), which summarizes the descriptive statistics of the variables used in the statistical models.

Figure 2: Example of a table summarizing the descriptive statistics of the variables used in the models.

Regarding Voice in Methods Writing: The Methods section should be written using past tense verbs, as it describes actions that have already been performed. Traditionally, the passive voice ('Samples were collected...') has dominated the Methods section to maintain objectivity. However, more and more journals and style guides encourage the use of the active voice ('We collected the samples...' or 'The team collected the samples...') when this improves clarity and conciseness, without sacrificing objectivity.

References

Ramírez-Castañeda V. 2020. Disadvantages in preparing and publishing scientific papers caused by the dominance of the English language in science: The case of Colombian researchers in biological sciences. PLoS ONE 15(9): e0238372. https://doi.org/10.1371/journal.pone.0238372

Van Belle S, Estrada A. 2020. The influence of loud calls on intergroup spacing mechanism in black howler monkeys (Alouatta pigra). International Journal of Primatology 41:265–286. https://doi.org/10.1007/s10764-019-00121-x

Van Belle S, Di Fiore A. 2022. Dispersal patterns in black howler monkeys (Alouatta pigra): Integrating multiyear demographic and molecular data. Molecular Ecology 31(1):391-406. https://doi.org/10.1111/mec.16227