Writing the Materials and Methods Section Without Losing Your Mind
The materials and methods section is where most researchers either document everything precisely enough to reproduce their work or dump a wall of text that makes peer reviewers want to quit mid-review. I've written enough of these to know the difference, and more importantly, I know what happens when you skip the small stuff. Here's how I approach it now, after three revisions of my own methods sections taught me that "it goes without saying" is the most dangerous phrase in scientific writing.
What a Materials And Methods Example Looks Like in Practice
A materials and methods section needs to answer three questions in order: what did you use, how did you use it, and in what order did you do things. That's it. Everything else is decoration that reviewers don't care about and you'll delete later when space gets tight. I keep a running lab notebook entry that follows the same structure as the final paper. When I'm done with an experiment, I write the methods section that day while the steps are still fresh. This cuts my drafting time from about two hours down to roughly fifteen minutes because I'm not reconstructing timeline memory. The real work of method documentation is the doing, not the writing. For the materials portion, list suppliers with city and country. Not just "Sigma-Aldrich" but "Sigma-Aldrich, St. Louis, MO, USA." When someone copies your work five years from now, they will not know that Sigma changed distribution agreements twice in that window. A catalog number is better than a trade name. If you used a specific kit, include the part number. I learned this after a collaborator tried to replicate my ELISA protocol and ordered the wrong buffer formulation because I wrote "PBS" instead of "PBS, pH 7.4, with 0.05% Tween-20 (Sigma P4488)." That cost him three weeks and four hundred dollars in reagents.
For methods, write in past tense. Passive voice is acceptable here and often clearer because the action matters more than who performed it. "Samples were centrifuged at 12,000 × g for 10 minutes at 4°C" is easier to scan than "I centrifuged the samples..." because every methods sentence follows the same grammatical pattern. Consistency beats personality in this section. Include parameters that seem obvious until they aren't. Incubation temperature, duration, agitation speed, instrument model and firmware version if relevant, software with version number, statistical test names with exact p-value thresholds. The threshold detail is something people routinely skip. Write "p
0.05 was considered statistically significant" explicitly. Don't assume the journal or reviewer will fill in that blank for you.
Get the Full Details
Common Mistakes That Waste Reviewer Patience
The most expensive mistake I see is omitting the statistical methods entirely and burying them in a results paragraph. This is backwards. Statistical analysis belongs in methods, not results. The analysis plan should be stated before the data is examined. If you describe your statistics after showing the numbers, it looks like you chose the test that gave you the result you wanted, even if that wasn't actually what happened. Another issue is providing too much detail on standard procedures and too little on what you actually modified. If you used a published protocol with one change, cite the original and specify the deviation. "We followed the method of Smith et al. (2019) with the following modification: heating was reduced from 95°C to 60°C to prevent protein denaturation." This is efficient and honest. It saves space and gives credit where it belongs. I once had a methods section rejected because I described a custom-built apparatus without providing a diagram or sufficient dimensional detail. The reviewer couldn't determine whether the inner chamber volume was 50 mL or 75 mL from my description alone. I spent two days redrawing the setup with labeled measurements and submitting it as supplementary material. My workaround going forward has been to photograph every custom piece of equipment I build and include a labeled schematic in the supplementary files. This eliminates the back-and-forth with reviewers who ask for dimensions.
Software and Version Numbers Matter More Than You Think
When you list software, include the version. R is not the same as R 4.3.2. Python scripts behave differently across versions. ImageJ has been renamed to Fiji and both have different default behaviors for background subtraction. I had a replication attempt fail because I cited "ImageJ" generically and the person reproducing my work was on a version with a changed threshold algorithm. The results diverged by 18% on the same images. Version numbers are not bureaucratic noise. They are essential replication data. For statistical software, state the package and version alongside the test names. "Two-way ANOVA with Tukey's post-hoc test was performed using Prism 10.2.1 (GraphPad Software, Boston, MA, USA)." That sentence tells someone exactly what to do if they want to repeat your analysis. Anything less is an invitation for clarification requests during review.
When to Include Raw Data and When to Summarize
Your methods section does not need raw data. It needs enough detail that someone with comparable equipment and reagents could repeat the procedure. Supplementary materials are for raw data, extended protocols, and additional figures. Keep the main methods section lean. I target 800 to 1,200 words for most biology papers. If you're over 1,500, you're probably including something that belongs elsewhere or describing standard steps that need no explanation. Not every method has a bright future. Some protocols simply cannot be reproduced from a written description because the skill is in the hands. Cell culture technique is the classic example. You can write about media composition, passage ratios, and incubation conditions, but the moment-to-moment judgment calls—when a flask looks "ready," how gentle the trypsinization should be—are impossible to fully capture in text. In those cases, acknowledge the limitation directly rather than pretending the method is more reproducible than it actually is. Reviewers respect honesty more than false precision.
