I’m well into my Scientific Writing course now, and I’ve just given the lecture that consistently annoys my faculty colleagues the most (well, it annoys many of them). It’s the one on writing the Methods section, and it’s heterodox in two rather different ways. This lecture stands out a bit – I don’t think my approach to IMRaD structure, or the content of the Discussion, or outlining, or writer’s block is all that different from the approach anyone else might take. But the Methods is different.
I said what I teach about the Methods is heterodox in two different ways. Let’s start with the active voice. I teach my students to write in the active voice (predominantly, not exclusively). That aligns perfectly well with virtually every bit of expert advice, as well as with the preference of almost every journal in my field (Biology)*. So in a way my teaching is perfectly orthodox – but every year, I ask my students how many of them have been taught to write in the passive voice and to never, ever, write “I” or “we”. This year, as usual, almost every hand went up. This is an interesting situation: experts are strongly agreed that we should write in the active voice, and yet we’re still teaching undergrads the opposite. This is a source of huge frustration for my students, as I’m asking them to unlearn what they’ve been taught. It’s a source of huge frustration for some of my colleagues, who then try to edit the active voice back out of the theses students write while (or after) taking my course. And it’s a source of huge frustration for my students – again – if they revise back to the passive at their supervisors’ behest, and then have me mark it wrong in my course!
I don’t know how we get past this. It seems that writing is different from anatomy, or genetics, or biostatistics, in the sense that we’re happy to let someone teach their opinion about it, whether or not that opinion has any actual knowledge behind it. And the preference for the passive is still widespread – perhaps because if all you do is read the past literature, that’s what you see. So I’ll continue, on this point, to be orthodox from the point of view of expert advice, but heterodox with respect to many of my teaching colleagues. And I’ll continue to be mystified about why that combination can exist.
The second kind of heterodoxy relates to the level of detail that belongs in a Methods section. We tell each other all the time that a Methods section should include enough detail that a reader could repeat the experiment for themselves. And most (but not all) writing guides concur with this. However, descriptive studies of what scientists actually do** find that this level of detail is rare. And I think that’s right, because replication attempts are very, very rare and therefore the vast majority of readers don’t need that detail – and thus are poorly served by a long, tedious Methods section that includes it. With this logic (developed in more detail here) behind me, I teach that a Methods section should include enough detail to establish your credibility as an investigator and enough to let the reader understand the results you’re presenting. Further detail, I suggest, is best included in an online supplement, where most of your readers can efficiently ignore it.
I guess in a way I’m only mildly heterodox on this point. I’m teaching what my colleagues (mostly) actually do, rather than what they (mostly) say we should do. And while I’m teaching much less detail in a Methods section than many writing guides recommend, I’m not teaching that one should hide the detail – only that one should present that detail in a supplement, so that only readers who need it have to grapple with it.*** Nonetheless: my students are often surprised, and my colleagues are sometimes shocked. And I think there’s an interesting phenomenon underlying this: the claim that detail for replication is important to science sounds so reasonable – as long as you don’t actually think about it carefully. As scientists, we’re very used to thinking carefully about statistics, and experimental design, and pipetting technique. It often seems that we’re not so used to thinking carefully about writing, philosophy of science, and the history and culture of scientific practice.
Next week in my course: the Results and Discussion sections. Everyone can breathe a sigh of relief – there’s less controversy there, so I don’t need to pick fights
© Stephen Heard January 31, 2023
*^I’m not going to provide a literature review here, or a big table of journal instructions to authors. If you doubt that active-voice writing is the modern best practice, you can start with my own writing book, The Scientist’s Guide to Writing, where I explore the history and rationales for our shifts from active to passive and back again. You can then continue with other writing guides and the literature they cite. I’m afraid I’m going to end this footnote with a sigh of frustration, though. I’ve found that those who don’t already believe that active-voice writing is a best practice are pretty consistently loath to do any reading about the matter, and dismissive of what they read when they do. Sigh.
**^While I won’t review the literature here, you can see The Scientist’s Guide to Writing and also citations in this older post as a way to start exploring this topic. I suspect you’ll find it as fascinating as I do.
***^This is almost exactly what happens with the Methods-last format used by many cell- and molecular-biology journals, plus some high-impact general biology journals. Nobody expects most readers to pay attention to the Methods section in a methods-last paper, which is why you’ll find enough methods information to understand the results in the Introduction and Results sections.
Steve, while you and I disagree on the passive voice in Methods section (I think methods should almost always be written in the passive voice, but that’s a matter for another discussion), I do agree that detailed methods should be put in Supplemental Material. The main body of the paper should give enough detail to let the reader know how you went about the business of gathering data. Science and Nature have been doing this for some years now, and I also ask authors to do it for The Scientific Naturalist series in Ecology, of which I am the founding editor. It shortens the paper and helps it flow more smoothly between the ideas, problems, and hypotheses in the Introduction and the Results and Discussion of those same things. In other words, better cadence and hence better writing.
LikeLiked by 1 person
As a medical editor who edits many, many journal manuscripts, I am very grateful to you for teaching this way. Thank you for helping to make science writing much more useful and even enjoyable.
I usually write in the active voice but I had a supervisor who would change all that back to passive and say it was too informal! But, sometimes it’s too much effort to argue about it, so I just went along with those suggestions at the time.
Yep. Your first point keeps surprising me too, Steve. My 7th grader and college freshman are both being taught to write their lab reports in the passive, decades after the convention was changed at the journal level. It’s hard to fight legacy convention.
I also recommend to write in the active voice, and when I ask students who write in the passive voice (which is, like, all of them) whether it’s a personal preference or they had been taught this way, the response is usually that the’ve been taught this way, probably in their first semester. I don’t make them change it but I always make this recommendation, as I believe it’s a matter of style and as long as the message gest across and can be understood, it’s fine. (And I think that in Portuguese the passive voice is less cumbersome than in English. I really, really hated it when I had to translate something in the passive voice into English.)
Regarding how detailed are the methods, the way I think about it is to include enough information that people may understand what has been done and not include details that would not change the outcome. I like when uncommon methods are briefly explained, so as to not make people read five papers to understand what a method does. And I recommend to not include too much detail on things that would not make a difference. For instance, does it matter whether the data were organized in Excel or in another spreadsheet software? Probably not, unless there is some bug related to this kind of data. But it does matter which statistical package was used for analysis because of different algorithms.
Or, it matters what was the distance between the transects used to sample vegetation and the quadrat size. It probably does not matter whether the transects were marked with flags, wooden poles or plastic poles (unless the study organism may be affected by them) or whether the trails were opened with a machete, hatchet or with one’s own bare hands. 🙂
LikeLiked by 1 person
Nicely explained on methods detail, Pavel!