29 September 2017

If you love science, love methods sections


Mensh and Kording (2017) have a new paper on scientific writing. It’s very good. I agree with most of their advice. But not this.

You should also allocate your time according to the importance of each section. The title, abstract, and figures are viewed by far more people than the rest of the paper, and the methods section is read least of all. Budget accordingly.

No. Do not skimp time spent on your methods sections.

I get where this advice is coming from. It’s the same sentiment that has lead some journals to put their methods section at the end, or to stuff parts of papers away in online “supplemental information.”

But we read papers for lots of different reasons. I read lots of papers that are only tangentially related to me out of curiosity. But when there is a paper that is in my field, that I need to understand, I dig deep into those methods sections.

I’ve run into so many cases where something that looked like a solid finding looked very shaky once you realized how the data were collected. While Mensh and Kording are right that few people read the methods, it neglects that those who do are going to be the most intense and critical readers.

A recent feature in Nature showed that weak detailing of methods was leading to irreproducible results (my emphasis).

In one particularly painful teleconference, we spent an hour debating the proper procedure for picking up worms and placing them on new agar plates. Some batches of worms lived a full day longer with gentler technicians. Because a worm’s lifespan is only about 20 days, this is a big deal. Hundreds of e-mails and many teleconferences later, we converged on a technique but still had a stupendous three-day difference in lifespan between labs. The problem, it turned out, was notation — one lab determined age on the basis of when an egg hatched, others on when it was laid.

The article give multiple examples of how hard it is to standardize methodologies, but how important it is to achieving consistent results. This older Drugmonkey post, makes a similar point.

The methods section is where the rubber meets the road in terms of actually conducting science. If you don’t get that methods section right, you’re wasting the time of people who come afterwards.

References

Mensh B, Kording K. 2017. Ten simple rules for structuring papers. PLoS Computational Biology 13(9): e1005619. https://doi.org/10.1371/journal.pcbi.1005619

Lithgow GJ, Driscoll M, Phillips P. 2017. A long journey to reproducible results. Nature 548: 387–388. https://doi.org/10.1038/548387a

External links

The most replicated finding in drug abuse science

1 comment:

klab said...

As one of the authors of the original study I want to wholeheartedly agree. The methods are crucial and a paper without good methods is worthless. Our point was exclusively about the writing. As long as the methods section completes all methods, it is fine if it is a bit dry. Once I read someone's methods section i am really motivated to dig hard. In other words, the content must all be there, but the writing (and local structure) is not that critical in my opinion.