The other month I was sitting down to read a journal article; the topic was dopaminergic signaling underlying both positive and negative motivational signals, which, all in all, seemed pretty darn interesting. Happily, I downloaded the article, and, feeling inspired and virtuous, downloaded the supplemental materials. Before settling in for an in-depth read, I skimmed over both files. Length of article: 6 pages.
Length of supplemental material: 34 pages.
In general, I am not a big fan to supplemental material. The above example is, admittedly, extreme, but anyone who has read a paper published since 2003 has experienced the frustration of wading through massive volumes of supplementary materials. As the daughter of an editor, I’m a fan of streamlined writing – I’m moderately irritated by tangential figures that clutter the logical flow of a journal article. Modest irritation becomes distinct dislike anytime supplemental figures vastly outnumber normal figures. Such unhappiness is only compounded when critical information (methods, interesting data, important analyses) is relegated to the supplemental material.
Normally, I ignore supplemental sections; when pressed for time, I am unwilling to synthesize the massive of information too often packed into the supplemental material section. As a reader, the large volume of supplemental material frustrates me, especially given a tendency for that material to be less carefully constructed than the prime-time figures. As a past (and future) author, I contemplate nervously the possible demands for supplementary results that reviews have (and will) make on me.
It turns out that the Journal of Neuroscience agrees with me on both counts.
The society of Neuroscience council came to the decision to remove the supplemental material section for multiple reasons, all of which I sympathize with as both a reader and writer. In the official notice, Maunsell notes that the volume of supplemental material associated with articles published in J. Neurosci. has grown exponentially since the section’s introduction in 2003, and appears on track to surpass the size ( in megabytes per article) of the main article itself (see figure). In the opinion of the journal, this sheer volume adversely affects the peer review process. Among the concerns cited is that the large volume makes it unlikely that reviewers will provide adequately in-depth evaluations of supplemental materials. Furthermore, the availability of a supplemental material section encourages reviewers to make additional (potentially excessive) demands for further analyses and experiments, the threat of which encourages authors to preemptively include extraneous, often tangential material in their initial submission. Furthermore, Maunsell and company are concerned with the tendency for critical material such as “methods that are essential for replicating the experiments, analyses that are central to validating the results, and awkward observations” to end up in the supplemental section. In their view, this “undermines the concept of a self-contained research report by providing a place for critical material to get lost.”
Maunsell states that the Journal of Neuroscience has considered, and rejected, several solutions to the problem, including setting limits on the amount and content of supplemental material, demanding thorough examination of material by reviewers, and official hosting of non-peer reviewed supplemental material. Therefore, in order to “maintain the integrity and value of peer-reviewed articles”, the Journal of Neuroscience has chosen to remove the supplemental material section, “requiring that each submission be evaluated and approved as a complete, self-contained scientific report”.
Authors will be allowed to list a URL pointing to supplemental material on a site they maintain, but this material will not be peer-reviewed, and will be labeled as such by the journal. One critical use for the supplemental material section is the inclusion of forms of data such as videos, which cannot be printed. To allow the inclusion of such data, the journal will allow authors to embed movies of 3-D models into the online PDF copy of their articles.
This is a major move on the part of an eminent scientific journal. Personally, I support their decision – and the responses from others in my lab have been similarly positive. I believe this move will force submissions to J. Neurosci to be constructed with more care. Furthermore, while the supplemental material section has the potential to serve a useful function, I agree with the journals attempt to reduce the reliance of authors on the presence of supplemental material to support their supposedly self-contained scientific stories.
What do you think? Is the wholesale removal of the supplemental material section a drastic response? Is the removal of the section from the Journal of Neuroscience less critical than if Science or Nature (who enforce stricter limits on article length) followed suit? Will this change make you more or less likely to submit to the Journal of Neuroscience? Would you be more likely to read/enjoy an article lacking supplementary figures due to official mandate? Please, discuss in the comments.
The full announcement can be viewed online at the Journal of Neuroscience: Maunsell, J. The Journal of Neuroscience, August 11, 2010, 30(32):10599-10600