Supplemental materials or no?

I was surprised when I read this DrugMonkey post on J Neuroscience's ending supplemental materials. In fields without significant open data repositories with required deposit prior to publication, supplemental materials may be the only way to get the data to check the work or to build upon it (authors aren't very good at replying to requests to share data - studies show).

I really don't know anything about neuroscience, is this field different or is this coming in other fields?  I know that astro and optics journals have been expanding their ability to take supplemental materials such that they are preserved and accessible.  Here are some of the reasons found in Drug Monkey's post:

  • they were representing the data as peer reviewed, but it isn't reviewed to the extent the text is and what does peer review of data mean anyhow
  • there's an arms race among authors and reviewers to throw in everything but the kitchen sink proactively to not be criticized and to request that more data and more experiments be included in the supplement
  • the text should stand on its own merits

Anyway, I hadn't heard this view and I didn't know this is the way it was working in this field. I kind of thought the paper was reviewed on its own merits and the supplemental data was like a bonus track added later. Once again, I'm thinking of astro and optics journals. So is this view common or does it work differently in different fields?  (Is there a paper the view of supplemental data in diverse areas of science?)

2 responses so far

  • DrugMonkey says:

    Read the editorial. See if you think the exceptions they mention are what you are imagining. This is not about archiving huge -omics or gene array tables. This is about normal data and methodological description being shunted off.

  • I agree with DM. It's not stuff that's traditionally been placed in appendices; it's really a second, substandard, not always easy to find paper.

    And speaking of -omics, or at least genomics, NIH now requires public deposition of raw data (or somewhat raw) and sequence.

    Kill supplemental methods. Kill them now.