Table and figures are a good thing. The tables, in particular, contain all the data that everything else in the paper is based on (I don't review pure theory papers. Theorists should be thankful). Frankly, I don't understand the trend towards depositing data and methods, rather than publishing them. Depositing a machine readable version of datasets IN ADDITION to publishing them (or a representative sample of data, for really large sets) is a sensible thing to do, but in general we would all be better off if papers contained data and results, and the speculative waffling, overly broad impacts, and long bows stretched to make the paper seem to appeal to fields it doesn't really effect were all banished to the online supplementary material.
Why do you differentiate between "publishing" and "depositing" data? Nowadays, the only distinction between the paper and the supplementary material is that the latter is only available online - and I suspect that the vast majority of papers are now accessed via the internet. I understand that some journals indicate that reviewers don't worry as much about the text in online materials, but surely the raw data would receive the same scrutiny whether it is available as a supplement or as a table in a paper. Frankly, I'd banish all data tables that don't directly aid the text when being read to supplementary material. Besides, publishing pages in a dead-tree journal costs money, and it's better spent on text and figures than long columns of numbers.
ReplyDeleteIn many cases, "supplementary" is "stuff we'll stick on the web sometime after the paper comes out." Especially for advanced online previews. I don't care what form the data takes- machine readable is in fact usually easier, but it need to be part of the overall logical argument, some something tacked on later, or missing from early publication and/or review.
ReplyDelete