Anton’s quick and dirty guide to publishing data

A journal has asked you to make your data publicly available as you submit your article.  This is a guide to how you can meet that requirement, not get your data stolen, and maybe even get more credit for your research.

Now, you should have thought about this when you started collecting your data, and you’d already be ready.  But you didn’t.  This should take a few hours to an afternoon to get together.

Step one, prepare your data.

This is the bit that takes time.  You need to get your data into a format that others can look at it, and satisfy themselves that you probably didn’t make it up.  One day there will be proper data reviews, like peer reviews, but they are really rare.

Wedging it from R or SPSS into a spreadsheet is the easiest.  After some cut and paste (or exporting it into excel format) create two new tabs.  The first should be called README, and it can have your name, permanent email address, the name of the paper, the abstract, keywords and acknowledgements. The second tab has a description of EVERY FIELD in your data.  So, what does each row represent, and what is each column.  (The columns have names right?  No?  Go put some in!)

I’m assuming your data is anonymised, and that any ethics approvals you sought said you’d publish the data.  See – thinking about this first would have been a good idea!

Step Two, publishing.

Go to your account.  Whoops, I mean, make a figshare account.  If you had an account then you have already done this before, and you wouldn’t be in this positon.

Create a new item.  We’ll create the record, and upload our file second.  Use your abstract for the description.  You only get a CC-BY license with a free account, but that the right one, so don’t worry about it.  Reserve a DOI, and and generate a private link.  Put the DOI in the README part of your data.

Now you save the item, but don’t check publish.  You can send the private link to the publisher, so they can see the data, but it isn’t in public and no one can steal it.  You can publish it when your paper is accepted (and you can use it to cite your data in the paper itself – see, you automatically get a citation as soon as you publish!  Hooray!).

Next time, you’ll think about getting your data in a publishable format while you collect it, right?

Leave a Reply

Your email address will not be published. Required fields are marked *