The OpenScience Project

I recently sent some advice to a colleague who is coming up for tenure at another university.  He’s quite well known in the Open Science community and is trying to figure out how best to make the case to his tenure committee that the open science contributions he has made in addition to his traditional journal publications are important.  We’re talking some major contributions here — lab protocols on OpenWetWare, open lecture materials on slideshare, data files released with CC0, videos of lab protocols on Benchfly, and he’s a regular contributor to science discussions on FriendFeed.

The advice I gave him was basically to make the committee’s job of measuring these contributions easier.  Here’s the advice (in a slightly edited form):

The audience for most tenure documents (and particularly the external letters) is a committee of non-specialists that advises the provost or other high-ranking administrator.  These committees are often somewhat skeptical of departments and candidates and are looking for external validation of what they are reading in the tenure dossier and the packet prepared by the departments.  They are swayed by real experts in the field (named chairs at other institutions, national academy members, people at top 10 institutions) and by things they can measure (publications, h-indices, grant money, citation counts). If you want to add a non-traditional contribution to a tenure dossier, you should also include a way of measuring the importance of that contribution.

First, if the rules of your institution allow it, make sure there is a strong defense of open ways of doing science in your dossier (1-2 paragraphs or so).  Make the case that it is important to consider non-standard contributions even though previous tenure committees did not.

Use as many metrics to back up your contributions as you can. Make a case that each of your software releases counts as much as a full publication, and use download statistics as if they were directly comparable to academic citations.  List external users of your software as if they were research collaborators, because they are!  If you can collect them, include download statistics on open contributions to sites like OpenWetware and Wikipedia.

If your institution’s rules allow it, make sections directly under your publications for  “Published Datasets”, “Contributed Software”, “Published Protocols & Notebooks”, “Scientific Videos”.  In each section, list authors, a title, description, and URL of the resource you have contributed along with a count of downloads or views, and a list of other groups using your data. Make this look as much like your publication section as possible, as you can then make the argument that these things should be treated with a similar weight to traditional academic publication.   Provide the metrics in the document so that your committees aren’t guessing about how important something is.  I can’t emphasize this enough – citation counts are easy for a committee to dig up – download stats are harder.  Do the measurement work for your committee and they’ll make the assumption that your metrics are important.

So that’s the advice.  I’ve been involved in a few internal tenure discussions, and the metrics are always important.  If there isn’t an easy analogy to something in my own experience, I look to the candidate’s documents and the external letters to tell me why something matters.