I want to note that basically everything Joe says for solar physics data also holds for Earth sciences data of various types here at the National Snow and Ice Data Center, though we do allow people to opt in to register to be notified of updates to data they are interested in (the second item on Joe's list below) and are working on some of the other types of suggestions too…  I will note that NASA Earth Sciences is headed towards requiring everyone to register to use their systems which seems somewhat counterintuitive since those data are truly supposed to be open data.

Ruth Duerr 
National Snow and Ice Data Center
University of Colorado at Boulder


On May 20, 2013, at 7:11 AM, Joe Hourcle <[log in to unmask]> wrote:

> On May 20, 2013, at 8:31 AM, Susan Manuel wrote:
> 
>> Hi all,
>> 
>> At Loughborough University we have been having some discussions about ways of assessing the usage and impact of data shared online. One of the things we talked about was having a registration system for accessing research data and we wondered what you think of this. We can see some obvious benefits to knowing who is accessing and using these data, particularly if individuals can be contacted and are willing to provide a testimonial/impact case study. Other benefits might be in providing evidence of use/impact to secure on-going institutional support for RDM and contact with data users may identify opportunities for future collaboration. However, there are also potential drawbacks to a system whereby users have to register for access to data:
>> 
>> 
>> ·         Would potential data users regard the request to register as a barrier to accessing data sufficient to put them off?
> 
> In the community that I work in (solar physics), yes.
> 
> You might be able to work something out so that the person logging in gets a benefit from it.
> 
> For instance, if you log in :
> 
> 	... you can set up alerts to 'watch' for new data that matches your desired parameters.
> 	... you'll be notified if something's done to deprecate the data you've downloaded.
> 	... you can select the data you're interested in, but then come back later (or from another computer) to download it.
> 	... you can set up identifiers to allow others to easily download some grouping of data.
> 		(so that you can then use that identifier in your published research)
> 	... you can set various preferences (how to display results, what search options are shown by default, etc.)
> 	... you can request that the data be transformed / repackaged, and then notified when it's ready to download.
> 
> ... but I'd still have a way to download the data anonymously -- you can always track by downloader IP address to get a general idea of who's downloading, even if it's not the specific person.
> 
> 
> 
>> ·         Similarly, how would people respond to having to provide their details for each dataset they wish to view/download?
> 
> 	It's even worse, particularly if you ask them for any significant amount of information, as you just end up wasting their time each time they go to download.
> 
> 	You also run into issues with *when* you ask them for their information.  (before they start searching, or when they're just about to download).
> 	
> 
>> ·         Would this be regarded as non-compliance with funders’ policies?
> 
> 	It might.  You can always get around this by making registration optional.
> 
> 
>> ·         Would managing registrations place too much of an administrative burden on the University?
> 
> 	For the US government, depending on what you collect, it could be considered 'personally identifiable information', and have to be stored & tracked differently, as it changes the rules of how we're supposed to secure our machines.
> 
> 	We intentionally *only* collect e-mail address when it's necessary for asyncronous processing / packaging, and dispose of it promptly.
> 
> 	(our system might also be influenced by the time in which it was built, due to the 2005 astronomy incident with Haumea : http://en.wikipedia.org/wiki/Controversy_over_the_discovery_of_Haumea )
> 
> 
>> ·         Would the administrative burden of managing registrations be offset by a reduction in workloads associated with the REF in providing evidence of impact?
> 
> Maybe.  But some of the proposals for citation of data would also reduce that burden.
> 
> For instance, the 'Landing Pages' proposal from the technical breakout of the BRDI meeting on data attribution & citation:
> 
> 	http://virtualsolar.org/citation
> 
> (basically, data providers need to make it easier for researchers to cite their data, by giving clear instructions on how they want it to be cited, and assign DOIs so that journals are more likely to accept it as a citation)
> 
> 
>> We would be interested to hear what your views are on registration for access to research data.
> 
> Our group just doesn't do it.  But we also don't accept any embargoed or otherwise restricted data.
> 
> 
> -----
> Joe Hourcle
> Programer/Analyst
> Solar Data Analysis Center
> Goddard Space Flight Center