On 2014-07-29 09:08, Andrea Splendiani wrote:
I agree with you. When I hear about "research" I tend to think about Life Sciences research (my bias). In this context, I see a lot of advocacy for reproducibility of research. But when I see what people do, for most of them, data elaboration is only one part of the process. So I was wondering, in this context, how relevant "publication" is on the reproducibility of results. Probably in proportion not much.Otherwise you are right. In a more computational/engineering oriented areas, where reproducibility only depends on information... information should be there.
I do not think there is a need to give different treatments to the types of science - is there? (It is an honest question, I am not an expert in this area). I figure that something either follows the scientific method, or it doesn't.
If how something is published hinders researchers from taking a closer look at what was done, should that not be a concern?
If a Web researcher can not deliver their work in full, it would be irresponsible to think that it is up to interested parties to get it in full. It either exists as complete as possible, or it just doesn't. And, frankly, if some research information can not be obtained or reproduced easily, it does not need to be cited. Otherwise, I do not see how that qualifies as science.
So, when publicly funded research gets summarized in a PDF, accessible through the publisher's site, and all of the dots to reproduce the work is unavailable at ease, then are we using the term Web Science in an honest way?
PDF is a desktop native document. It is true that all sorts of stuff can be jammed inside. Its supporters are trying to make sure that it plays well with the Web. While those are good efforts (making the best out of the situation), it is like placing a band-aid on a severe wound. PDF on and off the Web breaks good user-experience patterns no matter how you look at it. One can not simply navigate through a piece of research publication to all of its atomic parts in an ubiquitous fashion. PDF can not deliver that. It is more likely that your off the shelf Web browser can. Supporters of PDF will come out and say that, well, it can have hyperlinks, or it can even render in some Web browser. Do they really think that navigating between a PDF document (whether on desktop or viewed in a Web browser) and a hypermedia resource is considered good design?
Why shouldn't we remove all barriers to reproducibility? -Sarven http://csarven.ca/#i
smime.p7s
Description: S/MIME Cryptographic Signature
