Monday, 25 February 2013

ASA 2013: Jeremy Frey on Reputation, Responsibility, Reproducability and Reuse - A Scientist's Perspective on the Role of Publications

Jeremy Frey addresses the ASA conference audience
Jeremy Frey is Professor of Physical Chemistry at the University of Southampton. He covered a scientist's perspective on the role of publications and pointed out that the open access debate is actually an issue about the economics of who pays, as someone has to. However, he believes that there are different consequences around accessibility when you consider who does.

The consequences of not having open data is that your work cannot be checked by someone as bright as you. This could be an error made in good faith. There are ways of validating data, but they are dependent on having access to the data and a willingness to host that data. In his opinion, the Royal Society report has come up with the best phrase so far: 'intelligent open access to data'. It's not just access to the data, but it's also critical to record somewhere how to get use out of the data.

Some projects need large amounts of data from the literature. Access can be an issue. A digital repository is one answer, but does this lead to subversive and furtive sharing and exploitation of data in a virtual space? Can people understand how to use it?

One of the issues is that people in the scientific community put all emphasis on the answer, not the working out. And yet the process of coming to an answer is the most important part: 'methods are as important as the data'. The BBC 'Climategate' campaign is a good example of the benefit of doing this. Even Faraday in his notebooks included a huge amount of detail about how he went about his experiments.

What are researchers trying to get out of the whole publication process? Dissemination, reputation, and advancement. The scientific information supply helix was historically from researcher to literature to abstraction/search to literature. Currently you can add open notebook science. This makes the flow go from researcher to the web(?) to search to researcher, but this is rarely the only dissemination route.
There is a need for validation, traceability and accountability for recognition, promotion and future funding.

There are some journals that would limit the number of references within article as they are constrained by page size. This is a real problem in the interdisciplinary world. However, this is shifting as online removes space restriction. 

There are a number of problems with interdisciplinary research and dissemination. The methods and emphases varies with discipline:
  • journal versus conference
  • pure versus applied
  • research versus application
  • authorship and acknowledgements
  • who goes on the paper?
  • who is mentioned in the press release?
  • do you acknowledge the technician?
  • e.g. maths is like humanities - student works on their own while science is team work.

He believes that researchers also need to reflect on how they communicate the data. The Smart Tea Project was a way to create a harness to record data as you go along, but it was a bit overbearing for students. Then they moved to LabTrove which has more of a blog style with a highly interlinked system. Other issues to consider are diversity and disability. Another example is the BlogMyData Project. This type of project has an impact on researchers, can lead to a higher quality record, easier collaboration, improved planning, better proactive interaction and can run in a completely open or closed way.

Access to the discussion and the means of reproducibility of the analysis is really important to the way the community will be working in the future. Things have to change and at pace.

No comments:

Post a Comment