Research, Data and Publication

The evolution of the internet and the technologies for exploiting it are creating major changes in the availability of data, its use and the way it is reported. They are very important for creating a more productive and transparent society by making smarter use of data.

1. We are measuring things that were previously not measured, eg numbers of people reading articles on the web, rapid growth in surveys on opinion and behaviour
2. People are responding in real-time to data accessible because of the internet, eg traffic flows, weather warnings
3. Data previously collected and held by a single government department, are now being shared between departments and layers of government and often released for public scrutiny and use (see, for example Sites for data ). The UK White paper on Open Data talks about "making smarter use of data" and "a truly transparent society". See also a 2011 report by the Australian Government Issues Paper 2: Understanding the value of public sector information in Australia and a newspaper article Data is not a dirty word
4. Data collected about an individual by many different groups are able to be easily synthesised and exploited. Examples include (a) electronic health records and (b) integration of commercial transaction data and geographic location data. Many privacy issues arise and are yet to be resolved.
5. Newspapers are increasingly undertaking and publishing data analysis. Examples include the New York Times, Guardian, The Australian and The Age/Sydney Morning Herald. See Data Journalism links
6. Scientific journals are providing online repositories for 'supplementary material' which ensures that more data, software and details of analysis are available for use and scrutiny.
7. New research methods (particularly in areas like statistics) are being published not only as papers and algorithms but also as integrated packages (see repository of R packages for more than 4000 packages!). This is allowing these methods to be used and tested extremely quickly - often before a paper is published.
8. There are questions about the long-term commercial viability of scientific journals, based in part on the current experience of newspapers and book and music publishers and their rapidly changing business models
9. The increasing availability in shorter time of data and ideas is changing the nature of research partnerships and collaboration.

The links below explore some of these issues in more detail. They are only a fraction of the total material available and the volume is growing quickly. They have very significant implications for statisticians. Some statisticians are obviously active in participating in these changes though few statisticians are active in discussions or deliberately leading the change.

A session was held at the World Statistics Congress in Hong Kong in August 2013. See http://www.isi2013.hk/en/scientific_list_Aug_30.php#2 - Session IPS108. Links to the abstracts and some of the papers are given there. Papers based on these presentations are to appear in the Statistical Journal of the IAOS in June 2014.

Of course there are risks in pulling different sorts of data together. Are there biases inherent in the data collection process? Do variable selection methods lead to results which cannot be reproduced? These risks are presumably greater with observational data than data arising from designed experiments. See for example

• The Science of Getting It Wrong: How to Deal with False Research,
• The Truth Wears Off or
• Why Smart People are Stupid.
• Also, for example, David Hand on Classifier Technology and the Illusion of Progress

There is a discussion at http://www.straightstatistics.org/article/lot-science-just-plain-wrong which points to a paper by Begley & Ellis in Nature and Young & Karr in Significance

There is a growing activity on making research reproducible. On computational and statistical research, see http://researchcompendia.org/about/ and references given there.

1. Raise standards for preclinical cancer research, by C. Glenn Begley and Lee M Ellis, Nature 483, pp 531-33, 29 March 2012

2. Deming, data and observational studies, by S. Stanley Young and Alan Karr, Significance, September 2011, pp 116-120

Data Journalism

Reinventing Discovery

Michael Nielsen Blogs - Data Driven Intelligence and the impact of networks on research

UK White paper on Open Data

Future of Research Publication

Finch Report (2012) Accessibility, sustainability, excellence: how to expand access to research publications

See also a report on a 2 day workshop, "Implementing Finch" organised by the UK Academy of Social Sciences

Kent R Anderson (2000) From Paper to Electron - How an STM Journal Can Survive the Disruptive Technology of the Internet

Murray Krantz (2003) The Future of Research Publishing: The eReport and the eJournal

Michael Nielsen (2009) Is scientific publication about to be disrupted?

Malcolm Gladwell (2009) Priced to Sell

Force11 http://www.force11.org/white_paper