Correspondence

Nature 455, 461 (25 September 2008) | doi:10.1038/455461d; Published online 24 September 2008

Big data: teaching must evolve to keep up with advances

Samuel Donovan1

  1. Department of Biological Sciences and BioQUEST Curriculum Consortium, 4249 Fifth Avenue, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, USA
    Email: sdonovan@pitt.edu

Sir

The Editorial introducing your 'big data' special issue ('Community cleverness required' Nature 455, 1; 2008) highlights the entrenched "institutions and culture of science" as barriers to "taking full advantage of electronic data". Science education is part of that entrenchment, so more thought should be invested in the tight relationship between the advancement of science and our systems for educating future scientists.

The growth of web-based, collaboratively developed and publicly accessible research data provides an important opportunity to transform the ways in which students are taught science. Failure to challenge the status quo in science education will lead to an even greater divergence between scientific practice and education, which could potentially undermine the continued rapid evolution of scientific research. If the use of cyberinfrastructure is going to be "as integral to the practice of science as publishing and reading papers", as Lincoln Stein has suggested (Nature Rev. Genet. 9, 678–688; 2008), then we must invest in developing the human infrastructure necessary to realize these new approaches to doing science.

The opportunities to engage undergraduates in realistic scientific-research experiences using 'big data' and other e-science resources are as diverse as the research itself (see, for example, the BioQUEST Curriculum Consortium site http://bioquest.org). Unfortunately, computational infrastructure and successful model projects are still not adequate to ensure that the next generation of scientists will be prepared to work in a rapidly evolving research environment.

Most undergraduate science educators completed their graduate training more than 10 years ago and so have limited experience in working with modern e-science resources. Yet today's students are entering a scientific workforce in which they are expected to have skills in areas such as data mining, modelling, visualization and annotation.

Given the growing disparity between the rapidly evolving world of research and an entrenched culture of science education, the future of science depends less on "cleverness" than on our commitment to preparing future scientists to work with 'big data'.

See also Big data: open-source format needed to aid wiki collaboration.

Contributions may be sent to Email: correspondence@nature.com.