Comments on: Big Data: Opportunities for Computational and Social Sciences

By: Boris Shakhnovich

Boris Shakhnovich — Sat, 23 Oct 2010 16:21:51 +0000

I think that the future of science is really in inter-disciplinary research that can be used to create collaborations between folks that produce data and those that analyzie it. For example, check out the latest in collaborative grants and RFAs posted by organizations here: http://www.iamscientist.com/rfas

By: Ben Shneiderman

Ben Shneiderman — Mon, 03 May 2010 12:18:36 +0000

Go danah! Bravo for this thoughtful post. Combining quantitative & qualitative methods to from credible case studies is the challenge as Science 2.0 researchers press forward to understand the science of the made world. Big data is a great opportunity, but ethnographic methods are needed to make the results meaningful.

By: eliot

eliot — Fri, 23 Apr 2010 05:41:10 +0000

Danah, thanks for this blog posting, it dovetails in an uncanny way with Tim O’Reilley’s recent keynote at the MySQL CE 2010 conference http://www.youtube.com/watch?v=WqLB99dA48k about The Cloud – and the corporate use of data in order to facilitate consumer convenience (while accumulating an unprecedented amount of data that can be analyzed and monetized). It seems that many of the problems you note by scholars without a social science background who make erroneous assumptions using Big Data boil down to nothing more that logical flaws (intentional fallacies, proof by assertion, etc.) Perhaps a couple semesters of formal logical training would assist, in addition to more social science background!

By: Cornelius

Cornelius — Wed, 21 Apr 2010 14:06:23 +0000

Thanks for a very insightful and timely post, danah. I think many of the issues you raise about the suggestiveness of data and it’s representation will stay with us for many years to come. I also think they will “spill out” beyond compsci/socsci and beyond scholarship in general. I’m referring to the applications of social/human data modeling in things like profiling, context-sensitive ads and semantic web technologies that “infer” without understanding context. People will mistrust these technologies even if they become very reliable over time, and there will be a significant need for critical debate and mediation between those who build them and the public at large, which is something social scientists and humanities scholars who are tech literate can hopefully help with.

One thought came up when reading your piece and thinking about numerous other pieces on Big Data I’ve read over time (e.g. Chris Anderson’s article on “The end of theory”: http://www.wired.com/science/discoveries/magazine/16-07/pb_theory). Big Data ultimately strikes me as an incredibly male idea. Quantitative data is incredibly suggestive, esp. when visualized. There is the idea that it “shows” or “proves” something that precludes other interpretations and it conveniently provides rhetorical ammunition to get almost anything across (when misused). Being data literate is essential.

By: Jed Hallam

Jed Hallam — Wed, 21 Apr 2010 10:08:19 +0000

Hey Danah,
This is a brilliant post – thank you! I started writing up a comment basically talking about how Big Data is effecting social media analysis and tried to put a social media spin on things and it ended up being horrifically long so I published it as a post instead (here -> http://rock-star-pr.com/big-data-and-social-media-analysis/).

Thanks again!

By: zephoria

zephoria — Wed, 21 Apr 2010 02:49:15 +0000

CS grad student: Learn social science methodologies from social scientists. Develop a taste for the different methodological approaches, what questions can be addressed through what means, etc. Find a social scientist advisor who can help acculturate you.

By: e. pyatt

e. pyatt — Tue, 20 Apr 2010 20:49:23 +0000

Another good article. I’ve seen similar problems in linguistic research by non-linguists (from biologists in particular) and…linguists trying to answer archaeological problems (usually badly), so I sympathize.

There may be promising approaches in in combining information from Big Data and ethnography, but the researcher has to understand the methodologies BOTH disciplines so as to not create a too simplistic model (e.g. more daily contact = closer relationtionship).

I agree with commenters who point out that you tend to get different types of information from different methodologies, and the connection points are not always obvious. I think Big Data could be worth pursuing IF everyone understands what they’re doing (not saying I do by the way). I agree that assuming this will give us quick and simple answers can be dangerous.

By: marcus

marcus — Tue, 20 Apr 2010 15:11:25 +0000

nice article. thanks. has be pondering if quantitative and qualitative will ever co-exist. sort of like debating the existence of god in society. everyone see’s the same things, yet each has a very different perspective of what they are seeing. hope you are well.

By: tricia wang

tricia wang — Tue, 20 Apr 2010 07:12:03 +0000

This is a really great article to compliment your post dana – looks at 3 different interdisciplinary areas of research and one of those areas is ethnography in the IT industry! http://bit.ly/d5vakd

Barry, A., Born, G., and Weszkalnys, G. Logics of interdisciplinarity. Economy and Society 37, 1 (2008), 20-49.
“This paper interrogates influential contemporary accounts of interdisciplinarity, in which it is portrayed as offering new ways of rendering science accountable to society and/or of forging closer relations between scientific research and innovation. The basis of the paper is an eighteen-month empirical study of three interdisciplinary fields that cross the boundaries between the natural sciences or engineering, on the one hand, and the social sciences or arts, on the other….”

By: jkd

jkd — Mon, 19 Apr 2010 17:42:23 +0000

Seconded.