4 thoughts on “View From Nowhere

  1. I don’t disagree with the premise, but a single concrete example of bias in either the collection or analysis would have been nice. Without examples it comes off a little like postmodern hand-waving.

    The analogy that kept running through my mind was how scientists analyze the data sets nature offers (e.g. weather). What makes a natural data set “just data” but a social data set somehow suspect in, and of, itself? Certainly the analyst in either case can apply bias, but science is, over time, self-correcting. Others will analyze that same data and, perhaps, find different results.

    An example is the data collected by BICEP2 and how their analysis found evidence of primordial gravity waves. As it turns out, their analysis had some problems, but the data set remains for better study.

    1. Thanks for the comment! I agree with your march of science point – yes, others will hopefully analyze the same data set. The problem I worry the most about is that if we assume bigger is better, we start to ignore the incredibly crucial role that good method plays in our ability to draw conclusions from analyses. It doesn’t matter how big a data set is – garbage in, garbage out. The allure of big data obfuscates that principle.

      1. Agreed. “Big Data” is a marketing ploy, no doubt to help sell a fairly expensive enterprise. Even “The Cloud” is being marketed as if it were something really special and not just online disk space.

        Also agree it’s a shame to obfuscate the potential real value large data sets have — any statistical analysis benefits from the largest possible ‘N’.

  2. I have heard this kind of arguments over and over again. But there is one flaw to them. They tend to blame the data instead of the people who use wrong methodology. One of the best scholarly critiques of this problem was published in 2006 by Kittel in the article called Crazy Methodology.

    Kittel himself is known for leaving the big data research and moving to smaller quantitative problems, which he believes is better in answering research questions. I have to agree with him. Big data research brings a lot of caveats, especially if one tends to measure its validity by the Null ritual.

    I believe big data are useful, especially if they are visualized rather than used in the inferential research. Flying drones or smart applications for our smartphones over our heads can provide us with more and more data to be able to assess problems of pollution everywhere. Once we have access to these data on a global level, we will learn a lot from them.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s