Many articles in academia about serials and digital repositories mention measuring the ‘impact’ of a journal through various rubrics. Now that Google results are ‘personalized’ using data from prior searches, I wonder how that will skew the impact of articles. (Of course, Google is not alone in using cookies, search history, IP address, etc., to ‘personalize’ search results. Facebook is another major service that is known for this, too.)
I just finished reading Eli Pariser’s The filter bubble: what the internet is hiding from you, and he discusses at length the ways that this personalization is going to impact our knowledge. For instance, if you are an economist who mostly searches for, and clicks through to view, articles on a particular kind of economic theory, then you will see more results in that vein, and fewer results of other economic theories that do not fit into your search history. While this can save time when doing research, the problem is that we are going to be exposed less frequently to worldviews that do not coincide with our own.
Where I see this being relevant to serials and digital repository management is that articles are often used to measure the ‘impact’ of faculty writings, or of an institutional repository in general. Well, if many economists skew to one worldview, than more of those articles will be viewed – while equally good, but different economics articles will be viewed less, and therefore will be cited less frequently in papers, thereby lessening the impact of what could be a ground-breaking article.
It seems like we are moving from judging arguments on their merits to simply thinking along with the herd. Of course, this has been a problem for a while, since no one person or set of people could ever really sift through all of the literature in a particular subject area. But the problem is clearly going to be exacerbated by the ongoing personalization of search results. (And I fear it will not be long before such data is used by academic databases, such as EBSCO, to filter search results ‘by relevancy’.)
If you are interested in reading more about this, you can visit Pariser’s website: http://www.thefilterbubble.com/. He also has a page on “Ten things you can do to pop your filter bubble”: http://www.thefilterbubble.com/10-things-you-can-do.