This month’s post in our series of guest academic librarian bloggers is by Jessica Hagman, Reference and Instruction Librarian at Ohio University. She blogs at Jess in Ohio.
Last fall, I taught a one-credit learning community seminar. During the week where we discussed research and library resources, I showed the class this video from Google, describing how the search engine works. I suspected that most students had no idea how links come to the top of a Google search results page and no basis on which to begin evaluating the results beyond page rank, a suspicion confirmed by research from the Web Use Project (previously discussed here on ACRLog).
Yet, when I asked whether the video surprised them or if the search engine process was different than they had previously thought, I heard the proverbial crickets. Finally, one student spoke up with a shrug, “I guess I’ve just never thought about it before.” While I probably shouldn’t have been surprised that few students spent time thinking about the mechanics of Google, it was startling to hear it stated so clearly.
I thought about this comment again a few weeks ago when I ran across a link to Eli Pariser’s TED Talk “Beware Online Filter Bubbles.” In the talk and his new book elaborating on the subject Pariser argues that companies like Facebook and Google use the data we share online to build a personalized bubble around each person in which they only encounter information, news and links that confirm their already established world view and assumptions. And while the bubble is pervasive, it is mostly invisible.
After watching the talk, my thoughts turned to the undergraduate researcher writing about a contentious social issue like gun control or abortion whose browser history limits the scope of the results they see on Google. I’ve discussed Google searching in many library instruction sessions, but it’s usually been to point out the poor quality of some of the search results and to encourage students to look beyond the first link. Starting in the fall, I will mention the personalization of search results as well, so that students are at least aware that their search results reflect more than just the keywords they searched.
The implications of the filter bubble may go beyond the research for a freshman composition paper, however. In the later chapters of his book, Pariser argues that the pervasiveness of filter bubbles may hinder learning, creativity, innovation, political dialogue, and even make us more susceptible to manipulative advertising. It’s difficult to discuss these consequences in a one-shot library instruction session, but to know that the bubble exists is a powerful first step to escaping it when necessary.
I will be teaching the learning community seminar again this fall, and this year I will show them Pariser’s talk. While I think it’s important that they be aware of personalized search and its potential implications, I’m also very curious to hear what students think about personalized search and a world of filtered information. While they may not have spent much time thinking about Google in the past, I hope that seeing the video will encourage them to think about how their own search history and browsing data affect what see – or do not see – online.
Pariser has some fascinating points, but I’d be careful about showing him “unfiltered.” There are questions about his research methods and whether he relies on anecdotes rather than data. Scroll down the interview portion with Jacob Weisberg in this On the Media transcript: http://www.onthemedia.org/2011/jun/17/echo-chamber-revisited/transcript/
I feel like fears of the personal bubble are very overstated. Google is not yet capable of determining, for instance, whether someone is pro-choice or pro-life and then presenting only results that agree with one paradigm. Either way, Wikipedia articles will probably top search results, and those are fairly dual-sided. And one can always do an explicit search for inverted opinions to find them, the fact that one does not do so is a human fault and not Google. Also – using public computers, or a browser that isn’t signed into a Google account, obfuscates the construction of a bubble. Substantive, quantitative analysis (such as the First Monday on personalized search, which found the effects to be fairly minor: http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/3344/2766) seems like a better approach rather than simply speculating that a bubble exists.
I found that reading about a possible filter bubble really resonated for me, even if the internet effects may not be as strong as Pariser warns (though I do still find it a bit creepy that results differ for the same search terms in the same search engine). One thing I’ve noticed is that the more information I want to keep tabs on, the harder it is for me to allot space and time to seeking out information that runs counter to ideas I’m already invested in. Part of this is the desire to read/watch/listen to stuff that I like, which I think is a pretty natural thing for most folks. But with more and more sources for info every day, it’s easy to stick to the ones that will validate my viewpoints and I find it challenging to fight this tendency in myself.