This post is about research – and how the best of us can inadvertently misuse tools.
An advisee came into my office to complain about a teacher. Listing to a kvetching student is not an unusual occurrence – gripes about tough tests, unfair grades, perceived grudges and the like are common in our world. 99% of the time I’m able to roll it back into a mentoring session – study harder, engage the instructor, devise new study plans, become the willow not the oak, etc.
This case played out differently – one of my colleagues had assigned your basic research query: “Google x, then hand-write the top three results for class discussion tomorrow.”
Simple assignment, yes? Some of you have already jumped ahead, but stay along for the ride. The problem with this assignment, and the reason my advisee was grumpy from his grade of zero? Per the instructor’s viewpoint, only one student out of ten completed the assignment correctly. The remainder of the class did not complete the assignment to the instructor’s satisfaction.
The 30-something humanities teacher fell trap to google nostalgia. This person was familiar with page-rank, could preach the praises of the algorithm, and frequently offered campfire horror stories of searching the wild, wild web pre-google (hotbot.com, anyone?) As most of us know, google reinvented search by determining a site’s value via associations and popularity, applying a peer review model to the chaos of the early web.
So… the teacher assigned a great basic lesson and intro assignment for research, right? Not so fast. The hiccup is that modern googling has gotten personal (we’ll leave the argument of “perhaps too personal” for another post) over the past few years. Yes, page rank still applies. But google tries to be smarter than us, customizing our search results to yield stuff that we are likely to be looking for. The students in my advisee’s class all correctly completed the assignment – they just all retrieved different results based on past searching and browsing behavior, yielding wrong answers from the teacher’s perspective.
On the one hand, this personalization tends to provide more focused, targeted results in answer to our googling. On the other hand, it can lead to some fairly whacky results. Duck Duck Go provides a light-hearted info-graphic-mercial, demonstrating the self-reinforcing search bubbles we can live in nowadays. I see this bubbling all the time running research exercises in class – the most visible trends of divergence (anecdotally) seem to break down by demographics, with stereotypical clumping of returned results based on gender, national origin, or the likelihood of a searcher to click through to fox vs. al jazeera.
Clive Thompson wrote a nice wake-up call on search in November of 2011, discussing the importance of information literacy – being able to analyze the multitude of results delivered by search tools -and placing the onus on us as educators to guide our charges in this new world. I’ll add that it is our duty to stress the bias inherent in the system – any system – of research. Most of us are aware that our personal biases cloud our judgement and impact our work.
We all need to be more aware that today’s digital tools will willingly help us subtly automate and refine our pet biases. I’m sure you teach bias by author, gender, source, time period, culture, and innumerable other modifiers … please add automated tools and inadvertent personal bias to your bag of tricks.