There’s an interesting new study that was just published about how our brain mathematically handles online reviews that I wanted to talk about today.
But before I get to that, I wanted to talk about foraging a bit.
The story of how science discovered our foraging behaviors serves as a mini-lesson in how humans tick. The economists of the 1940s and '50s discovered the world of micro-economics, based on the foundation that humans were perfectly rational: we were homo economicus. When making personal economic choices in a world of limited resources, we maximized utility. The economists of the time assumed this was a uniquely human property, bequeathed on us by virtue of the reasoning power of our superior brains.
In the '60s, behavior ecologists knocked our egos down a peg or two. It wasn’t just humans that could do this. Pretty much any species had the same ability to seemingly make optimal choices when faced with scarcity. It was how animals kept from starving to death.
This was the birth of foraging theory. This wasn’t some homo-sapien-exclusive behavior that was directed from the heights of rationality downwards. It was an evolved behavior that was built from the ground up. It’s just that we humans had learned how to apply it to our abstract notion of economic utility.
Three decades later, two researchers at Xerox’s Palo Alto Research Center found another twist. Not only had our ability to forage been evolved all the way through our extensive family tree, but we seemed to borrow this strategy and apply it to entirely new situations.
Peter Pirolli and Stuart Card found that when humans navigate content in online environments, the exact same patterns could be found. We foraged for information. Those same calculations determined whether we would stay in an information “patch” or move on to more promising territory.
This seemed to indicate three surprising discoveries about our behavior:
-- Much of what we think is rational behavior is actually driven by instincts that have evolved over millions of years.
-- We borrow strategies from one context and apply them in another. We use the same basic instincts to find the FAQ section of a Web site that we use to find sustenance on the savannah.
-- Our brains seem to use Bayesian logic to continually calculate and update a model of the world. We rely on this model to survive in our environment, whatever and wherever that environment might be.
So that brings us to the study I mentioned at the beginning of this column. If we take the above into consideration, it should come as no surprise that our brain uses similar evolutionary strategies to process things like online reviews. But the way it does this is fascinating.
The amazing thing about the brain is how it seamlessly integrates and subconsciously synthesizes information and activity from different regions. For example, in foraging, the brain integrates information from the regions responsible for way-finding –- knowing our place in the world--– with signals from the dorsal anterior cingulate cortex, an area responsible for reward monitoring and executive control.
Essentially, the brain is constantly updating an algorithm about whether the effort required to travel to a new “patch” will be balanced by the reward we’ll find when we get there.
We don’t consciously marshal the cognitive resources required to do this. The brain does it automatically. What’s more, the brain uses many of the same resources and algorithm whether we’re considering going to McDonald’s for a large order of fries or deciding what online destination would be the best bet for researching our upcoming trip to Portugal.
In evaluating online reviews, we have a different challenge: How reliable are the reviews?
The context may be new – our ancestors didn’t have TripAdvisor or AirBNB ratings for choosing the right cave to sleep in tonight -- but the problem isn’t. What criteria should we use when we decide to integrate social information into our decision-making process?
If Thorlak the bear hunter tells me there’s a great cave a half-day’s march to the south, should I trust him? Experience has taught us a few handy rules of thumb when evaluating sources of social information: reliability of the source and the consensus of crowds.
Has Thorlak ever lied to us before? Do others in the tribe agree with him? These are hardwired social heuristics. We apply them instantly and instinctively to new sources of information that come from our social network.
We’ve been doing it for thousands of years. So it should come as no surprise that we borrow these strategies when dealing with online reviews.
In a neuro-scanning study from the University College of London, researchers found that reliability plays a significant role in how our brains treat social information. Once again, a well-evolved capability of the brain is recruited to help us in a new situation. The dorsomedial prefrontal cortex is the area of the brain that keeps track of our social connections. This “social monitoring” ability of the brain worked in concert with ventromedial prefrontal cortex, an area that processes value estimates.
The researchers found that this part of our brain works like a Bayesian computer when considering incoming information. First we establish a “prior” that represents a model of what we believe to be true. Then we subject this prior to possible statistical updating based on new information -- in this case, online reviews. If our confidence is high in this “prior” and the incoming information is weak, we tend to stick with our initial belief. But if our confidence is low and the incoming information is strong -- that is, a lot of positive reviews -- then the brain overrides the prior and establishes a new belief, based primarily on the new information.
While this seems like common sense, the mechanisms at play are interesting. The brain effortlessly pattern-matches new types of information and recruits the region most likely to have evolved to successfully interpret that information.
In this case, the brain had decided that online reviews are most like information that comes from social sources. It combines the interpretation of this data with an algorithmic function that assigns value to the new information and calculates a new model: a new understanding of what we believe to be true. And it does all this “under the hood," sitting just below the level of conscious thought.