December 15, 2011; Source: Miller-McCuneSharp polarization at all levels of politics might make us wonder, “Why is this happening, and why now?” School-based research conducted at Princeton University suggests that the problem may be too much information. This research posits that uninformed people play a critical role in political decisions because they temper extreme minority views. They tend to mirror majority views (for better and worse) rather than minority preferences: in other words, they democratize decision making. To the extent that the majority has wise views, the uninformed can have a positive social benefit.

To be clear, the school upon which the initial research is based is a school of fish. Specifically, the research led by Princeton ecologist Iain Couzin was designed to explore models of animal behavior and capture how animals make group decisions. In experiments with golden shiner fish, patterns in how individual animals “both influence and are influenced by those with whom they interact” were discovered.

Golden shiner fish are naturally attracted to the color yellow, so researchers used that preference as a proxy for “strong opinion.” After training one group of fish to swim toward the already preferred yellow colored target, and another group of fish to swim toward a blue target, they combined fish from both groups, with the yellow target fish in the minority. In over 80 percent of trials, the whole group of fish swam toward the yellow target. In other words, strong preference trumped the majority preference.

But when a sufficient number of untrained fish (a proxy for “uninformed”) were introduced into the mix, they changed the outcome of group decision making. The new fish tended to go with the flow of the majority and drive the group toward the blue target.

Is this fishy experiment really relevant to human behavior? The researchers have generated consistent results in simulations based on human behavioral studies from the social sciences. They’ve found that in very small or very large proportions, the uninformed don’t affect group decision dynamics. However, there’s a sweet spot in the middle where they do. A series of controlled human experiments is now planned to further explore what this means for human behavior.

These early results have intriguing implications for nonprofits’ drive to increase civic engagement in an era where we’re all swamped with information of varying reliability. If being informed (accurately or not) leads to more extremism and polarization among humans, what’s the remedy?—Kathi Jaworski