Cynics sometimes describe the electorate as sheep, but a new study suggests their behavior could mirror that of fish.

The idea comes from an experiment into how an influential minority affects the behavior of a large group. It involved the golden shiner, a fish that is naturally predisposed to link yellow to food, but can be retrained (which is why it is often used for behavior studies).

The researchers took two groups of the fish: a small group, which they gave extra training to reinforce their preference for yellow, and a larger group that was “reprogrammed” to go after blue targets. They then put the two groups together and slowly added extra fish that had no special training.

The results emerged in three distinct stages:

• As a few fish were added, they followed those with the “yellow training”.
• As more fish were added, the untrained fish started following those with the “blue training.”
• Eventually adding more fish led to a point at which the untrained fish stopped behaving in a consistent manner.

The researchers at Princeton University said the experiment, which was accompanied by computer simulations producing similar results, could be explained by how well informed the fish were. In the first stage, the smaller group was more confident about its choice because the training backed its own experience and past learning, with this confidence overcoming the numerical advantage of the “blue training” dish.

However, the more “uninformed” fish were added, the less weight was placed on this level of confidence and the more simply started following the numbers. Eventually, though, the proportion of fish who’d been trained one way or the other was so small among the total group that this training and confidence stopped being influential.

Iain Couzin, who led the study, believes the experiment questions the idea that a vocal minority can always dominate a debate and change the minds of those who don’t start out with strong views. He said that having a large number of participants who start out uncommitted can help break deadlocks and lead to a majority preference — though whether this preference is “right” is open to question.

A professor who studies voting systems told the BBC the experiment could be compared with the US electoral system by which a vocal minority can influence the primary process, but is less successful at doing so in the election itself when more people (with less political interest) take part.