In a nutshell, crowdsourcing is when you ask a large group of people (the crowd) for their answers to a question. Polling and surveys are forms of crowdsourcing. Social-media posts that ask for your opinion are another. And now companies are using this technique more and more for their innovation efforts.
The downside of many crowdsourcing initiatives is something called "mobsourcing." This is where the vocal few crowd out the rest of the responses. We saw this when California asked its citizens about how to solve its economic woes, and the majority of respondents said "Legalize and tax marijuana." The famous Boaty McBoatface story from the U.K. is another.
I've been talking about mobsourcing for ages, and now there is scientific research that backs me up, with some techniques to deal with it.
A recent research study on crowdsourcing said:
"Democratic methods... tend to favor the most popular information, not necessarily the most correct. The ignorance of the masses can cancel out a knowledgeable minority with specialized information of a topic, resulting in the wrong answer becoming the most accepted."
This is mobsourcing.
Researchers from Princeton and MIT wanted to see if there was a way to find the gold hidden among the duds. They started with a hypothesis and constructed an experiment, asking a group of people, "Is Philadelphia the capital of Pennsylvania?"
Not surprisingly (but maybe shockingly), a majority of people said "yes." Only a minority of the people knew that Harrisburg was the correct answer.
Here's where it gets fascinating.
In addition, people were asked to predict how popular the "yes" response would be. The people who said "yes, Philadelphia is the capital" predicted that most people would also say "yes." However, the people who knew the correct answer was "no" believed that most people would get it wrong and say "yes." Therefore, in total, most people "predicted" Philadelphia to be the popular answer.
Philadelphia was the most popular answer.
But the actual percentage of people who said "yes, Philadelphia is the capital" was quite a bit lower than the number of people predicted to say "yes."
Meanwhile, "no, Philadelphia is not the capital" exceeded predictions because very few people thought that others would give this response.
What they discovered is that the correct answer is not necessarily the most "popular" one. Rather, it is the one that is more popular than people "predict."
This knowledge could inform how we can conduct better experiments in the future. It could help companies innovate more efficiently. It might help political polling become more accurate. And it might even be useful for companies when they do customer surveys.
So the next time you ask your Facebook friends, "Do I look good in this outfit?" You might also want to ask them, "Do you think most people will say I look good?"