Crowdsourcing

Collecting Data

Crowdsourcing techniques do not always have to be digital. Before the Internet, it may have been more difficult to collect data from a large number of people, but it was not impossible.

In fact, Francis Galton used crowdsourcing in 1906 to accurately predict the weight of a bull. While advising a livestock fair in 1906, Galton observed a contest in which participants attempted to guess the weight of a particular ox that was on display. Out of the nearly 800 guesses made, nobody accurately estimated the exact weight of 1,198 pounds. Some guesses were too low, while others were too high. Galton surveyed the range of guesses and noted that out of the nearly 800 guesses, the mean (average) prediction was 1,197 pounds! While no single individual was able to make an accurate guess, the crowd, as a whole, was surprisingly accurate.

This “wisdom of the crowd” phenomenon is not restricted to the weight of bulls, however. It can be seen over and over again anytime a sufficiently large sample size of individuals is asked to estimate an unknown result and is the basis for the use of crowdsourcing.

Guess The Number of Jelly Beans

Let’s try this same experiment with virtual jelly beans.

We can crowdsource an estimated quantity using Google Forms and Spreadsheets to help gather a large number of responses. Make your best estimates and submit them to the collection. Then, examine the crowdsourced collection of estimates, and see how accurate the average estimate is to the actual amount.