The essay should be organized as a mini textbook chapter, including a statement of position, a logical organization of the concepts into a coherent story, and a conclusion demonstrating your own insights into the topic (see example essay on the home page of the course).
The essay should contain a definition box of terms that are technical and/or particularly important to understand the topic of the essay.
In-Class Notes on topic for reference:
Crowds -
More classes will cause crowds to get closer to 100%
This is the wisdom of crowds
If we put eveyrones ipinons together, we get a better and better image of the “truth”
Sir frances galton 1822 - 1911
Collected people’s guesses of a cow’s weight at the fair
Vox populi - “the voice of the people”
Real weight 1198 lbs
Average of guesses was 1207 lbs
All people tend to underestimate much more than be accurate or overestimate
Confidence isnt a factor in accuracy unless there is training in a related field
Central limit theorem (CLT)
Theres some randomness in the system (which produces unreliable estimates)
Many wrongs principle
Many wrongs makes a right
Everyone in the population makes incorrect estimates, but when you average them, they’ll get closer and closer to the truth
This is true as long as the population’s estimates are unbiased
Ex. in the who wants to be a millionaire show, between 85 and 95% of the time, the audience average is correct
Marquis de couderat
Nobleman in france
Born 1743 died 1794
French revolution was in 1789 due to many people being displeased with disproportionate division of wealth
Marquis supported this cause, was a humanist and was interested in creating a fair justice system where they jury is fair and more likely to make the right decision; but was ultimately executed because he was not “radical” enough
Couderat jury theorum
Wisdom of crowds
Often supported by the Many wrongs principle
You need many unbiased estimators to get the right answer
Supported by the central limit theorem
2 conditions to check? Independence and lack of bias.
Couderat’s jury theorem
If you know probability of correctness of each person’s choice being correct, you can calculate the probability of majority being correct/making the right decision
Can be applied to any kind of binary decisions
If people can be correct at least 50% of the time, then the larger the group the closer to 100% accuracy the group will be
Independence is required between individuals in order for Jury’s and Many wrongs to be applicable - without it there is biased
Amazon example
Bad reviews/good reviews can highly influence prob of buying
Down treated comment section resulted in an overshoot - where ppl came back to upvote and say “this product was much better than expected”
When looking at social information and transmission you must be critical in analyzing where the information came from because you must determine if it’s true, useful, is it coming from a trustworthy source?
Groupthink - bc people around you are telling you things, your perspective shrinks and the whole group ends up having a very narrow view
The reinforcement of one another’s opinions pushes them further and further from the “truth”
Bias -
our brains are biased - a cognitive bias (ex. Estimating quantities very quickly results in people underestimating the value)