A simple puzzle to tell whether you know what people are thinking

http://www.washingtonpost.com/graphics/business/wonkblog/majority-illusion/

by bemmu • 11 years ago

131 35 11 years ago

dalke• 11 years ago

This is the Friendship paradox (https://en.wikipedia.org/wiki/Friendship_paradox ) applied to polling. If the people with the most acquaintances have similar viewpoints, which differ from those with fewer acquaintances, then it will skew responses to questions concerning the viewpoints of the respondent's acquaintances.

Excavator• 11 years ago

Spiral of Silence¹ is basically the same idea, as I read it.

1: https://en.wikipedia.org/wiki/Spiral_of_silence

resonanttoe• 11 years ago

That second graph they use as an example (Same sex marriage one) adds up to 101%...

DoctorZeus• 11 years ago

Presumably that's from rounding errors... for example, the distribution could be 4.7, 38.7, and 56.6 ... which adds up to 100, but when those numbers are rounded up they then add up to 101.

kristopolous• 11 years ago

So a devious entity can shift mass opinion through selective filtering and promotion of content on a social network.

I wonder what the nature of a "Pravda effect" would be where people suspect their information source has been corrupted and made unrepresentative and thus will distrust many traditional mechanisms of authority.

Is there an actual assuaging of opinion, a disengagement of the process, an increase in conspiratorial beliefs ... what happens?

Propaganda does this through smear campaigns - e.g., painting an academic as being egotistical and seeking money (thus making them corrupted by undisclosed motives). I see this with regard to climate change, evolution, and formerly tobacco smoke.

jokoon2• 11 years ago

I'm wondering how this apply to reality.

zdean• 11 years ago

The article makes a passing mention of media being a hub of these networks. One effect of this is that people will believe that there are epidemics of sorts (violence, economic, etc.) when in fact there aren't. This will often happen with pessimistic viewpoints.

An interesting example of this is explored in this article:

http://www.nytimes.com/2015/10/01/opinion/nicholas-kristof-t...

jokoon2• 11 years ago

I wonder about reddit. I guess this doesn't apply to reddit? Or less ?

JoshTriplett• 11 years ago

So, if you're trying to change public opinion, rather than changing the minds of the most people, you need to change the minds of the people who others respect and whose opinions influence those of others.

This also explains why some issues seem difficult to change public opinion on. For instance, if you want to get a majority of people to care about privacy, or more specific issues such as encryption backdoors and spying, you need those people's social circles to include many people who care. Since those people also tend to disproportionately avoid social networks...

munificent• 11 years ago

This was basically official process for Scientology. That's why they deliberately aimed for celebrities—get them and (presumably) the rest will follow.

nhaehnle• 11 years ago

Twitter is another example of consciously making use of this.

saurik2• 11 years ago

Wait, but did that work?

ChuckMcM• 11 years ago

Yup pretty much. People wonder how to change the opinion of any large group, and the answer is to get the "thought leaders" to agree and the rest will follow. So a good skill to acquire is watching to see gets things done and who doesn't.

newscracker• 11 years ago

It's not necessary to have "thought leaders" to agree. What matters is how influential the person is among certain groups. This fact has been long used in getting "celebrities" endorse stuff to gain more publicity and momentum in several areas - be they product marketing or social causes - while the facts may be that these celebrities have absolutely no qualification or expertise whatsoever about the product or cause, and thus could not really be called "thought leaders".

P.S.: Downvoted.

joe_the_user• 11 years ago

This is essentially one of the first modern theories of communication, articulated by Lazarsfeld and Katz in the 1940s.

https://en.wikipedia.org/wiki/Two-step_flow_of_communication

chippy4• 11 years ago

I think the article hints that the reason why people voted the way they do is because they didn't have connections to the people close to them.

We should encourage and increase the connections and links between people, rather than disrupt or change the links of the few with the most connections.

drdeca2• 11 years ago

If one, instead of considering the ratio among those one know, what if people consider the ratio of the ratios of the people that the people that you know, know, can the same type of thing happen?

chippy4• 11 years ago

Could you rephrase the question, please? I was a bit confused by the wording.

drdeca2• 11 years ago

Ah, yes, sorry:

I meant:

If instead of each person checking if an opinion is held by the majority of the people they know,

and seeing what the typical result for that would be,

what if the operation was doubled?

So, each person would check what the majority opinion is among those they know, and then store that value.

Then each person would check what the majority of /those/ values are among the people they know.

I was wondering if this might partially undo the effects from some people being known to more people.

so, if a person x 's opinion on a topic is f(x), and the friends of a person is x.friends

then instead of majority(map(f,x.friends)) , instead use majority(map(lambda z:majority(map(f,z.friends)), x.friends))

So instead of using "what opinions do I see", instead using "what opinions do I see people saying they see". Would that be a more or less accurate measure of "what opinions do people have"?

zeveb• 11 years ago

I wonder to what extent this sort of thing is a self-fulfilling prophecy. If most folks don't want to feel like they're out of the norm (which I suspect is true), then if they think that most folks disagree with them, they might be willing to change their opinions in order to fit in with a perceived majority.

As a couple of folks have noted, this means that all one has to do is suborn a relatively few influencers in order to effect mass change, despite the fact that the majority would have resisted that change if they'd known the truth.

As the Hitchhiker's Guide to the Galaxy might have said, some suspect that this has already happened…

ThrustVectoring2• 11 years ago

What's funny is that you can guess the answer to these kinds of questions - if it was what you'd naively think it would be, then there wouldn't be an article, so you can easily guess the "counterintuitive" answer.

chippy4• 11 years ago

Hah! I also initially thought this, then I assumed it was because the most connected people were the most connected, and had more voting power.

But I missed the fact about the others who were not able to connect with others around them. Increasing connections between those around you is the solution here, not disrupting the connections of the highly networked ones.

ZenoArrow2• 11 years ago

Yeah, you can look at it like that, but if you don't mind me saying I'd suggest that attitude is indicative of our approach to formal education, which rewards being 'right' regardless of how deeply something is understood, rather than rewarding getting something 'wrong' and learning why in the process of exploring further.

ThrustVectoring2• 11 years ago

I completely agree - I was definitely taught the skill of guessing the teacher's password.

cubano• 11 years ago

It seems to me that opinion shaping on Twitter has been using this effect for at least the past year.

hsshah• 11 years ago

This post reminded me of the ideas discussed by Malcolm Gladwell in "The Tipping Point". IIRC, talks about highly networked individuals starting new trends.

Freeboots• 11 years ago

"Assume each person based their vote solely on how the people they know would vote"

This is misphrased, they vote based on their friends opinions, not the way their friends would vote. Otherwise it becomes a chicken and egg situation.

A simple puzzle to tell whether you know what people are thinking

Related Stories