"They studied the messaging habits of 100,000 Facebook users by zip code"
This skews the study substantially. My facebook friends are my old classmates, colleagues, family and friends. Most of them are people that live close to where I do.
My e-mail habits are very different. Businesswise I probably correspond with more people abroad than in my own country.
The sample determines the output - and in this case the sample isn't representative of all online communications.
The sample is also much smaller than the article suggests. The paper itself says they found about 1800 'linked pairs' of users with a stated ZIP code in that sample of 100K. And in the graphs for the larger (4500) sample of email data, the fit is quite bad, with a pronounced 'hump' in the tail. There are just far too few sample points here to make these kinds of sweeping conclusions.
Alternative title: Study shows we all gravitate to each other
m1 m2
p(r) = G -----
r^2
where p(r) is a probability of a social link between two specific individuals with a distance r between them and m1, m2 their individual susceptibilities of being linked.
http://arxiv.org/ftp/arxiv/papers/0906/0906.3202.pdf
I think the conclusion fits neatly in to a different idea: more time is spent on relationships that are of greater value. (I don't equate value and returns - you could have a harmful relationship that you greatly value)
And with communication getting easier, maybe so does the realization that it is easier to find value in in-person relationships.
I wonder why they did this just on Facebook. Why not get access to real email logs, look in the Received: headers and do geo-location on the IP addresses. I'm sure if you looked in my email you'd see all sorts of messages coming from all over the world.
Eric Gilbert has analyzed facebook friending vs geography and found it also followed a power law distribution. Wouldn't the correct interpretation be then that we're indifferent to geography for anyone we're already friends with?
Comments
"They studied the messaging habits of 100,000 Facebook users by zip code"
This skews the study substantially. My facebook friends are my old classmates, colleagues, family and friends. Most of them are people that live close to where I do.
My e-mail habits are very different. Businesswise I probably correspond with more people abroad than in my own country.
The sample determines the output - and in this case the sample isn't representative of all online communications.
The sample is also much smaller than the article suggests. The paper itself says they found about 1800 'linked pairs' of users with a stated ZIP code in that sample of 100K. And in the graphs for the larger (4500) sample of email data, the fit is quite bad, with a pronounced 'hump' in the tail. There are just far too few sample points here to make these kinds of sweeping conclusions.
Alternative title: Study shows we all gravitate to each other
where p(r) is a probability of a social link between two specific individuals with a distance r between them and m1, m2 their individual susceptibilities of being linked. http://arxiv.org/ftp/arxiv/papers/0906/0906.3202.pdf"Gravity" laws are actually common in many real case scenarios (like transportation networks, etc).
I think the conclusion fits neatly in to a different idea: more time is spent on relationships that are of greater value. (I don't equate value and returns - you could have a harmful relationship that you greatly value)
And with communication getting easier, maybe so does the realization that it is easier to find value in in-person relationships.
I wonder why they did this just on Facebook. Why not get access to real email logs, look in the Received: headers and do geo-location on the IP addresses. I'm sure if you looked in my email you'd see all sorts of messages coming from all over the world.
Eric Gilbert has analyzed facebook friending vs geography and found it also followed a power law distribution. Wouldn't the correct interpretation be then that we're indifferent to geography for anyone we're already friends with?