Kids these days

[gfish]

1) There is a natural cognitive bias, the older one gets, to see society as going downhill.

2) I have been increasingly worried about social trends for several years now.

It's getting very hard to write off my worries as just being a symptom of #1. All the antisocial behavior we kept excusing as something "just on the internet" has been leaking more and more into the physical world. We all spend all our time in an environment where the only response to the most hideous of attacks is just "oh, ignore it, it's not serious". Of course empathy is becoming increasingly unfashionable! Even traffic is getting more aggressive, with people breaking the speed limit much more consistently and to greater average degrees it seems. But obviously it's hard to trust those observations.

Is there an intellectually rigorous method for resolving this dilemma?

Comments

[eub.livejournal.com]

You know what somebody should do? Re-run the Milgram chain letter experiment again, and over time. This is the other one, the experiment where he made an estimate of social distance between two people: asked Alice to get a chain letter to Bob by sending it to one of Alice's acquaintances "on a first-name basis", asking that person to do the same.

Has the mean number of hops changed over time? Has the distribution broadened? Run it with two people chosen uniformly from the whole population[1], versus pairs chosen to be same or different in, say, urban/rural location, race, class. Which social divide is the biggest one, and how have their sizes changed over time?

(I came for the historical data, but I'd stay for data today about distribution and size of social divides. How hard would this really be to run?)

[eub.livejournal.com]

oops, Procrustesed the footnote:
[1] which Milgram did not. He chose people in cities: one in Boston, one in either Omaha or Wichita. And oh good grief man, they were respondents to advertisements for well-connected people.

(I see there's actually been quite a lot of related work. Haven't found ones in these particular directions though.)

[eub.livejournal.com]

Chain-lettering to myself here, Milgram had another experiment that kinda measures empathy toward a given group (or helpfulness, or sense of duty), the "lost letter" setup (PDF (http://www.communicationcache.com/uploads/1/0/8/8/10887248/the_lost-letter_technique-_a_tool_of_social_research.pdf)).

You could (IRB aside -- could you run this experiment if you can't debrief the subject?) do this where you drop the letters in different locations, and address them to different locations. You could racially code the recipient names like with experiments on resumes; the finder's race is harder to control outside of location. Income / class is hard.

(But are you measuring people's attitudes towards people, or people's habits towards paper mail?)

Here's a paper (https://www.jstor.org/stable/3006055?seq=1#page_scan_tab_contents) that looked at whether small-town folks are more helpful than city dwellers: nah, says the abstract. They varied the letter drop location, and held the destination address constant.