With only one student taking the survey, 70% of them stopped what they were doing and offered assistance. However, when there were two students taking the survey, this number dropped down dramatically. Most noticeably, when the group was two students—but one of the students was a stooge who was in on it and would always not respond, the response rate of the non-stooge participant was only 7%.
Somebody probably broke their leg next door behind just a curtain, and only 70% of the study subjects would go help? And only 7% would help if another person is in the room and the other person doesn’t go? Is anyone else very surprised by how low these numbers are? I would have expected something like 95% and 50%.
I am surprised, but unsurprised by my surprise. That is, I’ve come to expect that the results of these sorts of studies will consistently find that the “do the right thing” rates in the population are lower than I would naively expect them to be. That hasn’t really altered my intuitive sense that of course most people will “do the right thing,” although it has severely lowered my confidence in that intuition.
A psych teacher I had in college told the story of introducing a class decades earlier to the Milgram experiment, and (as he always did) asking for a show of hands of how many people in the room thought they would go all the way into the danger zone. Usually he got no hands; this year he got one hand. So after class he asked the guy (an older student) why he’d raised his hand, and the student explained that he was a war veteran, and knew perfectly well how easy it was to get him to do things that “nobody would ever do!”
The funny thing about Milgram is that the people who trusted the scientist and obeyed were factually right even if it didn’t feel that way to them at the time. No experimental subject was being hurt.
I’m just being overly literal because people believed that an experimental subject would be hurt and coincidentally it was the actual experimental subjects who were hurt. This would have been more apparent if Milgram had just hooked subjects up to real electrodes and told them he was testing reinforcement learning and then instructed them to actually shock themselves (which presumably many people would have done up to some pain threshold).
It would have been even more apparent if he could have lied about the voltage/current levels so that subjects received exactly as much physical pain (in terms of negative utility) to themselves from the actual shock as the emotional pain they would experience if they administered a shock of the imagined strength to another person.
Somebody probably broke their leg next door behind just a curtain, and only 70% of the study subjects would go help? And only 7% would help if another person is in the room and the other person doesn’t go? Is anyone else very surprised by how low these numbers are? I would have expected something like 95% and 50%.
I’m going to guess that it would vary not insignificantly depending on what the person pretending to break their leg looked like.
Before reading the paragraph with statistics I tried guessing what the results might have been, and I ended up with 50% for one person. It did not feel right at all, but just going off of similar studies quoted on this site I thought a healthy dose of pessimism was in order. The results just don’t seem surprising anymore, but its nice to be reminded that you have to try really hard to be cynical enough, and that if you want to be the kind of person who is 95% likely to help in a situation like this you need to do some major self modification or be self aware of your memetic and genetic programming often enough to veto it.
It wouldn’t hurt to get first-aid training. Knowing what to do is often more important (practically) than knowing that you should do something. Training for dealing with combative persons is probably useful too.
Random question (not expecting expert quality answers here, but curious)
If you have only minimal first aid training and no particular combat training, you are in a mostly deserted area, and you hear a gunshot somewhere nearby but out of sight, what do you think the rational altruistic response?
Edit: In this scenario, you either don’t have a phone, there are no police in the area, or otherwise “easy way out” are not available. You can choose to whether and how to help by yourself, and that’s it.
Is it really a gunshot?
Do you assume that the gunshot was at a person rather than a bear or mountain lion? (I don’t know where you are but “deserted” implies no people so an animal or an inanimate object is a more likely target.)
But let’s assume I am in an urban setting, the sound really is a gunshot, and your presumption that an innocent has been shot by an assailant with generally evil intent, e.g. the assailant is likely shoot me too if I appear, is correct. (OK, I am making a LOT of assumptions here.) In that case, the right answer is quite simple: get as far away as possible. Reasoning:
I have no way to combat the assailant successfully;
I have no way to assist the victim in any meaningful fashion;
there is a good chance I will also become a victim if I attempt to render aid.
In the end you will do no one any good, not even yourself. Also, you will not be able to do anyone any good in future.
Now, if you run away you have the opportunity to get First-Aid, EMT, and firearms training. You can get a concealed weapons permit and carry a gun with you at all times. (Well, in places that permit concealed carry.) Now if you happen to be in a similar situation again you will be in a position to both be an effective combatant AND be able to render aid to the victim. At that point in time you might decide to attempt to render aid. This becomes a much more difficult decision to make … but it wasn’t your original scenario.
Yeah. This is more or less what I’ve come to realize.
The story I was working on was originally prompted by an analogy about the weird psychological quirks that afflicted me when I was trying to decide where to donate money. I felt a compulsion to donate to a less efficient charity because it’s cause was “almost done,” and I wanted to finish it off and get a satisfying “Ding! Achievement” feeling.
This prompted someone to say “that’s like ignoring evidence that someone’s dying of a gunshot wound, so that you can finish throwing the last few starfish back into the ocean.” Which I then felt compelled to try and turn into a rationalist fan-parable.
A lot of pieces of it seem worthwhile, but the work is buckling and straining over the ridiculousness of a young girl and old man trying to do anything about a nearby gunshot. The gunshot itself isn’t all that important to the story, but I need something that:
a) is obviously more important than a few starfish b) time sensitive c) is plausibly addressable by a young girl and old man d) the evidence for said bad thing that needs addressing is uncertain.
Hi, I don’t know you, but I’d like to solve your puzzle.
A child opens a door to a house, and ten cats run out. The animal loving cat lady inside is disabled and so they have to get the cats for her. They’re rushing to get all ten cats back inside before any of them wander off too far and get lost. While they’re doing this, they see a child playing near an open manhole. She’s doing cartwheels and rolling on the ground. She’s too far to yell to. While they get the cats, they keep seeing her nearly missing the manhole. They keep thinking about going over there and putting the manhole cover back on. This would obviously take both the old man and the young girl—those are heavy… but they also want to get every cat before they’re gone...
I don’t think this is likely to come up as specified. In a rural area (of the United States, at least; other parts of the world might be different), isolated gunshots probably aren’t indicative of danger to a human and can be safely ignored unless you have good reason not to. In an urban area, you’re probably close to a payphone or another human with a phone, so even if you’re not carrying a phone yourself that should probably be your priority.
That being said, in the situation you’ve given any benefits to the victim would be overwhelmed by danger to you. You don’t know whether the assailant is still in the area, and you don’t know the disposition of anyone involved. It’d probably be slightly safer if you waited a few minutes before investigating (a single shot means a brief incident, and a sane attacker probably wouldn’t stick around long), but your scope for rendering aid is so limited that I still think I’d favor leaving the area and getting help if possible.
Additionally: Prioritize minimizing risk to yourself (and anyone else once you learn of their presence, if any) while grabbing any minimal-risk opportunity to obtain more information on the situation (e.g. seeing the face or hearing the voice of the person who shot and the person who was shot, if any).
Sounds about right. In the process of writing what’s sort of supposed to be a rationalist parable, but the takeaway lesson is sort of vague and I’m not sure what to encourage.
What are you going to do against a person with a gun even if someone’s been shot?
There’s a very good chance the person or animal targeted is already dead and it’s too late.
If you move toward a gunshot sound, and visibility isn’t perfect (say you’re going through the bushes) you may be shot because you’re mistaken for a wild animal or enemy.
The last thing you’d want to do is rush over and see what happened or get involved in a conflict when you have no idea how it started or who is telling the truth.
In this scenario I just don’t think “rational” and “altruistic” are compatible (not even using an approximation to an acausal decision theory and assuming that so do other people, which makes altruism rational in the prisoners’ dilemma). I’d just run the fuck away.
Anyway, this isn’t analogous at all to the scenario in the study. There’s no way you’ll harm anyone (including yourself) by asking the researcher if she’s OK and (say) calling an ambulance or something if she isn’t.
I think there is a very good case that a hypothetical rational agent motivated entirely by helping others would run away—that doesn’t mean it isn’t altruistic, just that there is realistically nothing to do to help others except leave to survive and find other others to help.
I’m surprised too, and I’d like to think I’d check on her in that situation; I’ve checked up on people with significantly less provocation.
OTOH, how good was the setup? Could the subjects have had a clear view of the actor the whole time, and actually been thinking, “Why is she pretending to be injured? … weirdo”
Peter Singer’s account of the experiment made it clear that all of the action occurred behind a curtain which made it so the subject could not be viewed. Unfortunately since the source is cited to be a book, I can’t follow up on this easily.
I suspect the figures would vary a lot depending on what country/culture the students are from.
(Also, maybe some of the students had guessed that they were being tested about that, and wanted to screw up the results—though 30% sounds like a lot for this hypothesis. What were they students of?)
I am curious if they collected data for how often they said something on the assumption the other researcher should go, especially if they saw the back as private.
Somebody probably broke their leg next door behind just a curtain, and only 70% of the study subjects would go help? And only 7% would help if another person is in the room and the other person doesn’t go? Is anyone else very surprised by how low these numbers are? I would have expected something like 95% and 50%.
I am surprised, but unsurprised by my surprise.
That is, I’ve come to expect that the results of these sorts of studies will consistently find that the “do the right thing” rates in the population are lower than I would naively expect them to be. That hasn’t really altered my intuitive sense that of course most people will “do the right thing,” although it has severely lowered my confidence in that intuition.
A psych teacher I had in college told the story of introducing a class decades earlier to the Milgram experiment, and (as he always did) asking for a show of hands of how many people in the room thought they would go all the way into the danger zone. Usually he got no hands; this year he got one hand. So after class he asked the guy (an older student) why he’d raised his hand, and the student explained that he was a war veteran, and knew perfectly well how easy it was to get him to do things that “nobody would ever do!”
The funny thing about Milgram is that the people who trusted the scientist and obeyed were factually right even if it didn’t feel that way to them at the time. No experimental subject was being hurt.
Of course, by the same token they were factually wrong: no experimental subject was learning to memorize a list of words.
I thought the experimental subjects reported feeling extreme duress about causing apparent harm to another human being.
You’re reffering to the experiment itself; they’re talking about the experiment within the experiment.
I was compelled to post that clarification after being primed for “helping behavior.”
I’m just being overly literal because people believed that an experimental subject would be hurt and coincidentally it was the actual experimental subjects who were hurt. This would have been more apparent if Milgram had just hooked subjects up to real electrodes and told them he was testing reinforcement learning and then instructed them to actually shock themselves (which presumably many people would have done up to some pain threshold).
It would have been even more apparent if he could have lied about the voltage/current levels so that subjects received exactly as much physical pain (in terms of negative utility) to themselves from the actual shock as the emotional pain they would experience if they administered a shock of the imagined strength to another person.
I’m going to guess that it would vary not insignificantly depending on what the person pretending to break their leg looked like.
Before reading the paragraph with statistics I tried guessing what the results might have been, and I ended up with 50% for one person. It did not feel right at all, but just going off of similar studies quoted on this site I thought a healthy dose of pessimism was in order. The results just don’t seem surprising anymore, but its nice to be reminded that you have to try really hard to be cynical enough, and that if you want to be the kind of person who is 95% likely to help in a situation like this you need to do some major self modification or be self aware of your memetic and genetic programming often enough to veto it.
It wouldn’t hurt to get first-aid training. Knowing what to do is often more important (practically) than knowing that you should do something. Training for dealing with combative persons is probably useful too.
Random question (not expecting expert quality answers here, but curious)
If you have only minimal first aid training and no particular combat training, you are in a mostly deserted area, and you hear a gunshot somewhere nearby but out of sight, what do you think the rational altruistic response?
Edit: In this scenario, you either don’t have a phone, there are no police in the area, or otherwise “easy way out” are not available. You can choose to whether and how to help by yourself, and that’s it.
Is it really a gunshot? Do you assume that the gunshot was at a person rather than a bear or mountain lion? (I don’t know where you are but “deserted” implies no people so an animal or an inanimate object is a more likely target.)
But let’s assume I am in an urban setting, the sound really is a gunshot, and your presumption that an innocent has been shot by an assailant with generally evil intent, e.g. the assailant is likely shoot me too if I appear, is correct. (OK, I am making a LOT of assumptions here.) In that case, the right answer is quite simple: get as far away as possible. Reasoning:
I have no way to combat the assailant successfully;
I have no way to assist the victim in any meaningful fashion;
there is a good chance I will also become a victim if I attempt to render aid.
In the end you will do no one any good, not even yourself. Also, you will not be able to do anyone any good in future.
Now, if you run away you have the opportunity to get First-Aid, EMT, and firearms training. You can get a concealed weapons permit and carry a gun with you at all times. (Well, in places that permit concealed carry.) Now if you happen to be in a similar situation again you will be in a position to both be an effective combatant AND be able to render aid to the victim. At that point in time you might decide to attempt to render aid. This becomes a much more difficult decision to make … but it wasn’t your original scenario.
Yeah. This is more or less what I’ve come to realize.
The story I was working on was originally prompted by an analogy about the weird psychological quirks that afflicted me when I was trying to decide where to donate money. I felt a compulsion to donate to a less efficient charity because it’s cause was “almost done,” and I wanted to finish it off and get a satisfying “Ding! Achievement” feeling.
This prompted someone to say “that’s like ignoring evidence that someone’s dying of a gunshot wound, so that you can finish throwing the last few starfish back into the ocean.” Which I then felt compelled to try and turn into a rationalist fan-parable.
A lot of pieces of it seem worthwhile, but the work is buckling and straining over the ridiculousness of a young girl and old man trying to do anything about a nearby gunshot. The gunshot itself isn’t all that important to the story, but I need something that:
a) is obviously more important than a few starfish
b) time sensitive
c) is plausibly addressable by a young girl and old man
d) the evidence for said bad thing that needs addressing is uncertain.
Why can’t I resist a puzzle?
Hi, I don’t know you, but I’d like to solve your puzzle.
A child opens a door to a house, and ten cats run out. The animal loving cat lady inside is disabled and so they have to get the cats for her. They’re rushing to get all ten cats back inside before any of them wander off too far and get lost. While they’re doing this, they see a child playing near an open manhole. She’s doing cartwheels and rolling on the ground. She’s too far to yell to. While they get the cats, they keep seeing her nearly missing the manhole. They keep thinking about going over there and putting the manhole cover back on. This would obviously take both the old man and the young girl—those are heavy… but they also want to get every cat before they’re gone...
A scream? Perhaps “HELP!”
A car crash?
I was thinking a scream earlier today, but the car crash actually works better for poetic reasons.
A child crying somewhere nearby?
I don’t think this is likely to come up as specified. In a rural area (of the United States, at least; other parts of the world might be different), isolated gunshots probably aren’t indicative of danger to a human and can be safely ignored unless you have good reason not to. In an urban area, you’re probably close to a payphone or another human with a phone, so even if you’re not carrying a phone yourself that should probably be your priority.
That being said, in the situation you’ve given any benefits to the victim would be overwhelmed by danger to you. You don’t know whether the assailant is still in the area, and you don’t know the disposition of anyone involved. It’d probably be slightly safer if you waited a few minutes before investigating (a single shot means a brief incident, and a sane attacker probably wouldn’t stick around long), but your scope for rendering aid is so limited that I still think I’d favor leaving the area and getting help if possible.
Call 911 (or local equivalent) and ask what to do?
Additionally: Prioritize minimizing risk to yourself (and anyone else once you learn of their presence, if any) while grabbing any minimal-risk opportunity to obtain more information on the situation (e.g. seeing the face or hearing the voice of the person who shot and the person who was shot, if any).
Sounds about right. In the process of writing what’s sort of supposed to be a rationalist parable, but the takeaway lesson is sort of vague and I’m not sure what to encourage.
Meant to specify that you didn’t have a phone, or that you are in a place without easy access to cops.
Leave and get help.
What are you going to do against a person with a gun even if someone’s been shot?
There’s a very good chance the person or animal targeted is already dead and it’s too late.
If you move toward a gunshot sound, and visibility isn’t perfect (say you’re going through the bushes) you may be shot because you’re mistaken for a wild animal or enemy.
The last thing you’d want to do is rush over and see what happened or get involved in a conflict when you have no idea how it started or who is telling the truth.
In this scenario I just don’t think “rational” and “altruistic” are compatible (not even using an approximation to an acausal decision theory and assuming that so do other people, which makes altruism rational in the prisoners’ dilemma). I’d just run the fuck away.
Anyway, this isn’t analogous at all to the scenario in the study. There’s no way you’ll harm anyone (including yourself) by asking the researcher if she’s OK and (say) calling an ambulance or something if she isn’t.
I think there is a very good case that a hypothetical rational agent motivated entirely by helping others would run away—that doesn’t mean it isn’t altruistic, just that there is realistically nothing to do to help others except leave to survive and find other others to help.
Call the police and run?
I’m surprised too, and I’d like to think I’d check on her in that situation; I’ve checked up on people with significantly less provocation.
OTOH, how good was the setup? Could the subjects have had a clear view of the actor the whole time, and actually been thinking, “Why is she pretending to be injured? … weirdo”
Peter Singer’s account of the experiment made it clear that all of the action occurred behind a curtain which made it so the subject could not be viewed. Unfortunately since the source is cited to be a book, I can’t follow up on this easily.
I suspect the figures would vary a lot depending on what country/culture the students are from.
(Also, maybe some of the students had guessed that they were being tested about that, and wanted to screw up the results—though 30% sounds like a lot for this hypothesis. What were they students of?)
I am curious if they collected data for how often they said something on the assumption the other researcher should go, especially if they saw the back as private.
The subjects were (falsely) told that the other researcher was another student taking the same survey.
Interesting. This raises my level of surprise a bit.
Serously—they audibly complained about not feeling or moving their foot, and no one did anything? This sounds fishy.