Debunking the Bystander Effect
July 16, 2019 11:03 PM   Subscribe

New research analyzing camera footage shows the bystander effect may be largely a myth.
posted by blue shadows (22 comments total) 17 users marked this as a favorite
 
This is almost a perfect storm of pop-science, because I think these outcomes have actually been changed by the act of observing them*.

What I mean is, the bystander effect comes from an era of increasing cultural focus on urban life and its dangers. Pop-culture was obsessed with the city as a place of opportunity and anonymity, both considered against small towns where everyone knows each other's business and reinvention is impossible. It comes from the 80s and 90s. The stabbing incident most associated with the idea comes from 1996, but it's a cherry picked occurrence that gave shape and weight to anxieties that had brewed for decades, like any good scary story.

It was also a story told JUST before the advent of the internet and realization of full urban surveillance, Also at the cusp of the new 24 hour news cycle. These trends would lead to the age of camera phones and instant notoriety.

It's very hard to say how often people ignored visible city crime in 90s, or whether a dip up or down would have strayed far outside the norm. We were coming out of a decade where cities were cast as lawless, graffiti-strewn, hellscapes in the popular imagination and moving into the decade in which every sitcom took place in NYC or some other major city and portrayed as both aspiration and refreshingly amoral.

But certainly in an age of cellphones, police vest cameras, and the funk of social media exposure always following us, it makes sense that people are publicly responsive to obvious injustice and crime. This might be a good thing? Social awareness by way of Big Brother?

*my mom thinks I'm clever
posted by es_de_bah at 11:24 PM on July 16, 2019 [10 favorites]


Everyone I know who studies the bystander effect is super annoyed right now at the media coverage of this study. I know some folks are working on a response right now. I'll post it asap.
posted by k8t at 11:29 PM on July 16, 2019 [16 favorites]


In the US at least, awareness of the bystander effect as a phenomenon might also push people to intervene because they think that nobody else will or because they don't want to be a bystander.
posted by storytam at 11:32 PM on July 16, 2019 [3 favorites]


I couldn't get into the full article by Philpot, et al., but a couple of things occur to me just based on the abstract and the article you linked to.

1. The new study looks at behavior in the UK, Netherlands, and South Africa. I'm almost certain Darley and Latane were looking at Americans. This makes me wonder about the extent to which diffusion of responsibility varies with cultural context.

2. The newer research seems to define "intervention" quite broadly.

3. I agree with es_de_bah that expectations about surveillance are changing, and that may have an impact on how people behave in public. It still surprises me to read about how prevalent public surveillance is in the UK now.
posted by Nerdy Spice at 11:33 PM on July 16, 2019


Advice to anyone writing about science, including in metafilter posts:
A scientific experiment never 'shows' anything. It indicates, or gives evidence in favour of a claim- and even that usually requires definite interpretation of results.
posted by leibniz at 11:34 PM on July 16, 2019 [25 favorites]


The stabbing incident most associated with the idea comes from 1996,

The Kitty Genovese incident occurred in 1964.
posted by Mitheral at 11:54 PM on July 16, 2019 [27 favorites]


Point taken, indicates would definitely be a better word.
posted by blue shadows at 12:10 AM on July 17, 2019 [2 favorites]


And, as the third link points out:
Initially reported in a 1964 New York Times article, the story sensationalized the case and reported a number of factual inaccuracies. An article in the September 2007 issue of American Psychologist concluded that the story is largely misrepresented mostly due to the inaccuracies repeatedly published in newspaper articles and psychology textbooks.
With the real truth being that the attack did not happen as popularly portrayed and several people did intervene either through shouting or calling the police. The major factors for the attack not being taken as seriously as it was resulted from people in the neighborhood thinking it was either a domestic assault or just being afraid to call the police in general because they were homosexuals in an era when simply existing could, and did, result in an arrest.

So if the Kitty Genovese story tells us anything its that a major way to combat any sort of bystander effect is to eradicate misogyny and homophobia. People are more likely to intervene when they don't take interpersonal violence for granted or fear state violence against them.
posted by Panjandrum at 1:28 AM on July 17, 2019 [47 favorites]


On a related issue: In a disaster, mass panic is rare.
posted by Mr.Know-it-some at 5:48 AM on July 17, 2019 [4 favorites]


I don't think anyone who has been the target of racist aggression or microaggression would say the bystander effect is dead. It's been around since long before Kitty Genovese -- James Garfield has a scathing speech in 1947's Gentleman's Agreement about how evil happens when people stay silent. And if we're talking about just physical violence and not verbal, a hell of a lot of child abuse happens in plain sight because of cultural norms about parental authority.

On the flip side, slogans like "See something, say something," are supposed to be a sort of anti-bystander effect thing, but have ended up turning into tools of racist abuse anyway. Humans are really awful sometimes.
posted by basalganglia at 6:06 AM on July 17, 2019


This research doesn't seem to indicate the bystander effect is a myth at all. The majority of the bystanders didn't intervene. What they found was that there was a good chance someone would intervene and the more people were there, the higher the chance was. The study actually seems to back up the idea that most bystanders won't act, but contributes the reassuring observation that the fraction who will act is usually high enough to result in some intervention taking place.
posted by Redstart at 6:17 AM on July 17, 2019 [1 favorite]


What they found was that there was a good chance someone would intervene and the more people were there, the higher the chance was.

That's the opposite of what the bystander effect predicts, so I'm not sure how it isn't indicative of the bystander effect being a myth.
posted by Bugbread at 6:46 AM on July 17, 2019 [4 favorites]


Didn't look at the whole study, so correct me if I'm wrong, but isn't it also possible that the fact that the idea of the Bystander Effect (and the widespread story of Kitty Genovese) is now so well-known, that people are now more likely to intervene because they don't want there to be another horrific preventable occurrence on their watch? I was born years after her murder and her story was always told as a morality tale about how ordinary people are more terrible than we think we are -- and it certainly stuck with me and made me always try to step up in a crowd if it seemed needed - I can't believe i'm alone in this.
posted by Mchelly at 7:13 AM on July 17, 2019 [1 favorite]


Bugbread, I don't think that's quite right. As I understand it, the bystander effect predicts that individual people are inhibited from helping as a function of increasing group size. So long as inhibition increases at a lower rate than group size the chances are that at least one person in the group will help. Some statistics would come into play, because not everyone in the group will be inhibited to a similar extent. Some people will be very inhibited, some people will not be inhibited at all, and most people will be inhibited somewhat. Thus, as group size increases and the effect of inhibition increases, more and more people will be beyond the threshold of inhibition and will not help. However, if there is a normal distribution of inhibition we can predict that a larger group is more likely to have individuals present who are a sufficient number of standard deviations from the mean that they fall below the threshold of inhibition and step forward to help. Thus, even though more people in a large group are inhibited from helping, a large group still has a better chance of producing a helper.
posted by slkinsey at 7:41 AM on July 17, 2019 [4 favorites]


That's the opposite of what the bystander effect predicts, so I'm not sure how it isn't indicative of the bystander effect being a myth.

The opposite of the bystander effect would be for the increasing number of bystanders to make individuals more likely to take action. I don't think this study showed that that happened. As slkinsey explains above, even if the increasing size of the group makes each individual less likely to help, the probability that someone will help can still go up. Let's say the probability that any one person will help is .5 in a group size of 4 and only half that (.25) in a group size of 20. That still means you'll have 2 people on average intervening out of a group of 4 and 5 intervening out of a group of 20, so you're more likely to get help from the larger group, even though each individual is less likely to help.
posted by Redstart at 8:17 AM on July 17, 2019


Oh, this means I get to plug one of my favorite podcasts, You're Wrong About. Here is their episode about Kitty Genovese. Some key things that I learned from that episode:
- The murder happened in the middle of the night, and there's no reason to think that lots of people saw it happen.
- One of the people who delayed calling the police was gay, and this was the 60s, so there was plenty of reason for him to think twice about having the cops there. This was basically borne out when the police focused most of their investigation on Genevese's girlfriend (she was a lesbian).
- Police at this time routinely did not come when called.
- In fact, someone did call the police to report the attack, and they didn't come.
- A fair amount of people who heard the struggle thought it was domestic violence, which was legal at the time.

But the most important thing I learned is that Kitty Genovese was awesome. Well worth a listen.
posted by Ragged Richard at 8:25 AM on July 17, 2019 [15 favorites]


Or you could have something like the incident Monday on Highway 101 in Paso Robles, California, where two Good Samaritans got out of their cars to help the victim of a crash with a wrong-way hit-and-run driver and got run over by a van that lost control after one of his tires burst from the first crash's debris. Now that was a Good Samaritan worst-case scenario.
posted by oneswellfoop at 8:36 AM on July 17, 2019


A classmate of mine was killed in a similar way. He got out to help a woman whose car was stranded in the snow, and another driver lost control and hit him. Helping someone can be legitimately dangerous.

Even when not, though - I've been in situations where I'm not sure intervening would make things worse or not. I'm still not sure what the "right" call is in some of those situations, because I'm pretty sure the "right" call is down to a lot of factors that are hard to predict.
posted by Kutsuwamushi at 9:37 AM on July 17, 2019 [1 favorite]


Advice to anyone writing about science, including in metafilter posts:
A scientific experiment never 'shows' anything. It indicates, or gives evidence in favour of a claim- and even that usually requires definite interpretation of results.
The difference between "showing" something and "providing very strong evidence" for something is pretty subtle. The former is a very common phrase in lots of professional, scientific contexts and publications. Whether it implies too much confidence in this case isn't something I know enough to comment on. But, it's not an unreasonable word to choose.

Continuing to respect experimental psychology as a legitimate science grows harder with every paper I read.
posted by eotvos at 12:15 PM on July 17, 2019 [1 favorite]


slkinsey: " As I understand it, the bystander effect predicts that individual people are inhibited from helping as a function of increasing group size. So long as inhibition increases at a lower rate than group size the chances are that at least one person in the group will help."

That makes sense, but it doesn't map to the conventional understanding of the Kitty Genovese case, which is what got the whole bystander effect ball rolling (as has been pointed out, the reality of the case is different from the conventional understanding).

If that interpretation were true, you would have something more along these lines:
  • Normally, if 4 people see a crime, one calls the police, so the call likelihood that any individual will call is 25%
  • If 40 people see a crime, the likelihood of any individual calling the police drops to 10%
  • The likelihood of any individual calling the police is lower, but the total likelihood that anyone will call the police actually rises, in this case to 40%
  • Therefore, on net average, if 4 people see a crime, 1 will call the police. If 48 people see a crime, 4 people will call the police.
But the story is that zero people called -- in other words, it posits that inhibition increases at a faster rate than group size, so the larger the group, the lower the likelihood of action, not just for any individual, but for the group as a whole.

Otherwise, it wouldn't be such a dramatic story - "Big groups are more likely to act than small groups, but not quite by as large a margin as you might expect" just doesn't make for a really wow meme.
posted by Bugbread at 9:32 PM on July 17, 2019 [1 favorite]


(Whoops "48 people" -> "40 people")
posted by Bugbread at 9:42 PM on July 17, 2019


Kitty Genovese's brother made an Oscar-nominated film investigating the claims that people did nothing and pretty much debunking it via primary sources. It was on Netflix last year; I don't know about now, but it's well worth the time if you can find it.
posted by DarlingBri at 5:12 AM on July 18, 2019 [2 favorites]


« Older "It's called a green screen Morty"   |   How to, home and car edition Newer »


This thread has been archived and is closed to new comments