Why good people stand by and let bad things happen

You may have already seen the YouTube video of Michael Nodianos – if you have the stomach for it.

In what looks like a basement rec room, the teenager – his cheeks red with alcohol, and looking much younger than his Ohio State T-shirt suggests – wobbles on a chair, firing loathsome jokes to a too-willing audience of boys.

"That girl," he says, giggling, is "deader than Obi-Wan Kenobi when Darth Vader cut his head off. Deader than a door nail."

And this: "She is so raped right now."

This goes on for 12 minutes, in a video taped the night two football players did indeed rape a 16-year-old girl, who was so drunk she didn't know what happened until the next day.

There is no victim in the video. It's unclear whether the perpetrators of the assault are present. Mr. Nodianos is the star here, the rest of the boys (seen only when the one holding the camera laughs so hard he can't keep it straight) are passive voices from the sidelines.

But then, you can't get the full story of what went down last August in Steubenville, Ohio, without seeing the crowd who stood on the sidelines. By the time the video was shot, the victim had been taken to at least three parties, attended by dozens of teenagers. She was photographed being carried by her wrists and ankles, clearly unconscious. Between parties, she was sexually assaulted in a car with others watching.

All night long, at all the points when a rape could have been prevented, in a town small enough that teenagers at a summer party would know each other, no one stepped in to help. Last weekend, the perpetrators were convicted in large part because of the digital trail of evidence from hundreds of texts, incriminating video and pictures.

Ohio law makes it a crime not to report knowledge of a felony, such as rape, to the police. But morality has to come into play too. So what happened in Steubenville? Why did so many people document, rather than stop, what was going on? Was it just "human nature," or were other outcomes possible?

Psychologists talk about the "bystander effect," the phenomenon in which people in a crowd are less likely to risk their necks for someone in danger. But new research suggests that being in a group can also work to embolden us – to facilitate "acts of moral courage." And new work is under way to test how, with the right training and education, we might engineer a more positive bystander effect.

The "bystander effect" was first coined in 1968, by two American psychologists, John Darley and Bibb Latane, who began conducting lab experiments which simulated an emergency and then monitored whether people reacted differently alone or in a crowd. Their research was spurred by the prominent murder of Kitty Genovese, a New York woman whose neighbours supposedly ignored her screams for help. That part of the story has since been cast into doubt, but the theory held up in studies, and there have been endless real-life cases that seem to support their findings.

In 2009, people watched as a 15-year-old girl was raped outside a homecoming dance in California. That same year, a man stepped in to help two children being attacked by thugs at a German train station – only to be killed by those thugs in front of witnesses who did not come to his aid. Just last year, a photographer in New York made headlines when, instead of helping a man pushed onto the subway tracks, he took a picture of his last minutes.

Certainly, in the Steubenville context, it's not hard to recognize key elements of the "bystander effect" at work. First, there is a "diffusion of responsibility," where witnesses assume someone else, perhaps a friend, will step into take care of the victim. Then there's "evaluation apprehension," anxiety about the social cost of intervening (in this case, standing up to football players revered as demigods). The group may also enforce a "pluralistic ignorance," where witnesses fail to even recognize a crime (if everyone is laughing and texting while a girl is dragged around and abused for kicks, the message is that the behaviour is acceptable).

In Ohio, the witnesses were beer-chugging, peer-pressured teenagers. But adults have proved just as susceptible to the fear of being humiliated or judged for misinterpreting a dangerous situation, overcoming reason and common sense. In one of the most famour bystander experiments, participants were asked to fill out a questionnaire in a room that began filling with smoke. If they were alone, they sought help almost immediately. If they were with a group in which everyone else ignored what was happening, though, the majority of participants squirmed in their seats and, later, many didn't even report seeing the smoke.

But while human behaviour tends to fall into patterns, they are more complex than we might think. In a recent study, German psychologists pored through the results of more than 100 bystander experiments and data from 7,700 participants going back half a century. They concluded that under certain (rather surprising) circumstances, being in a group can also be a powerful motivator to take action.

It turns out that when a situation is especially dangerous, and when the risk to the person helping a victim is highest, the bystander effect actually melts away. In these cases, researchers theorize, the emergency itself tends to be less ambiguous, there is a greater cost in not helping, and the bystanders expect help from others in the crowd.

One of the most courageous examples is the events on United Flight 93 on 9/11. Realizing what the hijackers intended, a few passengers chose to take action, correctly assumed support from the group, and brought down the plane so it couldn't take more lives by hitting another populated target.

In fact, even the Genovese case may be more revealing of new thinking than the traditional view of a passive crowd. For decades, the story held that as many as 30 witnesses had heard her screams or seen the attack, and done nothing. But new research has found that many residents didn't recognize the scream – the nature of the emergency wasn't clear. Reports now suggest that a few people did dial 911, and an eyewitness called out to stop the attacker, but it was too late.

Figuring out what successfully spurs timely action is the next step for researchers. They are now testing programs designed to improve the odds that bystanders will act.

Giving bystanders specific, non-threatening strategies for stopping bad behaviour – creating a distraction to stop an altercation, or pretending to know the victim – can help. More important, though, is reframing what we consider dangerous situations, by addressing "pluralistic ignorance."

At the University of Windsor, for example, psychologist Charlene Senn is studying a new campus program to help students recognize the warning signs of sexual assault, so that, unlike the kids in Steubenville, they do not shrug off a drunk girl being led away from a party. It also debunks prevailing myths about rape (that victims are to blame for drinking, or lying, or "asking for it" somehow).

The key is help bystanders recognize troubling behaviour, and feel confident that others will support them if they stand up and act. "We are not talking about prevention at the moment of a sexual assault," says Ann Coker, a public health researcher at the University of Kentucky who is evaluating a high-school program called Green Dot, which includes motivational speeches and small group workshops to teach practical intervention techniques. "We are talking about behaviour that makes you feel like you should say something."

It's also vital to widen the net of responsibility, to make safety everybody's business. As Prof. Senn says, "There is so much we won't notice because we don't feel responsible."

How responsible we feel for others at risk depends on how closely we relate to them. We are far more likely to help if we see a victim as part of our in-group, and other bystanders as allies. British researchers found that Manchester United fans, for instance, were more likely to help someone in a bar fight if they were wearing Manchester United jerseys. They were also more likely to step in if other bystanders were wearing Manchester jerseys.

Social media can play a part: In Steubenville, technology seemed to distance observers from what they were seeing and reinforce, through texts, that there was nothing alarming about what was happening. But group dynamics can also strengthen our sense of responsibility, that we are being watched by those with shared values.

In a recent American experiment, for instance, participants were asked to make comments on an online forum for people struggling with severe depression. When their names were highlighted onscreen, or if they felt more visible to observers, they left more comments. In other words, a sense of group cohesion spurred action.

In Steubenville, silence was the norm. (Or, too often, when people did speak up, they blamed the victim.) More than a dozen potential witnesses refused to testify. Rather than going to the police, incriminating video was deleted. As the town's sheriff told The New York Times, "If you could charge people for not being decent human beings, a lot of people would have been charged that night."

But the law, in these cases, arrives too late, for victims, and even for the perpetrators – as researcher Ann Coker points out, "I would be willing to bet those football players wish someone had intervened that night."

That's the "silver bullet," as she calls it, to uncover the key to making us step in when it really counts, before the damage is done.