A series of experiments I'd like to see.


PDA






Warren
May 15, 2007, 06:24 AM
Take a group of random people, give them saftey glasses have them change into different clothes and put them into a mocked-up school or office environment.

Assign them all roles to play.

Tell them they get X amount of dollars even if they are "killed" but X*3 if they "survive".

Release a "gunman" into the environment. He is armed with simuntions or a laser tag type thing or even a regular paint-ball gun.

He gets $Y for each person he marks.

He could start anywhere. He could come from outside or inside. He could be any of the roles. The others do not know. In fact the "gunman" was randomly chosen from the overall group.

They key here is that none of the others has a "gun". There are many things about that could be used as weapons but no guns.

In the interest of safety likely items to be used as improvised weapons would be "break-away" or made of styrofoam or other light materials.

Umpires would determine the effect of the use of any such items.

The test would be to see how many people he could mark up to the time the "police" arrive or he is stopped by other means. The police response time would be a random amount of time but withen a spread of actual response times to such incidents.

Run the test multiple times with different groups of people and see what the average number of "victims" is.

Next run the test with one other random person having a "gun". Run it the same number of times as the other test and then compare the number of victims.

Finally additional groups of tests would be run: Each time increasing the number of "good guy guns".

I believe that there would be a marked difference in the casualty rates between the first two scenarios. As the number of good guy guns increase the number of casualties would decline toward zero. Also we would see if "good guys" shoot each other.

A variant could be police response (using real cops?) and "good guy" defender in the same building and see if the cops shoot the good guy.

Anyone have any ideas to make this more realistic or impatial?

This would be expensive, any ideas as to who could fund it?

If you enjoyed reading about "A series of experiments I'd like to see." here in TheHighRoad.org archive, you'll LOVE our community. Come join TheHighRoad.org today for the full version!
TCB in TN
May 15, 2007, 07:01 AM
No idea who would/could fund it, but I think your conclusions are pretty spot on. More good guy guns= lower casualty rates.

Mannix
May 15, 2007, 07:13 AM
Too many variables to control, and for it to be anywhere near being useful, it would have to be unethical. I also have a feeling either telling people what they are in for is going to mess it up, or people fighting back after being hit is going to mess it up. IMO, it would make a great commercial, but the idea is not very feasible for a study.

DragonFire
May 15, 2007, 08:32 AM
I think one thing that will mess up the results is mindset.

No matter how much or how little you tell the participants they will be thinking differently than people do/have done in similar situations.

They will have to know that there is no chance of real harm, so they may take chances they wouldn't take in a real situation. Or they may think they have to fight back, where we've seen that few people do take action in real life.

Plus they will understand the mindset of the "shooter". They know his intent will be to tag as many people as he can, and act accordingly, while in real life people get into a denial mode, where they believe that the shooter won't shoot them (even if he's shot the person next to them).


Besides, no matter what results you do get, the anti-gun crowd will twist them to try to show they're right.

Zen21Tao
May 15, 2007, 08:33 AM
Such an "experiment" would fail to test what it was designed for. As others have mentioned, even if research participants didn’t know what to expect, just knowing they were part of a study would put them "on alert" such that they wouldn't act the same as if something completely unexpected happened. Also, such an experiment doesn't test reaction under real world conditions, only under the conditions controlled for. People are liable to respond quite differently to a real life-and-death situation than to a controlled condition. Anything that could succeed in making the condition any more realistic (e.g. deceptive study, blind study, etc.) would also make in highly unethical.

In light of such shortcomings the best approach wouldn’t be an experiment but would be a descriptive study analyzing numerous real world incidents. Such a study could also include self-report questionnaires of criminals and victims.

October
May 15, 2007, 11:37 AM
Tell them they get X amount of dollars even if they are "killed" but X*3 if they "survive".

Seems to me all you've done is set up a big hide and seek game. With the incentives you provide, everyone's primary goal is to avoid contact with anyone until the game ends (i.e., the police arrive). Even if you provide the victims with guns, game theory says anyone with a gun should still hide rather than trying to confront the shooter since doing so would risk the higher payout.

Warren
May 16, 2007, 06:33 PM
I though about some things last night, after I had
shut down the computer of course.

To disguise the nature of the experiment the
participants would not be told what was going to
happen.

They would know it was to test for survival instincts
or methods in a crisis situation.

So it could be a building fire, earthquake, tornado,
bomb threat, vermin incursion, NBC attack, mass
illness and so on. A mass shooting need not even be
mentioned.

The bad guy would be the only one who would know what
was about to happen.

The random people given guns would not be told why
they have a gun, just that they could, if needed, use
it to ensure their survival.

Other variants could be set in a mall, aircraft,
sporting event or concert.

Or a multi-shooter incident, or a crazy man with a
blade of some sort.


To switch it around the experiment could be centered
on the bad guy.

To test the incentive process of an "evil" shooter he
could be given four choices of victim groups.

Group A would have the most people but also 80% of the
people would be armed.

Group B would have less people but only 10% would be
armed.

Group C would have even less people and only one
person would have a gun.

Group D would have the least people but also would
have no guns.

The shooter would get $X for every victim so I would
be interested to see what group gets hit the most.

This would not simulate insanity of course but we would be able to explore the why of the decision.

230RN
May 17, 2007, 01:39 AM
Y' know what, though? This ought to be done routinely anyhow to alert people to the possibility that they can, indeed, fight back.

Maybe every office manager ought to set up a training situation like that just in case.

'Course some teachers doing this in schools have got in trouble for it --and I can understand why, since there are no guns allowed inside schools anyhow, so there's no purpose to it there.

Zen21Tao
May 17, 2007, 09:02 AM
Let me say that I have a BS in (Research) Psychology and a BS in Physics. I am just giving constructive criticism so please don’t take negatively as just bashing your ideas.


To disguise the nature of the experiment the
participants would not be told what was going to
happen.

They would know it was to test for survival instincts
or methods in a crisis situation.

So it could be a building fire, earthquake, tornado,
bomb threat, vermin incursion, NBC attack, mass
illness and so on. A mass shooting need not even be
mentioned.

The bad guy would be the only one who would know what
was about to happen.


The experiment you talking about would either fall under Behavioral Psychology or (most likely) Social Psychology and as such would have to be approved by a research ethics board/committee. The type of study you are talking about is a deceptive study. These types of studies have extremely high ethical restrictions and I can assure you, the stress you would have to put people under to simulate a 'real world' situation would prevent any such study from being approved by any ethical research board. You simply can not replicate the environment necessary for your study. Any environmental conditions you could set up and get approved for wouldn't generalize to those found in the real world.

But lets say you could get approved for realistic environmental conditions. The next question is whether you could succeed at testing the reactions of everyday people that might find themselves in such an environment.

The random people given guns would not be told why
they have a gun, just that they could, if needed, use
it to ensure their survival.

Just giving people guns increases their state of alertness. They know they are part of a study, when you hand them a gun they know that the gun will have some role in the experiment. They will not be of the same mind set of someone legally carry a firearms with absolutely no reason to believe they will need to use the gun. This difference further complicates the study before it even begins. What you would be testing is the reaction of people given guns (not that chose to obtain guns) in an environment of a perceived unknown danger not the reaction of people armed in an environment where unperceived danger occurs.



Or a multi-shooter incident, or a crazy man with a
blade of some sort.


To switch it around the experiment could be centered
on the bad guy.

To test the incentive process of an "evil" shooter he
could be given four choices of victim groups.

Group A would have the most people but also 80% of the
people would be armed.

Group B would have less people but only 10% would be
armed.

Group C would have even less people and only one
person would have a gun.

Group D would have the least people but also would
have no guns.

This is a much better experiment but it still has its drawbacks. The shooter will know that this isn’t a life or death situation. He will know that the most he stands to loose is some amount of money. This does pose a real problem. Research has show people to be much less risk averse when it money at stake than their life. If study risk aversion you will find the mathematical formula that shows as the perceived risk increases the willingness to undergo the risk decrease. Quite simply, the people in you study would more likely choose group C or B thinking they would willingly risk 1% or even 10% armed people for the better payoff because they know it is only money they stand to lose. However, the mathematical formula for risk aversion would more than likely say that when their life is on the line they would chose group D or C (if the rewards is high enough). I think that historical studies are much more likely to confirm what the math predicts than experimental studies.

One should also ask how many times the shooter gets to “risk his luck.” This also makes a meaningful difference. A risk adverse person would gladly choose multiple participations with little to no risk than a single participation with moderate to high risk.

There is also a valid reason to think a criminal by his very nature is not as risk averse as the average citizen. He is willing to risk incarceration (if not death) for a monetary payoff. Such an experiment as performed above would test the ‘average’ person no the ‘risk taking’ person that commit most crime. To form such an experiment you would have to pre-select people already known to be ‘risk takers’ and then run them through the experiment. In this case you would say many more Bs and many less Cs.

Again, this has all been just my 2 cents. Feel free to disregard.

If you enjoyed reading about "A series of experiments I'd like to see." here in TheHighRoad.org archive, you'll LOVE our community. Come join TheHighRoad.org today for the full version!