Shoot/Don't Shoot With a Cultural Twist

[Guardian]

This from one of the 'Usual Suspects'.....but the conclusion that this 'experiment' is proof of the individual bias of the tested is a leap in logic.....it's based on the flawed ASSumption that I want to shoot black people more than white people....when in the 'test' i'm LOOKING for a gun...pure and simple. And in doing so am FOCUSED on looking at the hand, and hitting the button as SOON as I can visually identify the threat.

So, using OCCAM's RAZOR, which provides a more RATIONAL explaination for differences?

A) That some SUBCONCIOUS mechanism drives us to 'shoot black people' against our will.

B) That visual cues in the test, different physical color combinations, make it take longer to visually identify weapons in certain frames of the test versus the other. Where the issue is ONLY VISUAL ACCUITY!

My suspicion is that there IS BIAS involved.....RESEARCHER BIAS!

I agree with Sgtmac here, your not looking for the individuals color, your looking for the weapon and that specifically, I took this test and my whole time was spent looking for the weapon (the threat), not the individuals color, this test is flawed beyond measure in my view.

 

[punisher73]

One of the basic considerations in the design of a 'good' scientific experiment is the limitation of the number of variables.

If this test were to be done for maximal accuracy and validity, I would recommend a double-blind study with two groups taking two almost-identical tests, just with groups One and Two seeing the EXACT same pictures, except the race of the person holding the gun/object is reversed. Then I would do an analysis of the results to see if the difference in reaction time is within the statistical variation in the data. This would remove the artifacts Carol alludes to (like the fact that the gun the last guy is holding is camoflauged against the train in the background - it would be camoflauged identically for both test Groups).

You beat me to it. They should have the same pictures/poses and type of gun for both groups. You should have them as close as possible in appearance except for skin color if that is what you are really testing.

 

[CoryKS]

Game Over
Your Score: 180
Average reaction time:
Black Armed:850.56ms
Black Unarmed:1057.16ms
White Armed:813.48ms
White Unarmed:1060.36ms

Had a rough start - was down like -120 before I realized I was reacting too slow and getting docked for it.

I suspect that Carol has it right. It did seem like some photos were clearer than others. The cellphone raised against the wooded background was a nice touch.

Obviously, I'm racist. I have no other explanation for why I was so eager to pop a cap in whitey. :shrug:

 

[Deaf Smith]

The study is very poor. I can see it's done by 'book' people and not street people. Why? There is no context. Images do not convey the actual context of the encounter.

Like why did the person come up to you?

Were are you (bad part of town)?

What time of day or night?

Was it a familiar area for you? What was said?

What was the body language the other person was using (indicators gents, indicators(!)?

Did they have friends? Did you have friends or loved ones to protect?

Just showing someone with different attire, different race, different objects is not enough. It was a poor study done poorly but academics with no experience outside of their schools.

Oh, and not all people with guns, expecially since THEY ARE NOT POINTING THEM AT YOU, are bad guys. All I did was look for the weapon, I didn't look at their faces or color or atire.

Didn't shoot any without a gun, but I was slow a few times holstering or fireing. But that's not a fault. You should identify what you are shooting anyway.

Poor study it was.

Deaf

 

[Mr G]

I'm sorry to say, but I think it would be pretty hard to prove anything with a computer / internet basis scenario like this. It doesn't make it without value though. If we all look at the test and think reflectively about the results, valid or not, it did something... What made me fire? How hard is it to tell the cell phone from the gun? What if...? What if...?

Please note: think about your own results, not everyone else's....

 

[championmarius]

I'll admit, I felt the test was a bit... skewed.
I think I know why, in my case.

I was scanning for objects in hands, then focusing in on it. I had a hell of a time making contrast between the hands of the blacks and the items as opposed to the hands of the whites...

My times:
Black Armed:658.92ms
Black Unarmed:676.6ms
White Armed:606.92ms
White Unarmed:634.72ms

I could identify and track the hands of the whites a helluva lot faster than I could the blacks. an almost 50ms difference. either way, both armed and unarmed.

 

[Archangel M]

I also dont quite understand the point of "holstering speed" in "real life" all that matters is if you do/dont pull the trigger. How fast you reholster is irrelevant.