[Broken_Hippo]

How would you even trust such results? I mean, lets look at where they come from.

Reported crimes in systems that are biased against some folks, with laws that are easier to break and get caught with when you are poor. Some of the folks are poor because of the situations that have happened to them - war, civil unrest, domestic violence, racism and discrimination of the past still affecting them. And lets not forget, a lot of folks don't report crimes like rape - this is higher in some countries than others. In some cases, reporting a rape might ruin your life.

How can you trust numbers coming from such systems when they say that skin color x or cultural background y does more crimes? We've had reports that do say such things, and the inherent biases in them generally make them untrue.

[jbmorgado]

You are setting up a framework of thought where no matter how much data you gather that shows some group is indeed more violent, you are allowed to just dismiss it on the basis of some situational perceived discrimination.

[Broken_Hippo]

Not really, I'm just aware that a lot of these sorts of grouping stereotypes are wrong - and much of the past studies have been flawed by discriminatory views.

For example, take the "poor people commit more crimes". Now, I don't know that this is true, though I know folks believe it and some studies have seemed to back it up. But I also know that a lot of poor folks lack ways to hide their crime (no private place to smoke pot in, for example) and live in areas more likely to be policed heavily. The only thing I can really come up with is that poor people get caught for crimes more often, and so are easily over-represented in the prison system. The conclusion that "poor people commit more crime" is one of bias.

This sort of thing is enough to cast doubt on a lot of these sorts of statements and check for such biases. It is a bit different from the data actually being true with the cause of the trend being discrimination. For example, "Felons often return to crime after release in the US" is likely true. Part of the reason for that is because of discrimination and biases against the ex-felon (especially in certain subsets of felons). I'd not ignore the data, but address the discrimination when trying to fix it.

[powertower]

> How would you even trust such results?

If those results are universal (meaning they are evident in all locations regardless of the location's history), then its likely extremely true.

Another way is to simply use your eyes, ears, and common sense.