Kevin Tarr wrote:
>
> I found a web site which had an excel file with international crime data:
>
> http://www.aic.gov.au/stats/international/
>
> Hardly a rigorous study: I took the data and did some things with it. I
> divided each incident of crime by the population, averaged each crime
> statistic and 'normalized'* the crime average for each country to the
> average homicide rate. Then I averaged the crimes for each country.
>
...
>
> England & Wales 8.37% 7.77% 7
> U.S.A. 5.61% 8.08% 6
>
> *I took the average of the crime statistic, divided the homicide average,
> then multiplied by the specific country's rate for that crime.
> _______________________________________________
Kevin-- Please explain your procedure a bit more clearly. What do you
mean by "average", anyway? The original table has overall numbers, and
(at least for homicide) per capita incidence.
You produce some counter-intuitive results. I'm just
looking at the U.S. and England above. As one might expect, the U.S.
has over 3 times the per capita homicide rate of England. But its
rate for "Table 1 crimes" is almost 3 times less. To me, this
means that different countries put different crimes on "Table 1".
That is supported by footnote (1), which says exactly that.
I bet one would obtain much more useful results by simply
extrapolating the number of "serious crimes" from the number of
homicides. That is, just comparing based on the per capita incidence
of homicides.
David Hobby
My table is at work*. I know that some countries tabulate crimes
differently, but overall there would be a normal or average definition of
each crime category. So if a country had more assaults, it may just be
counting types of assaults that other countries don't. (But now I just
realized that I may be compounding that error). So I was comparing each
countries individual crime amount, for each crime, against the average of
all countries crime amount. The counting procedure may be inflated for some
crimes, but a stolen vehicle is a stolen vehicle, a rape is a rape, it
would be hard to say that a country has inflated ways of counting in ALL
categories. So each country ends up with a certain number for each crime,
and I compared each country from all the crime statistics, not just one.
Again, it wasn't a rigorous study, I just played with the numbers in a way
that looked right in my mind, and ended up with those number. I'm very
surprised that NZ and Aus are so high,. tomorrow I'll look to see why. On
the other hand, yes U.S.A. homicide is high, but I believe other countries
have overall higher crime rates. So I don't think your method sounds right.
One category, homicide, would skew everything else.
Hey I'm just fooking around at work, waiting for the next project.
Kevin T.
YES! Yanks went ahead!
*I sent myself the e-mail from work, then forwarded to list.
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l
