On Jun 16, 2006, at 7:42 AM, Ted Winslow wrote:
> Doug Henwood wrote:
>
>> How's that? You have some examples that show both similiarities
>> and differences. Similarity: lots of guns. Differences: on the
>> dependent side, murders; on the independent side, imperial
>> history, frontier culture, inequality, racial heterogeneity... You
>> control for the similarities and try to see if the dissimilarities
>> are significant. That sounds like standard stats to me.
>
> This implicitly assumes that individuals are everywhere and always
> the same so that differences in behaviour are explainable solely by
> differences in circumstances.
I don't see how differences among individuals are explicable by anything but circumstances (allowing for "temperament," aka Freud's "constitutional factors," and random variation). History, inequality, family patterns, racial hetero/homogeneity - these are precisely the differences among societies you'd want to investigate to explain differences in crime rates.
Doug