""Keith, I am prepared to accept that this over representation exists in the specified places. This "over-rep" in itself does not "prove" the theory though.""

Added to which, the "over representation" is such that, in addition to being statistically dodgy because of the sampling method and the sample size, it is just as likely to be a case of criminal activity of a localised group of interacting gangs who are Pakistani because gangs tend to consist of just one ethnic group.

This is a concept which Keith will not even discuss, but is a valid alternative to his "cultural predilection" theory, and about ten times as likely to be true.

And in addition, he has once again avoided a direct answer to any of my points because he HAS no answer.

Don T.