Re: Cohen data suspect at best

[RuthWeiner@AOL.COM]

In a message dated 1/28/02 4:04:04 PM Mountain Standard Time, healthrad@HOTMAIL.COM writes:



> .  For example, Dr. Field told me that if you were using 
>
> census track data to determine how many black people were in a certain area, 
>
> it would be correct less than 20% of the time


On the contrary. If you look at the Bureau of Census tables on the Web you will find them very precise as far as residence is concerned.  Census "undercounting" does occur: the homeless in cities, many tribal people on reservations, some of the rural poor.  Undercounting has been estimated as high as 10% among the urban homeless, but even that is not exclusively African American (come to Albuquerque if you want to see "white" homeless people).


If census tabulations were off by 80% there would be no point in funding a census or in using census data.  


Ruth Weiner, Ph. D.

ruthweiner@aol.com