Re: 2000 Education by Race

[SamL <saml@demog.berkeley.edu>]

I am not saying they are inferior.  What I said is that "their educational
attainment does not reflect the performance of the U.S. system of
education."  That goes regardless of whether the numbers suggest high or
low ed attainment--one still can't use Census Bureau numbers to assess
the U.S. system.

Using these data is like judging a hospital's effectiveness by calculating
the death rate of people in the surrounding county.  Only some of the
people actually went to the hospital, whereas others may have gone to
other treatment facilities, or received no treatment, or died before they
could get treatment.  One doesn't have to suggest anything about those
other possibilities--not going to the hospital, going to some other
hospital, or dying before one can get to a hospital--to affirm that the
county-level data will *not* provide good data on the effectiveness of the
hospital, which could have a death rate of zero, and just look bad because
of the people who don't get to the hospital.  Or the hospital could have a
horribly high death rate, but look good because of all the people who
survive by going to other facilities.  The county-level data will mask the
hospital's performance, and we will have no way to unmask it if we treat
the county-level data as indicative of the hospital's performance.  The
same goes for using the census data to assess schools in the U.S.; the 25+
year olds could have been educated anywhere, and we have no way to find
that out.  So, the data can't address how the U.S. education system is or
has been working.

We should be talking about good data and what features it needs to have,
not assuming anyone is denigrating anything.  And, the Harvard/Urban
Institute data--I can't speak to that, because I wasn't talking about
*all* data.  I was talking about data from the census.  My own preference
would be something like NELS, or the soon-to-be-released ELS, or local
studies of cohorts.  Those are all much better for addressing the
question, because they tell us, of the students who attended school in the
location of interest, what proportion did not graduate, what proportion
went to college, and so on.  To clarify, note that the census data allows
someone to drop out of school in 1960 in Holland, immigrate to the U.S.
sometime in the intervening years, and show up at age 58 or so in the 2000
census as a drop-out.  How does this tell us *anything* about the
functioning of the U.S. school system?

All the census data can tell us, at best, is what proportion of people of
a certain age/sex/race in the society now have what level of education.
That is an important question in terms of what the labor market situation
looks like, but it's very bad data for assessing the work of schools.  I
apologize for assuming that the presentation of the table on this list was
to help assess schools.

Sam

On Sun, 29 Feb 2004, John Lawhead wrote:

>
> >Migration means that the people in the data may not have even been
> >children in the U.S., and thus their educational attainment does not
> >reflect the performance of the U.S. system of education.  Hence, one
> >cannot really use the data below to say much of anything.
> >
>
> The schools in Latin American countries aren't so grossly inferior that
> teenage immigrants can't claim the right to get a diploma on time with the
> proper support, including bilingual ed.   That there are kids with
> interrupted schooling is a different issue and in my experience it's a
> relatively small number.  Our students from Ecuador come to Brooklyn with
> math and science knowledge ahead of the rest of school.  They end up getting
> hammered in 11th grade by the English Regents exam, especially the boys.
> According to the Harvard/Urban Institute study New York state has the lowest
> rate of Latinos getting diplomas in the country.  In this case, I think the
> number means something.
>
> John
>
> _________________________________________________________________
> Watch high-quality video with fast playback at MSN Video. Free!
> http://click.atdmt.com/AVE/go/onm00200365ave/direct/01/
>