SJL wrote:
*** Please pay close attention when replying to a message on this list!
*** If you want the reply to go to the list, use REPLY-TO-ALL
*** Recruiters may advertise only permanent employment positions in this list.
All -
See the full paper here:
http://www.urban.org/UploadedPDF/411562_Salzman_Science.pdf
Study Conclusions excerpted below.
Please don't hijack this thread with unnecessary rants, and don't bother
commenting unless you have actually read the study at the link above.
I will have more to say on this topic soon...
I'm not yet clear on how to interpret much of it.
After a solid introductory section, the paper starts getting into
details that attempt to explain various international testing results. A
couple testing groups are discussed, along with comparative results. The
U.S.A. seems to compare well and to be on slowly rising trend. (I.e., as
long as it's understood, according to the study, that the U.S.A. has
been slowly increasing the participation of lower achievement students,
that is.)
On page 22, it is discussing some likely reasons for the U.S.A. PISA
scores. One significant factor:
"
The United States is one of six OECD countries where socioeconomic
level has a strong impact on student performance, where the average
score will be more affected by the size of the low socioeconomic
population, or, conversely, where schools are least effective in
providing an education that improves the test performance of
students from low socioeconomic, immigrant, and/or single-parent
families (OECD 2004, 182).
"
In other words, the U.S.A. has a major gap between what it provides to
the privileged portion of our population and what it provides to the
poor or disadvantaged. The study suggests that we actually do reasonably
well, on the whole, in education. We do well enough to be in the upper
group of nations.
Now, given that a gap does exist (according to the research), the
quality of K-12 for the privileged manages to raise them into the
world-elite range while the poor necessarily get educational treatment
equivalent to lower tier nations. The upper group must be much higher in
order to keep our average up.
This effect is apparently magnified -- "the average score will be more
affected by the size of the low socioeconomic population". If it's
pulled down to a greater degree than it's pulled up, and we still rank
well among nations, then K-12 for the privileged must be far superior to
have such an ability to offset the stronger downward pull.
...and later...
"
One could argue that it is the diversity and openness of the United
States that both contribute to its high economic performance and its
lower average educational performance.
"
...and especially...
"
The test results indicate that, rather than a policy focus on
average science and math scores, there is an urgent need for
targeted educational improvement to serve low performing
populations, such as recent immigrants and some minorities.
"
I have a hard time making sense of what this 'study' actually is. At
points, it seems to work at demonstrating that the U.S.A. has always
succeeded because of and is getting stronger due to significant
immigration. It _seems_ to lay out an argument for encouraging
immigration and then for raising our efforts to educate immigrants.
Unfortunately, it does a bad job of it. It can't count and it can't keep
track of its numbers.
It can't even demonstrate that it knows what to count nor what numbers
should be used:
"
Of course, net employment growth is not a direct measure of
employment demand or total job openings, since net growth does not
include replacement for retirements or occupational quits, nor do
these aggregate numbers indicate the types of workers sought
(education level, experience, etc). Moreover,...
"
If we think in terms of 40-year careers, then we immediately must
consider that the top one-fortieth of the whole population (S&E careers)
is dying or retiring every year. But how large is that population? The
numbers change significantly in different parts of the study.
Most strangely, there are footnotes:
"
[19]This simple calculation appears not to square with a comparison of
the annualized growth rate of S&E graduates and jobs from 1980 to
2000. That calculation finds that the annual growth rate of S&E
graduates at all degree levels is about one third that of S&E
employment growth (1.5 versus 4.2 percent annually). But the rate of
growth argument is somewhat misleading insofar as the slower growth
rate of S&E graduates is, as noted here, a far larger number than
the smaller but more rapidly growing number of S&E jobs.
"
From the spreadsheets noted later in that footnote:
2002 S&E employed: 5,412,000 (4.2%=227,304)(1-in-40=135,300)
2002 Bachelor's: 415,611
2002 Master's: 99,173
2002 Doctoral: 24,571
2002 total S&E degrees: 539,355 (33%=177,987)
http://www.nsf.gov/statistics/seind06/pdf_v2.htm
Let's see... almost five and a half million S&E jobs growing at a rate
of 4.2% is somehow "smaller" than the the number of graduates of just
over half a million? The study lightly dismisses quits, retirements and
deaths, and then dismisses a growth rate that's almost three times the
growth rate of graduates, and it's a higher growth rate for a population
that's a full order of magnitude larger!
Much of the study is concerned with why such a high number of S&E
graduates (two-thirds of them) don't go into S&E jobs. Some degrees are
awarded to those who already have S&E jobs (continuing education). Some
graduates use their S&E degrees for non-S&E jobs (e.g., 'patent
attorneys). Various reasons contribute to the drain of graduates into
other areas.
What we end up with are 178000 graduates per year entering S&E fields.
When there are 227000 new jobs, we already have a shortage. When we
think there are reasonably another 135000 job openings from deaths and
retirement, we find a basic shortage of 184000 jobs every year and a
shortage that is increasing far faster than the number of graduates.
At least, those are the numbers according to the spreadsheets referenced
in footnote [19]. Those numbers were used there to make a particular
point. When a different point is made in the study...?
Then comes "Table 3." It's included a page or so later in the study. It
shows numbers for 1995 and 2001 that are significantly higher than the
numbers for 2002 that we just spoke about, higher by perhaps 50%.
Different sources, of course... Unfortunately, the study doesn't provide
a link for Table 3 figures; they're referenced as "special tabulations",
whatever those are.
It's not clear why there were 844,500 graduates with Bachelor's and
Master's degrees in 1995, while 2002 only had 539,355 even including
Doctoral degrees. In 2001, the Table 3 Bachelor's plus Master's total is
918,400. Clearly, there would have to have been a precipitous drop to
fall all the way back to 540,000 in 2002 (including Doctorals!); yet the
study assures us that no such drop occurred. In fact, it assures us that
the number has continued to rise.
There is no shortage!
Well, maybe there is and maybe there isn't.
There certainly was a shortage of competence in compiling this study. It
makes no sense.
It's inconsistent from page to page. It excuses its own shortcomings. It
can't recognize a "smaller" number. It can't draw a correct conclusion
from different growth rates in differently sized populations. When it
talks /about/ numbers as in footnote [19], it provides a link for you to
dig through numerous downloadable spreadsheets to see if you can find
the numbers. When it provides numbers as in Table 3, it doesn't give
information for you to determine what the numbers mean.
Overall, I can't comprehend how this study managed to get itself made
public. I'd consider it an embarrassment.
Tom Liotta
midrange.com
