Malcolm's data

1 view
Skip to first unread message

Ron McDermott

unread,
Oct 21, 1999, 3:00:00 AM10/21/99
to
I've been looking at the data Malcolm posted relating to math scores
and district size as well as starting age of schooling, and I've
noticed a couple interesting things which some of you may be able to
help me with.

First of all, there are 9 states missing from the list - why? These
states are: Idaho, Illinois, Nevada, New Hampshire, New Jersey, Ohio,
Oklahoma, Pennsylvania, South Dakota. This omission is equal to
roughly 22% of the provided data, which means that it is a significant
omission.

Then there's an additional relationship in the data. If you look
closely at the omitted states, they lie primarily in the northeastern
part of the country. Those states which have late starting ages
and/or small districts are similarly clustered regionally.
Specifically, there is a large cluster in the upper western states
(excluding California), another large cluster in the upper midwest,
virtually all of the extreme northeast, and another cluster in what is
often called the deep south. Now we know that all regions of the
country are not demographically comparable; specifically, the extreme
northeast, the upper midwest, and the upper western states are going
to have relatively small minority populations. This isn't true of the
deep south, but overall, the states which have late starting ages
and/or small districts tend to ALSO have small minority populations,
and those are likely to be concentrated in relatively few districts
for a given state (urban).

The population dense states: New York, California, Ohio, etc are going
to have large minority populations AND those populations are going to
be found in many or most of the districts in those states (due to the
number of large cities and the larger percentage of minority
population).

The comparisons, therefore, are not only based upon age and district
size, but are ALSO racial comparisons and comparisons between city
dwellers (high density populations) and suburban to rural dwellers.
Anyone who believes that these are inconsequential factors is
seriously deluded!

Now I notice that the correlations are based on the numbers found
under the percentiles, but those numbers are clearly not students, so
I have to guess that they represent districts which house 8th grade
students and it is the DISTRICTS which are being ranked as to
percentile (if I'm wrong here, I'd appreciate a "heads up"). So... If
one operates on that basis, one notices that the total number of
districts is pretty much the same from one state to another,
regardless of population. This means that populous states MUST have
larger districts, and thus Malcolm has now produced an artificial
separation of states by population, with the sparser states being the
ones which Malcolm uses to make his point. Any state with a large
number of districts which are in improverished areas, or which house
primarily minority populations, will tend to have a higher percentage
of its districts low in the percentiles, while those which have small
minority populations (or in which the minority populations are
concentrated in a small number of districts) will appear to be
"better".

The other issue, which I haven't explored, is the variation in the
percentages of districts which fall into the various percentiles from
state to state. I wonder if the range of variations is even
significant given the uncertainty in the scores themselves? For
example, if the degree of uncertainty in a given score is plus or
minus a percentage point, is a difference of 1 percentage point in the
breakdown of district percentages in a given percentile meaningful?

Imo, for Malcolm to claim that the data indicates that early schooling
leads to poor grades, or that large districts inhibit learning is
ludicrous; all the data appears to indicate is that minority students
do poorly in relation to non-minority students, and THIS we already
know!

panther

unread,
Oct 21, 1999, 3:00:00 AM10/21/99
to
In article <87XdtVqF0GQ7-pn2-JNZMdITBNY20@localhost>,

rmc...@nospam.banet.net (Ron McDermott) wrote:
> I've been looking at the data Malcolm posted relating to math scores
> and district size as well as starting age of schooling, and I've
> noticed a couple interesting things which some of you may be able to
> help me with.
>
> First of all, there are 9 states missing from the list - why?...
>
MK. Discussion deleted...
>
MK. Not all States participate.

>
> Then there's an additional relationship in the data. If you look
> closely at the omitted states, they lie primarily in the northeastern
> part of the country. Those states which have late starting ages
> and/or small districts are similarly clustered regionally.
> Specifically, there is a large cluster in the upper western states
> (excluding California), another large cluster in the upper midwest,
> virtually all of the extreme northeast, and another cluster in what is
> often called the deep south. Now we know that all regions of the
> country are not demographically comparable; specifically, the extreme
> northeast, the upper midwest, and the upper western states are going
> to have relatively small minority populations. This isn't true of the
> deep south, but overall, the states which have late starting ages
> and/or small districts tend to ALSO have small minority populations,
> and those are likely to be concentrated in relatively few districts
> for a given state (urban).
>
MK. You see the same relationship (earlier compulsory attendance and
lower scores) if you look at mean scores for Whites, or Blacks, or
Hispanics, or Asians alone.

>
> The population dense states: New York, California, Ohio, etc are going
> to have large minority populations AND those populations are going to
> be found in many or most of the districts in those states (due to the
> number of large cities and the larger percentage of minority
> population).
>
MK. Why do you call NY, Calif., and Ohio "population dense"? These are
large States, but not particularly dense. There are a number of
exceptions to your generalization. Hawaii, not particularly urban, with
one single district, mostly non-White. New Jersey, the most densely
populated State, with few large districts.

>
> The comparisons, therefore, are not only based upon age and district
> size, but are ALSO racial comparisons and comparisons between city
> dwellers (high density populations) and suburban to rural dwellers.
> Anyone who believes that these are inconsequential factors is
> seriously deluded!
>
MK. The effect remains when measured within racial groups. Except for
children of White, college-educated parents, whose scores rise as
districts increase in size.

>
> Now I notice that the correlations are based on the numbers found
> under the percentiles, but those numbers are clearly not students, so
> I have to guess that they represent districts which house 8th grade
> students and it is the DISTRICTS which are being ranked as to
> percentile (if I'm wrong here, I'd appreciate a "heads up")...
>
MK. Heads up!. These are aggregate student scores.
>
> ...So... If

> one operates on that basis, one notices that the total number of
> districts is pretty much the same from one state to another,
> regardless of population. This means that populous states MUST have
> larger districts, and thus Malcolm has now produced an artificial
> separation of states by population, with the sparser states being the
> ones which Malcolm uses to make his point. Any state with a large
> number of districts which are in improverished areas, or which house
> primarily minority populations, will tend to have a higher percentage
> of its districts low in the percentiles, while those which have small
> minority populations (or in which the minority populations are
> concentrated in a small number of districts) will appear to be
> "better".
>
> The other issue, which I haven't explored, is the variation in the
> percentages of districts which fall into the various percentiles from
> state to state. I wonder if the range of variations is even
> significant given the uncertainty in the scores themselves? For
> example, if the degree of uncertainty in a given score is plus or
> minus a percentage point, is a difference of 1 percentage point in the
> breakdown of district percentages in a given percentile meaningful?
>
> Imo, for Malcolm to claim that the data indicates that early schooling
> leads to poor grades, or that large districts inhibit learning is
> ludicrous; all the data appears to indicate is that minority students
> do poorly in relation to non-minority students, and THIS we already
> know!
>
MK. OK. Since the mistaken assumption about percentiles being rankings
of districts makes most of the above irrelevant, I will address only
your last point: Whites are a minority in Washington DC, yet the DC
White mean score is the highest in the nation. Whites are a minority in
Hawaii. Hawaiian/part Hawaiian are the largest ethnic group in Hawaii's
State schools. The White mean score (Numbers and Operations, parents
are high-school educated) is the lowest in the nation. The Black mean
score is the lowest in the nation. NAEP does not give separate scores
for Hawaiians and Asians, so I use Stanford Achievement Test scores,
here. A school's mean score rises as the Asian:Pacific Islander ratio
rises. Asians are the politically adept group in Hawaii. Hawaiians,
Blacks, and Whites are not.
>
Take care. Homeschool if you can.
>
www.schoolchoices.org (Massive site. Useful links).
>
www.hslda.org (Very useful links, for prospective homeschoolers)
>
www.rru.com/~meo/hs.minski.html (One page. Marvin Minsky comment on
school. Please read this.)


Sent via Deja.com http://www.deja.com/
Before you buy.

Ron McDermott

unread,
Oct 22, 1999, 3:00:00 AM10/22/99
to
On Thu, 21 Oct 1999 15:15:02, panther <igon...@my-deja.com> wrote:

> In article <87XdtVqF0GQ7-pn2-JNZMdITBNY20@localhost>,
> rmc...@nospam.banet.net (Ron McDermott) wrote:
> > I've been looking at the data Malcolm posted relating to math scores
> > and district size as well as starting age of schooling, and I've
> > noticed a couple interesting things which some of you may be able to
> > help me with.
> >
> > First of all, there are 9 states missing from the list - why?...
> >
> MK. Discussion deleted...
> >
> MK. Not all States participate.

Ok.. A potentially important 25% of the data is missing then.

> > Then there's an additional relationship in the data. If you look
> > closely at the omitted states, they lie primarily in the northeastern
> > part of the country. Those states which have late starting ages
> > and/or small districts are similarly clustered regionally.
> > Specifically, there is a large cluster in the upper western states
> > (excluding California), another large cluster in the upper midwest,
> > virtually all of the extreme northeast, and another cluster in what is
> > often called the deep south. Now we know that all regions of the
> > country are not demographically comparable; specifically, the extreme
> > northeast, the upper midwest, and the upper western states are going
> > to have relatively small minority populations. This isn't true of the
> > deep south, but overall, the states which have late starting ages
> > and/or small districts tend to ALSO have small minority populations,
> > and those are likely to be concentrated in relatively few districts
> > for a given state (urban).
> >
> MK. You see the same relationship (earlier compulsory attendance and
> lower scores) if you look at mean scores for Whites, or Blacks, or
> Hispanics, or Asians alone.

You didn't post this data; I can't respond to stuff which isn't there,
and the data you posted is not by student (as nearly as I can
determine).

> > The population dense states: New York, California, Ohio, etc are going
> > to have large minority populations AND those populations are going to
> > be found in many or most of the districts in those states (due to the
> > number of large cities and the larger percentage of minority
> > population).
> >
> MK. Why do you call NY, Calif., and Ohio "population dense"? These are
> large States, but not particularly dense.

I'm not thinking people per square mile here, but more along the lines
of the number of districts which exist in "large" cities. The Eastern
seaboard, particularly, is littered with a great many large population
cities, and this tendency decreases as one moves inland.

> There are a number of exceptions to your generalization.

There are exceptions to everything, in my experience.

<discussion deleted>

> > The comparisons, therefore, are not only based upon age and district
> > size, but are ALSO racial comparisons and comparisons between city
> > dwellers (high density populations) and suburban to rural dwellers.
> > Anyone who believes that these are inconsequential factors is
> > seriously deluded!
> >
> MK. The effect remains when measured within racial groups.

Data not provided. If you wish to alter the discussion to encompass
different data, you will have to provide that data.

> Except for children of White, college-educated parents, whose scores
> rise as districts increase in size.

Which strongly suggests a racial (and perhaps dwelling location)
factor is being evaluated here and not a beginning age or district
size factor.

> > Now I notice that the correlations are based on the numbers found
> > under the percentiles, but those numbers are clearly not students, so
> > I have to guess that they represent districts which house 8th grade
> > students and it is the DISTRICTS which are being ranked as to
> > percentile (if I'm wrong here, I'd appreciate a "heads up")...
> >
> MK. Heads up!. These are aggregate student scores.

You're going to have to clarify that. The totals of the numbers in
the percentiles are close are the same, yet the populations of the
states are very different. How it this "aggregate" defined?

We'll see; that depends on how the "aggregate" is being defined and
generated.

> I will address only your last point: Whites are a minority in Washington
> DC, yet the DC White mean score is the highest in the nation. Whites
> are a minority in Hawaii.

The implication intended is that all WHITES are the same everywhere?
That must mean all BLACKS are the same as well <g>?

> Hawaiian/part Hawaiian are the largest ethnic group in Hawaii's
> State schools. The White mean score (Numbers and Operations, parents
> are high-school educated) is the lowest in the nation. The Black mean
> score is the lowest in the nation.

So let's see now, what can we conclude from this single data point?
That Hawaii schools are poor? That kids don't attend school because
the weather is mostly good? That pineapples reduce IQ? That kids in
Hawaii are struck by falling coconuts at an early age? Maybe
saltwater is inimical to educational success? I'm sure there
excellent correlations between the scores and the weather, the
presence of pineapples and seawater, etc. Unfortunately, I suspect
those correlations have absolutely no causal relationship and are
merely chance juxtapositions of unrelated factors - as I believe yours
to be.

panther

unread,
Oct 23, 1999, 3:00:00 AM10/23/99
to
In article <87XdtVqF0GQ7-pn2-hokduL0QLsdw@slip166-72-232-
98.ny.us.prserv.net>,
rmc...@nospam.banet.net (Ron McDermott) wrote:

> panther wrote:
>
> > > I've been looking at the data Malcolm posted relating to math
scores...
>
MK. Discussion deleted...

>
> > > First of all, there are 9 states missing from the list - why?...
>
> > MK. Discussion deleted...
>
> > MK. Not all States participate.
>
> Ok.. A potentially important 25% of the data is missing then.
>
MK. A world of data are missing. I could give you other years (1990,
1992). The data look pretty much the same. The 1996 (posted) results
are the most recent NAEP 8th grade math data. You got something better?
>
MK. Discussion deleted...
>
> > >...Now we know that all regions of the country are not

demographically comparable; specifically, the extreme
> > > northeast, the upper midwest, and the upper western states are
going
> > > to have relatively small minority populations. This isn't true
of the
> > > deep south, but overall, the states which have late starting ages
> > > and/or small districts tend to ALSO have small minority
populations,
> > > and those are likely to be concentrated in relatively few
districts
> > > for a given state (urban).
> > >
> > MK. You see the same relationship (earlier compulsory attendance and
> > lower scores) if you look at mean scores for Whites, or Blacks, or
> > Hispanics, or Asians alone.
>
> You didn't post this data; I can't respond to stuff which isn't there,
> and the data you posted is not by student (as nearly as I can
> determine).
>
MK. You can respond to data I don't post: you can get it yourself.
However, I have since posted mean scores by parent's race and level of
education.
>
MK. Discussion deleted...

>
> > MK. Why do you call NY, Calif., and Ohio "population dense"? These
are
> > large States, but not particularly dense.
>
> I'm not thinking people per square mile here, but more along the lines
> of the number of districts which exist in "large" cities. The Eastern
> seaboard, particularly, is littered with a great many large population
> cities, and this tendency decreases as one moves inland.
>
> > There are a number of exceptions to your generalization.
>
> There are exceptions to everything, in my experience.
>
> <discussion deleted>
>
MK. Here Ron deletes exceptions which dispute the "urban density" as
opposed to "school district size' explanation for poor NAEP test
performance. I expect New Jersey and Florida provide illuminating
contrasts, also. Florida commits 80% of its students to large districts.
New Jersey, almost none (Newark and Jersey City). The average school
district in New Jersey is under 6,000 students, if I recall correctly.
Dade county, Fla., and Broward county, Fla., are among the nation's top
ten largest school districts.

>
> > > The comparisons, therefore, are not only based upon age and
district
> > > size, but are ALSO racial comparisons and comparisons between city
> > > dwellers (high density populations) and suburban to rural
dwellers.
> > > Anyone who believes that these are inconsequential factors is
> > > seriously deluded!
> > >
> > MK. The effect remains when measured within racial groups.
>
> Data not provided. If you wish to alter the discussion to encompass
> different data, you will have to provide that data.
>
MK. No, -you- brought race into it. I answered. Fornately, I also did
provide the data. Go back to "The data, the model, and the proposal".

>
> > Except for children of White, college-educated parents, whose scores
> > rise as districts increase in size.
>
> Which strongly suggests a racial (and perhaps dwelling location)
> factor is being evaluated here and not a beginning age or district
> size factor.
>
MK. It strongly suggests that district size and age are not the only
factors, but who ever said they were?

>
> > > Now I notice that the correlations are based on the numbers found
> > > under the percentiles, but those numbers are clearly not
students, so
> > > I have to guess that they represent districts which house 8th
grade
> > > students and it is the DISTRICTS which are being ranked as to
> > > percentile (if I'm wrong here, I'd appreciate a "heads up")...
> > >
> > MK. Heads up!. These are aggregate student scores.
>
> You're going to have to clarify that. The totals of the numbers in
> the percentiles are close are the same,...
>
MK. ??? Typo. Does not compute.
>
> ...yet the populations of the

> states are very different. How it this "aggregate" defined?
>
MK. Do you know what "percentile score" means? These are student
percentile scores: The score of the 10th percentile students, 25th
percentile students, 50th percentile students, etc., in the State.
>
> > > ...So... If
> > > one operates on that basis...
>
> > MK. Discussion deleted...

>
> > MK. OK. Since the mistaken assumption about percentiles being
rankings
> > of districts makes most of the above irrelevant,
>
> We'll see; that depends on how the "aggregate" is being defined and
> generated.
>
> > I will address only your last point: Whites are a minority in
Washington
> > DC, yet the DC White mean score is the highest in the nation. Whites
> > are a minority in Hawaii.
>
> The implication intended is that all WHITES are the same everywhere?
> That must mean all BLACKS are the same as well <g>?
>
MK. Get serious, or leave.

>
> > Hawaiian/part Hawaiian are the largest ethnic group in Hawaii's
> > State schools. The White mean score (Numbers and Operations, parents
> > are high-school educated) is the lowest in the nation. The Black
mean
> > score is the lowest in the nation.
>
> So let's see now, what can we conclude from this single data point?
> That Hawaii schools are poor?...
>
MK. "Wretched", I would say.
>
>...That kids don't attend school because the weather is mostly good?...
>
MK. The climate's nice in Singapore, also. Cutting a terrible school
probably boosts NAEP scores.
>
MK. The rest is even more frivolous. Inane sarcasm deleted...
>
MK. Is there a serious critic out there?

>
Take care. Homeschool if you can.
>
Or throw your baby to the sharks, if you think that will do the other
swimmers any good.

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
www.schoolchoices.org (Massive site. Useful links).
>
www.hslda.org (Very useful links, for prospective homeschoolers)
>
www.rru.com/~meo/hs.minski.html (One page. Marvin Minsky comment on

schooling. Please read this.)

Bob LeChevalier

unread,
Oct 24, 1999, 3:00:00 AM10/24/99
to
panther <igon...@my-deja.com> wrote:
>MK. Here Ron deletes exceptions which dispute the "urban density" as
>opposed to "school district size' explanation for poor NAEP test
>performance. I expect New Jersey and Florida provide illuminating
>contrasts, also. Florida commits 80% of its students to large districts.

No, Florida has legal district boundaries that people have settled in to
create large districts. Probably most of the lines were drawn when those
districts were much smaller than they are now (Miami had a population under
2,000 in 1900; Newark's population has grown 5% since then and Jersey City
10% (and both are smaller than they were a couple decades ago).

>New Jersey, almost none (Newark and Jersey City.

Whereas New Jersey has no center cities, only suburbs.

>The average school
>district in New Jersey is under 6,000 students, if I recall correctly.
>Dade county, Fla., and Broward county, Fla., are among the nation's top
>ten largest school districts.

Florida has 67 school districts and 67 counties; the counties were likely
set up long before anyone knew that some of them would come to house a
hundred thousand students. In short, they have chosen to reduce
bureaucracy by having school districts match up with the other legal
division of the state. Makes sense to me.

New Jersey is the inexplicable one, having hundreds of districts while only
having 19 counties. Think of all that bureaucracy, all those redundant
multi-layered tax-collecting authorities.

>> > Except for children of White, college-educated parents, whose scores
>> > rise as districts increase in size.
>>
>> Which strongly suggests a racial (and perhaps dwelling location)
>> factor is being evaluated here and not a beginning age or district
>> size factor.
>>
>MK. It strongly suggests that district size and age are not the only
>factors, but who ever said they were?

You have made claims that "schools are lethal institutions", and used
language such as "poison" within the last couple of days. This indicates a
claim that the district size and age are the EFFECTIVE cause "damaging
kids". You make this clear in your "Proposal":
>6. States with large school districts tend (strongly) to have low
>scores and a large gap between the top and bottom scores (Hawaii is no
>exception.)
>7. States with large school districts tend to have high rates of
>juvenile crime (Hawaii and Utah are exceptions.)
>8. Politically powerless minorities are poorly served by large school
>districts (Hawaii is not an exception.).

You use the word "tend" which by my dictionary implies a direct and/or
innate result. If you admit other factors, then you should not be making
claims of "tendency" as opposed to mere "correlation".

If these district parameters are not the only ones, then you must establish
that they are the most important or causal factors in order to justify your
vituperative remarks. So far you have presented no evidence that "school
districts" or tax money has anything to do with lethality, poison, or
sharks.

>> > I will address only your last point: Whites are a minority in Washington
>> > DC, yet the DC White mean score is the highest in the nation. Whites
>> > are a minority in Hawaii.
>>
>> The implication intended is that all WHITES are the same everywhere?
>> That must mean all BLACKS are the same as well <g>?
>>
>MK. Get serious, or leave.

You are the one who made the comparison, not Ron.

>> > Hawaiian/part Hawaiian are the largest ethnic group in Hawaii's
>> > State schools. The White mean score (Numbers and Operations, parents
>> > are high-school educated) is the lowest in the nation. The Black mean
>> > score is the lowest in the nation.
>>
>> So let's see now, what can we conclude from this single data point?
>>

>>...That kids don't attend school because the weather is mostly good?...
>>
>MK. The climate's nice in Singapore, also. Cutting a terrible school
>probably boosts NAEP scores.

I didn't notice any NAEP scores for Singapore in your data. How is it
relevant to the discussion?

>MK. The rest is even more frivolous. Inane sarcasm deleted...

Inane conclusions so far being ignored. And given your most politic
"proposal", I don't blame the legislature for ignoring you.

"The cost of this system includes losses due to crime,
and the cost of prison for the poor children whose lives you trash".
Indeed.

>MK. Is there a serious critic out there?

>...


>Or throw your baby to the sharks, if you think that will do the other
>swimmers any good.

Not when you make remarks like this one.

lojbab
----
lojbab ***NOTE NEW ADDRESS*** loj...@lojban.org
Bob LeChevalier, President, The Logical Language Group, Inc.
2904 Beau Lane, Fairfax VA 22031-1303 USA 703-385-0273
Artificial language Loglan/Lojban:
see Lojban WWW Server: href=" http://xiron.pc.helsinki.fi/lojban/ "
Order _The Complete Lojban Language_ - see our Web pages or ask me.

Ron McDermott

unread,
Oct 24, 1999, 3:00:00 AM10/24/99
to
On Sat, 23 Oct 1999 14:09:40, panther <igon...@my-deja.com> wrote:

> In article <87XdtVqF0GQ7-pn2-hokduL0QLsdw@slip166-72-232-
> 98.ny.us.prserv.net>,
> rmc...@nospam.banet.net (Ron McDermott) wrote:

> > > MK. Not all States participate.
> >
> > Ok.. A potentially important 25% of the data is missing then.
> >
> MK. A world of data are missing. I could give you other years (1990,
> 1992). The data look pretty much the same. The 1996 (posted) results
> are the most recent NAEP 8th grade math data. You got something better?

Nope, but it makes me cautious about inferring too much from the data.
The fact that the missing data is largely clumped in a specific
region is cause for concern.

> MK. Discussion deleted...
> >
> > > >...Now we know that all regions of the country are not
> > > >demographically comparable; specifically, the extreme
> > > > northeast, the upper midwest, and the upper western states
> > > > are going to have relatively small minority populations. This
> > > > isn't true of the deep south, but overall, the states which
> > > > have late starting ages and/or small districts tend to ALSO
> > > > have small minority populations, and those are likely to be
> > > > concentrated in relatively few districts for a given state
> > > > (urban).
> > >
> > > MK. You see the same relationship (earlier compulsory attendance and
> > > lower scores) if you look at mean scores for Whites, or Blacks, or
> > > Hispanics, or Asians alone.
> >
> > You didn't post this data; I can't respond to stuff which isn't there,
> > and the data you posted is not by student (as nearly as I can
> > determine).
> >
> MK. You can respond to data I don't post: you can get it yourself.

You're making the claims, it's up to you to substantiate them.

> However, I have since posted mean scores by parent's race and level of
> education.

I'll be looking at this as time permits.

> > > MK. Why do you call NY, Calif., and Ohio "population dense"? These
> > > are large States, but not particularly dense.
> >
> > I'm not thinking people per square mile here, but more along the lines
> > of the number of districts which exist in "large" cities. The Eastern
> > seaboard, particularly, is littered with a great many large population
> > cities, and this tendency decreases as one moves inland.
> >
> > > There are a number of exceptions to your generalization.
> >
> > There are exceptions to everything, in my experience.
> >
> > <discussion deleted>
> >
> MK. Here Ron deletes exceptions which dispute the "urban density" as
> opposed to "school district size' explanation for poor NAEP test
> performance.

Presenting a couple anecdotal cases is hardly compelling..

> I expect New Jersey and Florida provide illuminating contrasts, also.

Which we cannot establish since New Jersey is among the missing
states..

> Florida commits 80% of its students to large districts.

And has a huge latino population as well as some other minority
components, again clustered in the cities.

> New Jersey, almost none (Newark and Jersey City).

And we know that the cities perform poorly (at least I know since New
Jersey is "next door" to me).

> > > > The comparisons, therefore, are not only based upon age and
> > > > district size, but are ALSO racial comparisons and comparisons
> > > > between city dwellers (high density populations) and suburban
> > > > to rural dwellers. Anyone who believes that these are
> > > > inconsequential factors is seriously deluded!
> > > >
> > > MK. The effect remains when measured within racial groups.

To be continued after I look at THAT data.

> > > Except for children of White, college-educated parents, whose scores
> > > rise as districts increase in size.
> >
> > Which strongly suggests a racial (and perhaps dwelling location)
> > factor is being evaluated here and not a beginning age or district
> > size factor.
> >
> MK. It strongly suggests that district size and age are not the only
> factors, but who ever said they were?

You did when you claimed the data "proved" that early attendence and
large districts impede learning. It, of course, does no such thing,
and the fact that the pattern does not hold for whites, who are still
the majority of the population, makes it pretty clear that what is
showing up is something else, and that it is racially related.

> > > > Now I notice that the correlations are based on the numbers found
> > > > under the percentiles, but those numbers are clearly not
> > > > students, so I have to guess that they represent districts which
> > > > house 8th grade students and it is the DISTRICTS which are being
> > > > ranked as to percentile (if I'm wrong here, I'd appreciate a "heads
> > > > up")...
> > > >
> > > MK. Heads up!. These are aggregate student scores.
> >
> > You're going to have to clarify that. The totals of the numbers in
> > the percentiles are close are the same,...
> >
> MK. ??? Typo. Does not compute.
> >
> > ...yet the populations of the
> > states are very different. How it this "aggregate" defined?
> >

> These are student percentile scores: The score of the 10th percentile
> students, 25th percentile students, 50th percentile students, etc., in
> the State.

Ok, I missed the forest for the trees the first time around; I never
noticed that the numbers rose in each row; I was too busy getting them
lined up. Averages then? Ok, I'll consider the implications of that
information. Off the top of my head it occurs to me that the
percentage of minorities and inner-city dwellers in the state is going
to be telling factor. If this is the case (as I suspect it may be),
then I'd expect to see low scores in such states regardless of
district size or entering age.

> > > I will address only your last point: Whites are a minority in
> > > Washington DC, yet the DC White mean score is the highest in
> > > the nation. Whites are a minority in Hawaii.
> >
> > The implication intended is that all WHITES are the same everywhere?
> > That must mean all BLACKS are the same as well <g>?
> >
> MK. Get serious, or leave.

I'm deadly serious; your assuption above is that there are no
demographic differences in white populations from state to state, nor
in black, nor asian, etc. IS that your view, since you are using it
to try to support your argument? Surely you are well-aware that the
white population of DC is largely college-educated, professional, etc?
You wouldn't have been trying to slip a change-up by me?

> > > Hawaiian/part Hawaiian are the largest ethnic group in Hawaii's
> > > State schools. The White mean score (Numbers and Operations, parents
> > > are high-school educated) is the lowest in the nation. The Black
> > > mean score is the lowest in the nation.
> >
> > So let's see now, what can we conclude from this single data point?
> > That Hawaii schools are poor?...
> >
> MK. "Wretched", I would say.
> >
> >...That kids don't attend school because the weather is mostly good?...
> >
> MK. The climate's nice in Singapore, also. Cutting a terrible school
> probably boosts NAEP scores.
> >
> MK. The rest is even more frivolous. Inane sarcasm deleted...

Aw gee, and I especially liked the "falling coconut" correlation..
Aren't you the one who believes correlation implies causality? The
logical extension of your statement as to the effect of starting age
on math results would be that having "kids" enter school at 50 or 60
is preferable to having them enter at 5 or 6. Personally, I think
THAT'S pretty inane, but I hadn't said so until now.

> MK. Is there a serious critic out there?

You don't care to have your analysis technique used for satirical
effect? Perhaps if it were not flawed to begin with, the potential
for satire wouldn't exist. If you approached your data with even a
modicum of objectivity, and were even SLIGHTLY honest in your
appraisal of the meaning of the data, I wouldn't have to point out the
logical acrobatics you have to undergo to get the conclusions you
want; acrobatics that, used by me, result in "inane sarcasm". The
inanity comes from the flawed reasoning and analysis techniques.

panther

unread,
Oct 24, 1999, 3:00:00 AM10/24/99
to
In article <3816cfb5....@newsreader.cais.net>,
loj...@lojban.org (Bob LeChevalier) wrote:

> panther <igon...@my-deja.com> wrote:
> >MK. Here Ron deletes exceptions which dispute the "urban density" as
> >opposed to "school district size' explanation for poor NAEP test
> >performance. I expect New Jersey and Florida provide illuminating
> >contrasts, also. Florida commits 80% of its students to large
districts.
>
> No, Florida has legal district boundaries that people have settled in
to
> create large districts. Probably most of the lines were drawn when
those
> districts were much smaller than they are now (Miami had a population
under
> 2,000 in 1900;
>
MK. So why: "No"? Utah also commits a large fraction of total State
enrollment to large districts, which were laid out before the
districysreceived today's population. The point is not how those
districts came to be, but how large they are.

>
> >Newark's population has grown 5% since then and Jersey City
> 10% (and both are smaller than they were a couple decades ago).
>
> > New Jersey, almost none (Newark and Jersey City.)

>
> Whereas New Jersey has no center cities, only suburbs.
>
> >The average school
> >district in New Jersey is under 6,000 students, if I recall
correctly.
> >Dade county, Fla., and Broward county, Fla., are among the nation's
top
> >ten largest school districts.
>
> Florida has 67 school districts and 67 counties; the counties were
likely
> set up long before anyone knew that some of them would come to house a
> hundred thousand students. In short, they have chosen to reduce
> bureaucracy by having school districts match up with the other legal
> division of the state. Makes sense to me.
>
MK. It makes sense to a lot of people, but it doesn't work out. With
the command of aggregated resources, the bureaucracy is better able to
command political influence, and pad the administration with useless
warm bodies.

>
> New Jersey is the inexplicable one, having hundreds of districts
while only
> having 19 counties. Think of all that bureaucracy, all those
redundant
> multi-layered tax-collecting authorities.
>
MK. As I said, it doesn't work out that aggregation is more efficient,
that large districts employ fewer bureacrats per 100 students.

>
>
> >> > Except for children of White, college-educated parents, whose
scores
> >> > rise as districts increase in size.
> >>
> >> Which strongly suggests a racial (and perhaps dwelling location)
> >> factor is being evaluated here and not a beginning age or district
> >> size factor.
> >>
> >MK. It strongly suggests that district size and age are not the only
> >factors, but who ever said they were?
>
> You have made claims that "schools are lethal institutions", and used
> language such as "poison" within the last couple of days. This
indicates a
> claim that the district size and age are the EFFECTIVE cause "damaging
> kids". You make this clear in your "Proposal":
> >6. States with large school districts tend (strongly) to have low
> >scores and a large gap between the top and bottom scores (Hawaii is
no
> >exception.)
> >7. States with large school districts tend to have high rates of
> >juvenile crime (Hawaii and Utah are exceptions.)
> >8. Politically powerless minorities are poorly served by large
school
> >districts (Hawaii is not an exception.).
>
> You use the word "tend" which by my dictionary implies a direct and/or
> innate result. If you admit other factors, then you should not be
making
> claims of "tendency" as opposed to mere "correlation".
>
MK. We differ. Firing at random into a crowd "tends" to kill people.
There are other factors, but it's still murder.
>
MK. Discussion deleted...

>
Take care. Homeschool if you can.

Bob LeChevalier

unread,
Oct 24, 1999, 3:00:00 AM10/24/99
to
rmc...@nospam.banet.net (Ron McDermott) wrote:
>> I expect New Jersey and Florida provide illuminating contrasts, also.
>
>Which we cannot establish since New Jersey is among the missing
>states..
>
>> Florida commits 80% of its students to large districts.
>
>And has a huge latino population as well as some other minority
>components, again clustered in the cities.
>
>> New Jersey, almost none (Newark and Jersey City).
>
>And we know that the cities perform poorly (at least I know since New
>Jersey is "next door" to me).

Indeed, one obvious flaw with MK's conclusions is that where you have
cities with slums, they will (by many earlier analyses) do worse than the
surrounding suburbs. NJ has vast areas of considerable wealth that are NOT
slums, so it is reasonable to expect those wealthy to insist on and get
good schools. Florida has a smaller per capita income, and we know that a
significant portion of its wealthier people are retirees who don't have
much concern with school quality. (Indeed this is a problem in many
sunbelt states; I have little doubt that Hawaii has the same problem to
some extent.)

>> > > I will address only your last point: Whites are a minority in
>> > > Washington DC, yet the DC White mean score is the highest in
>> > > the nation. Whites are a minority in Hawaii.
>> >
>> > The implication intended is that all WHITES are the same everywhere?
>> > That must mean all BLACKS are the same as well <g>?
>> >
>> MK. Get serious, or leave.
>
>I'm deadly serious; your assuption above is that there are no
>demographic differences in white populations from state to state, nor
>in black, nor asian, etc. IS that your view, since you are using it
>to try to support your argument? Surely you are well-aware that the
>white population of DC is largely college-educated, professional, etc?
> You wouldn't have been trying to slip a change-up by me?
>
>> > > Hawaiian/part Hawaiian are the largest ethnic group in Hawaii's
>> > > State schools. The White mean score (Numbers and Operations, parents
>> > > are high-school educated) is the lowest in the nation. The Black
>> > > mean score is the lowest in the nation.

Furthermore this analysis seems to assume that racial groups inherent have
common interests, and indeed, based on MK's conclusion (not quoted), that
people can be grouped as to political potency based merely on their race.

>> MK. The climate's nice in Singapore, also. Cutting a terrible school
>> probably boosts NAEP scores.
>> >
>> MK. The rest is even more frivolous. Inane sarcasm deleted...
>
>Aw gee, and I especially liked the "falling coconut" correlation..
>Aren't you the one who believes correlation implies causality? The
>logical extension of your statement as to the effect of starting age
>on math results would be that having "kids" enter school at 50 or 60
>is preferable to having them enter at 5 or 6. Personally, I think
>THAT'S pretty inane, but I hadn't said so until now.

I'm working on proof, but it seems quite clear that there is no significant
attendance in kindergarten in states that have differing ages of mandatory
education.

panther

unread,
Oct 24, 1999, 3:00:00 AM10/24/99
to
In article <87XdtVqF0GQ7-pn2-UEc6fYSm4qzF@localhost>,
rmc...@nospam.banet.net (Ron McDermott) wrote:
> panther wrote:
>
MK. Discussion deleted...
>
MK (Scores fall as districts increase in size. Scores fall as States
compel attendance earlier).

>
> > > > MK. The effect remains when measured within racial groups.
>
> To be continued after I look at THAT data.
>
MK. OK.

>
> > > > Except for children of White, college-educated parents, whose
scores
> > > > rise as districts increase in size.
> > >
> > > Which strongly suggests a racial (and perhaps dwelling location)
> > > factor is being evaluated here and not a beginning age or district
> > > size factor.
> > >
> > MK. It strongly suggests that district size and age are not the only
> > factors, but who ever said they were?
>
> You did when you claimed the data "proved" that early attendence and
> large districts impede learning. It, of course, does no such thing,
> and the fact that the pattern does not hold for whites, who are still
> the majority of the population, makes it pretty clear that what is
> showing up is something else, and that it is racially related.
>
MK. (a) You put "proved" within quotation marks. Author = igonaked@my-
deja.com; date = ___??? Is this "deliberately deceptive" (i.e., a lie)?
(b) The pattern -does- hold, across the US, for Whites, overall. It
does not hold for children of White, college-educated parents. Even
these, when a few States are subtracted, parallel the trend (not as
strongly). My model is that the politically adept parents shield their
offspring fpreom the depradations of the State school bureaucracy.
Hawaii is confirmation, since whites do not do so well here, and whites
are not the politically adept group. The high-school educated Black
parents in Hawaii are probably mostly military, therefore full-time
employed, and above average even for that population (they did well
enough on the military's BTB). So the fact that their mean score is the
lowest in the nation for that group (h.s.-educated Black) is telling.
Military Blacks in Hawaii are thrice unfortunate, wrt political
influence: There are few Blacks, military personnel usually vote
elsewhere, and politicians and the bureaucracy know they'll rotate out
of here in two or three years.
>
MK. Discussion deleted...

>
> > These are student percentile scores: The score of the 10th
percentile
> > students, 25th percentile students, 50th percentile students, etc.,
in
> > the State.
>
> Ok, I missed the forest for the trees the first time around; I never
> noticed that the numbers rose in each row; I was too busy getting them
> lined up. Averages then?...
>
MK. NO! The scores you inspected were student percentile scores.
>
>... Ok, I'll consider the implications of that

> information. Off the top of my head it occurs to me that the
> percentage of minorities and inner-city dwellers in the state is going
> to be telling factor. If this is the case (as I suspect it may be),
> then I'd expect to see low scores in such states regardless of
> district size or entering age.
>
> > > > I will address only your last point: Whites are a minority in
> > > > Washington DC, yet the DC White mean score is the highest in
> > > > the nation. Whites are a minority in Hawaii.
> > >
> > > The implication intended is that all WHITES are the same
everywhere?
> > > That must mean all BLACKS are the same as well <g>?
> > >
> > MK. Get serious, or leave.
>
> I'm deadly serious; your assuption above is that there are no
> demographic differences in white populations from state to state, nor
> in black, nor asian, etc. IS that your view, since you are using it
> to try to support your argument? Surely you are well-aware that the
> white population of DC is largely college-educated, professional, etc?
> You wouldn't have been trying to slip a change-up by me?
>
MK. I do assume that the distribution of genetic endowment is the same,
nationwide, within racial groups. I do not assume that "culture"
(whatever that means) is the same, nationwide. Both the mean score of
children of White, college-educated parents AND the mean score of
children of White, high-school educated parents in DC are the highest
in the nation.
>
MK. Discussion deleted...

>
> Aw gee, and I especially liked the "falling coconut" correlation..
> Aren't you the one who believes correlation implies causality? The
> logical extension of your statement as to the effect of starting age
> on math results would be that having "kids" enter school at 50 or 60
> is preferable to having them enter at 5 or 6...
>
MK. Why: "logical"? We only have data within a narrow range of starting
ages. It would be my preference to repeal compulsory attendance laws
entirely, but that's another argument.

>
> THAT'S pretty inane, but I hadn't said so until now.
>
MY. Your "logical" extension is silly.

>
> > MK. Is there a serious critic out there?
>
> You don't care to have your analysis technique used for satirical
> effect?
>
MK. You haven't done that. I do not see any tabular data with model and
proposal, from your side.
>
MK. Discussion deleted...

>
Take care. Homeschool if you can.

panther

unread,
Oct 24, 1999, 3:00:00 AM10/24/99
to
In article <38142312....@newsreader.cais.net>,
loj...@lojban.org (Bob LeChevalier) wrote:
> Ron McDermott wrote:
>
> >> (MK.) I expect New Jersey and Florida provide illuminating

contrasts, also.
> >
> >Which we cannot establish since New Jersey is among the missing
> >states..
> >
> >> Florida commits 80% of its students to large districts.
> >
> >And has a huge latino population as well as some other minority
> >components, again clustered in the cities.
> >
> >> New Jersey, almost none (Newark and Jersey City).
> >
> >And we know that the cities perform poorly (at least I know since
New
> >Jersey is "next door" to me).
>
> Indeed, one obvious flaw with MK's conclusions is that where you have
> cities with slums, they will (by many earlier analyses) do worse than
the
> surrounding suburbs. NJ has vast areas of considerable wealth that
are NOT
> slums, so it is reasonable to expect those wealthy to insist on and
get
> good schools....
>
MK. This is in no way a "flaw in with MK's conclusions". It is entirely
consistent with the model. The school districts of Newark and Jersey
City get more money than the average for the State. Poor minorities are
less adept at political action. Political control of schools does the
most harm to the children of the least politically adept parents. The
damage done to children is directly related to the enrollment (and so
to aggregate budget, and so to the incentive for non-parent groups to
take school policy away from parents).

>
Florida has a smaller per capita income, and we know that a
> significant portion of its wealthier people are retirees who don't
have
> much concern with school quality. (Indeed this is a problem in many
> sunbelt states; I have little doubt that Hawaii has the same problem
to
> some extent.)
>
MK. Redicing the number of allies parents might have in their contest
with the representatives of the NEA/AFT/AFSCME cartel.
>
MK. Discussion deleted...

>
> I'm working on proof, but it seems quite clear that there is no
significant
> attendance in kindergarten in states that have differing ages of
mandatory
> education.
>
MK. (a) What would that prove? There are very few States that compel
attendance at 5 (Virginia, Arkansas, Delaware, ???). Of the States that
participate in NAEP, more States compel attendance at 7 than at 6. The
extra year of parent control enhances 4th and 8th grade performance.
(b) It will be interesting to see what you call a "proof", in light of
the objections you expressed about not considering all possible factors
that might influence the result. If you hold yourself to the same
standard you hold me, I expect you'll get back to us in about thirty
years.

>
Take care. Homeschool if you can.

Ron McDermott

unread,
Oct 25, 1999, 3:00:00 AM10/25/99
to
On Sun, 24 Oct 1999 09:56:40, loj...@lojban.org (Bob LeChevalier)
wrote:

> New Jersey is the inexplicable one, having hundreds of districts while only
> having 19 counties. Think of all that bureaucracy, all those redundant
> multi-layered tax-collecting authorities.

Yes, but think of a bedroom community of professional people forming
in New Jersey. They look around and see stuff they don't like in the
local schools, so decide to set up their own district. Perhaps to
make it lily-white, perhaps to make it high-powered, etc. Not too
dissimilar to what happened in my area a couple decades ago when IBM
was growing; communities grew overnight, and those communities didn't
like the idea of sending their kids to schools with large black
populations (for a variety of reasons; I don't mean to imply that this
was STRICTLY racially motivated; simply racially related), so we ended
up with a couple brand new districts. The new districts were almost
entirely white, had the latest "stuff", and had a high-powered
educational program. As the influx of IBM employees aged, the
enrollments dropped, IBM instituted cutbacks, and these districts are
now just barely hanging on.

Bob LeChevalier

unread,
Oct 25, 1999, 3:00:00 AM10/25/99
to
panther <igon...@my-deja.com> wrote:
>In article <3816cfb5....@newsreader.cais.net>,
> loj...@lojban.org (Bob LeChevalier) wrote:

>> panther <igon...@my-deja.com> wrote:
>> >MK. Here Ron deletes exceptions which dispute the "urban density" as
>> >opposed to "school district size' explanation for poor NAEP test
>> >performance. I expect New Jersey and Florida provide illuminating
>> >contrasts, also. Florida commits 80% of its students to large
>districts.
>>
>> No, Florida has legal district boundaries that people have settled in
>to
>> create large districts. Probably most of the lines were drawn when
>those
>> districts were much smaller than they are now (Miami had a population
>under
>> 2,000 in 1900;
>>
>MK. So why: "No"? Utah also commits a large fraction of total State
>enrollment to large districts, which were laid out before the
>districysreceived today's population. The point is not how those
>districts came to be, but how large they are.

The objection is to the word "commit". The school dsitricts were set up on
a basis of counties. People came later. No one "committed them" to large
districts. They were born there or moved there after the boundaries were
set up. Hence it was a parental and not a state decision to "commit" to
the district by moving there. Just as people who immigrate to America
commit themselves to our form of government, which was here before they
arrived; We/government don't commit them to one of the 50 states - they
choose to live in one.

>> >Newark's population has grown 5% since then and Jersey City
>> 10% (and both are smaller than they were a couple decades ago).
>>
>> > New Jersey, almost none (Newark and Jersey City.)
>>
>> Whereas New Jersey has no center cities, only suburbs.
>>

>> >The average school
>> >district in New Jersey is under 6,000 students, if I recall
>correctly.
>> >Dade county, Fla., and Broward county, Fla., are among the nation's
>top
>> >ten largest school districts.
>>

>> Florida has 67 school districts and 67 counties; the counties were
>likely
>> set up long before anyone knew that some of them would come to house a
>> hundred thousand students. In short, they have chosen to reduce
>> bureaucracy by having school districts match up with the other legal
>> division of the state. Makes sense to me.
>>
>MK. It makes sense to a lot of people, but it doesn't work out. With
>the command of aggregated resources, the bureaucracy is better able to
>command political influence, and pad the administration with useless
>warm bodies.

Proof by assertion.

>> New Jersey is the inexplicable one, having hundreds of districts while only
>> having 19 counties. Think of all that bureaucracy, all those redundant
>> multi-layered tax-collecting authorities.
>>

>MK. As I said, it doesn't work out that aggregation is more efficient,
>that large districts employ fewer bureacrats per 100 students.

I thought we were talking about "choice", not "efficiency" (a concept that
is subjective when you start including quality in the measurement). If I
am moving to a state that has 800+ school district boundaries, none of
which have anything to do with the county and city lines, it is quite
unlikely that I will be able to make an informed choice among them.

>> >> > Except for children of White, college-educated parents, whose scores
>> >> > rise as districts increase in size.
>> >>
>> >> Which strongly suggests a racial (and perhaps dwelling location)
>> >> factor is being evaluated here and not a beginning age or district
>> >> size factor.
>> >>
>> >MK. It strongly suggests that district size and age are not the only
>> >factors, but who ever said they were?

If you admit that there are multiple factors, then you have to show how
those factors interact. You might find that one of the variables is highly
dependent on the other and hence both are not justifiably linked.

>> You have made claims that "schools are lethal institutions", and used
>> language such as "poison" within the last couple of days. This indicates a
>> claim that the district size and age are the EFFECTIVE cause "damaging
>> kids". You make this clear in your "Proposal":
>> >6. States with large school districts tend (strongly) to have low
>> >scores and a large gap between the top and bottom scores (Hawaii is >no
>> >exception.)
>> >7. States with large school districts tend to have high rates of
>> >juvenile crime (Hawaii and Utah are exceptions.)
>> >8. Politically powerless minorities are poorly served by large >school
>> >districts (Hawaii is not an exception.).
>>
>> You use the word "tend" which by my dictionary implies a direct and/or
>> innate result. If you admit other factors, then you should not be making
>> claims of "tendency" as opposed to mere "correlation".
>>
>MK. We differ. Firing at random into a crowd "tends" to kill people.
>There are other factors, but it's still murder.

Firing into a crowd does indeed "tend". But merely given that there are
people who fire into crowds, are you going to claim that joining into a
crowd "tends" to get one killed? Especially since there are also people
who fire at individuals?

You are claiming causality, while not showing it, and while admitting above
that there are other factors - factors that might be found causal to large
districts as well as to school starting age and test scores, in which case
all of these are results and not causes.

Bob LeChevalier

unread,
Oct 25, 1999, 3:00:00 AM10/25/99
to
panther <igon...@my-deja.com> wrote:
>> Indeed, one obvious flaw with MK's conclusions is that where you have
>> cities with slums, they will (by many earlier analyses) do worse than the
>> surrounding suburbs. NJ has vast areas of considerable wealth that are NOT
>> slums, so it is reasonable to expect those wealthy to insist on and get
>> good schools....
>>
>MK. This is in no way a "flaw in with MK's conclusions". It is entirely
>consistent with the model. The school districts of Newark and Jersey
>City get more money than the average for the State. Poor minorities are
>less adept at political action.

Sounds contradictory to your claim, if a school district with lots of poor
minorities gets more money than a district with fewer. Government money
tends to be a direct result of political action.

>Political control of schools does the
>most harm to the children of the least politically adept parents.

That is a causal claim. You have proven no causality, and have admitted
that there are multiple factors correlated with school performance, some of
which you have not identified. They are not ALL causes.

> The
>damage done to children is directly related to the enrollment (and so
>to aggregate budget,

You have not proven any DIRECT relation. There might be an indirect
relation (assuming we could agree on the phrase "damage done to children").

>and so to the incentive for non-parent groups to
>take school policy away from parents).

Non-parent groups have every incentive to take school policy away. It is
THEIR tax money too.

>Florida has a smaller per capita income, and we know that a
>> significant portion of its wealthier people are retirees who don't have
>> much concern with school quality. (Indeed this is a problem in many
>> sunbelt states; I have little doubt that Hawaii has the same problem to
>> some extent.)
>>
>MK. Redicing the number of allies parents might have in their contest
>with the representatives of the NEA/AFT/AFSCME cartel.

That assumes a contest. I, as a parent, am having no contest with a
figment of your imagination. (And it could well be that Florida is a state
without teachers' unions, like Virginia).

>> I'm working on proof, but it seems quite clear that there is no significant variation in


>> attendance in kindergarten in states that have differing ages of mandatory
>> education.

I left out a couple of words above.

>MK. (a) What would that prove? There are very few States that compel
>attendance at 5 (Virginia, Arkansas, Delaware, ???). Of the States that
>participate in NAEP, more States compel attendance at 7 than at 6.

If kindergarten attendance is the same in schools that compel attendance at
5, at 6 and at 7, then there is no reason to believe that the difference in
compulsory education ages has significant difference on parental or student
behavior. The kids are going to school whether the law makes their parents
send them or not. (All districts offer K thru 12, whether or not their
states compel it, so far as I know).

>The extra year of parent control enhances 4th and 8th grade performance.

My point would be that the supposed extra year of parental control has no
impact on parental behavior, because parents who can choose not to send
their kids to kindergarten are still sending their kids. (They may think
it helps their kids, or they may feel compelled by one of those other
forces in life - social conformity - which government has no monopoly on).

If there is no change in parental behavior as a linked to the differing
legal situation, a claim that differing legal situations change school
performance is nonsense.

>(b) It will be interesting to see what you call a "proof", in light of
>the objections you expressed about not considering all possible factors
>that might influence the result. If you hold yourself to the same
>standard you hold me, I expect you'll get back to us in about thirty
>years.

All I have to prove is that your supposed "cause" has no "effect" on
behavior, and therefore cannot be the cause of some result further down the
chain.

You love analogies. If states with high murder rates enact gun control
laws that are not successfully enforced (i.e. result in no behavior
change), YOU will find a correlation between gun control laws and murder
rate (and perhaps claim that "gun laws kill people"). That the gun control
laws were instituted to reduce the murder rate raised by other causes is
information missing from the statistics, and not detectable by your
inadequate statistical approach.

What I am showing is roughly equivalent to showing that the gun laws did
not affect gun ownership, which removes the possibility that the gun laws
are the cause of the murder rate.

Bob LeChevalier

unread,
Oct 25, 1999, 3:00:00 AM10/25/99
to
panther <igon...@my-deja.com> wrote:
>In article <87XdtVqF0GQ7-pn2-UEc6fYSm4qzF@localhost>,
> rmc...@nospam.banet.net (Ron McDermott) wrote:
>> panther wrote:
>> > > > Except for children of White, college-educated parents, whose scores
>> > > > rise as districts increase in size.
>> > >
>> > > Which strongly suggests a racial (and perhaps dwelling location)
>> > > factor is being evaluated here and not a beginning age or district
>> > > size factor.
>> > >
>> > MK. It strongly suggests that district size and age are not the only
>> > factors, but who ever said they were?
>>
>> You did when you claimed the data "proved" that early attendence and
>> large districts impede learning. It, of course, does no such thing,
>> and the fact that the pattern does not hold for whites, who are still
>> the majority of the population, makes it pretty clear that what is
>> showing up is something else, and that it is racially related.
>>
>MK. (a) You put "proved" within quotation marks. Author = igonaked@my-
>deja.com; date = ___??? Is this "deliberately deceptive" (i.e., a lie)?

These two lines are incoherent. You appear to be asking questions, but I
have no idea what those questions are.

>(b) The pattern -does- hold, across the US, for Whites, overall. It
>does not hold for children of White, college-educated parents. Even
>these, when a few States are subtracted, parallel the trend (not as
>strongly).

In other words, if you fudge the data enough by selectively omitting points
that you don't like, you can force your conclusion, sort of.

>My model is that the politically adept parents shield their
>offspring fpreom the depradations of the State school bureaucracy.

If you have a model, a scientist should look for tests that might disprove
the model. I have not seen you include those tests.

>Hawaii is confirmation, since whites do not do so well here, and whites
>are not the politically adept group.

a) this assumes that one can stereotype races as a "group" for purposes of
political action when neither whites nor blacks nor any other race
necessarily has common political goals in any state.

b) this assumes that as a race, whites are somehow politically unadept
merely because they have moved to Hawaii. I wonder what happens when a
politically adept white moves from New Jersey to Hawaii. Will he suddenly
lose political adeptness?

>The high-school educated Black
>parents in Hawaii are probably mostly military, therefore full-time
>employed, and above average even for that population (they did well
>enough on the military's BTB). So the fact that their mean score is the
>lowest in the nation for that group (h.s.-educated Black) is telling.

Oh, this is a funny new twist you throw in. Starting from an unsupported
"probably", you reach a conclusion without looking for comparable data
elsewhere in the nation. How do blacks elsewhere in the nation do on the
BTB (whatever that is)? How do you know that black military in Hawaii are
"above average for that population" (what population is "that" one, and how
did you determine the average?)? And how well do children of the military
do elsewhere in the country (it would seem to me that mobility might be a
factor reducing scores more than the quality of the school system, given
parents without a lot of extra resources)?

>Military Blacks in Hawaii are thrice unfortunate, wrt political
>influence: There are few Blacks, military personnel usually vote
>elsewhere, and politicians and the bureaucracy know they'll rotate out
>of here in two or three years.

And 4th, they might not have any interest in voting as a political bloc
based on your racist stereotype, a stereotype that implies that they SHOULD
form such a bloc. Witrh racists like you in the state that damn them for
not fighting for their racial interests, they haven't a chance.

If you haven't gotten the point, you have injected racial politics into the
issue and you can expect that stereotyping people based on race as to their
political beliefs or clout will draw heavy fire.

>> > > > I will address only your last point: Whites are a minority in
>> > > > Washington DC, yet the DC White mean score is the highest in
>> > > > the nation. Whites are a minority in Hawaii.
>> > >
>> > > The implication intended is that all WHITES are the same everywhere?
>> > > That must mean all BLACKS are the same as well <g>?
>> > >
>> > MK. Get serious, or leave.
>>
>> I'm deadly serious; your assuption above is that there are no
>> demographic differences in white populations from state to state, nor
>> in black, nor asian, etc. IS that your view, since you are using it
>> to try to support your argument? Surely you are well-aware that the
>> white population of DC is largely college-educated, professional, etc?
>> You wouldn't have been trying to slip a change-up by me?

Furthermore, by what he has written above, he believes that the black
population of Hawaii is demographically quite different from the rest of
the country. Presumably he therefore knows that the white population is
also likely to be demographically different. (I hadn't thought to look for
correlations of test scores by percentage of the population that is in the
military. That might be at least as causal as his large district
correlation, since he says that the military is not politically adept -
which itself is news to me).

>MK. I do assume that the distribution of genetic endowment is the same,
>nationwide, within racial groups.

Why would you ASSume that? Alan L. has been posting about the researchers
who believe that they have proven otherwise.

>I do not assume that "culture"
>(whatever that means) is the same, nationwide.

But you assume that "black" is the same, nationwide. Likewise "white". I
suspect that in Hawaii, people classify themselves differently because of
differences in the politics and culture.

>Both the mean score of
>children of White, college-educated parents AND the mean score of
>children of White, high-school educated parents in DC are the highest
>in the nation.

By the time you have broken things down that fine, you have probably
reached the point where the "mean" score has no statistical validity. How
many children of white, high-school-educated parents in DC do you think
took the test?

I looked in the 1996 Math summary report, that does not break things down
that far. It appears that aroun 1700 8th graders took the test in DC, of
which 6% were white - maybe 100. A different table says that 19% of DC
test takers had no more than high school education, but you should think
this is probably very high for whites given your assumption above that
whites in DC are mostly college-educated professionals. So your
conclusions about white high school parent students in DC is probably based
on a sample size of under 10 kids.

It should come as no surprise to you that the NAEP does not consider such a
small sample size to be adequate to reach any conclusions whatsoever.
Indeed in Appendix A to the summary report they discuss the statistical
issues at length. It becomes clear on even a brief skimming why they don't
release raw data to people like you who have no sensitivity to the proper
treatment and use of statistical data. The cutoff to report average scores
at all seems to be somewhere between 2% and 3% of the student population.
So, unless you have clear reason why I am wrong, however you got an mean
score for DC high-school-parented kids, your number is not statistically
valid.

Of course this won't stop you from trying to overturn the system based on
your invalid statistics.

>> Aw gee, and I especially liked the "falling coconut" correlation..
>> Aren't you the one who believes correlation implies causality? The
>> logical extension of your statement as to the effect of starting age
>> on math results would be that having "kids" enter school at 50 or 60
>> is preferable to having them enter at 5 or 6...
>>
>MK. Why: "logical"? We only have data within a narrow range of starting
>ages.

You don't seem to have any compulsion against extrapolating outside the
range of your available data.

>It would be my preference to repeal compulsory attendance laws
>entirely, but that's another argument.
>>
>> THAT'S pretty inane, but I hadn't said so until now.
>>
>MY. Your "logical" extension is silly.

So is your entire statistical edifice on which the argument is based. That
is our point. Your model sounds as ridiculous to us as this one sounds to
you.

>> You don't care to have your analysis technique used for satirical
>> effect?
>>
>MK. You haven't done that. I do not see any tabular data with model and
>proposal, from your side.

Maybe we don;t think that such tabular data can prove anything useful and
would not deign to prepare a proposal based on faulty statsitics.

panther

unread,
Oct 25, 1999, 3:00:00 AM10/25/99
to
In article <38197d86....@newsreader.cais.net>,
loj...@lojban.org (Bob LeChevalier) wrote:
> panther wrote:
>
MK. One post per day to this thread from MK, henceforth, since it's
getting tangled. I'll try to read all posts, and address the points I
remember. The ad hominem arguments of only the post to which I directly
respond will receive a response.
>
MK. Discussion deleted...
>
MK. First, Ron's post about IBM. Worth consideration. I have no
argument with it.

>
> >> > MK. It strongly suggests that district size and age are not the
only
> >> > factors, but who ever said they were?
> >>
> >> You did when you claimed the data "proved" that early attendence
and
> >> large districts impede learning. It, of course, does no such
thing,
> >> and the fact that the pattern does not hold for whites, who are
still
> >> the majority of the population, makes it pretty clear that what is
> >> showing up is something else, and that it is racially related.
> >>
> >MK. (a) You put "proved" within quotation marks. Author =
igonaked@my-
> >deja.com; date = ___??? Is this "deliberately deceptive" (i.e., a
lie)?
>
> These two lines are incoherent. You appear to be asking questions,
but I
> have no idea what those questions are.
>
MK. Where (date, forum, thread) did MK. ("igon...@mydeja.com") claim a
"proof" of anything? is my question.

>
> >(b) The pattern -does- hold, across the US, for Whites, overall. It
> >does not hold for children of White, college-educated parents. Even
> >these, when a few States are subtracted, parallel the trend (not as
> >strongly).
>
> In other words, if you fudge the data enough by selectively omitting
points
> that you don't like, you can force your conclusion, sort of.
>
MK. You-all were happy to endorse deleting one or two States when Ron
claimed (falsely) that it reversed the sign of the correlations for
(test score, district size). Odd cases are very instructive: HIV
positive survivors, good minority urban schools, States with large
districts and high test performance (Utah), polite defenders of the
NEA/AFT/AFSCME cartel. Well, it -would- be interesting to meet one of
these, anyway.

>
> >My model is that the politically adept parents shield their
> >offspring from the depradations of the State school bureaucracy.
>
MK. Bob's comment deleted...

>
> >Hawaii is confirmation, since whites do not do so well here, and
whites
> >are not the politically adept group.
>
> a) this assumes that one can stereotype races as a "group" for
purposes of
> political action when neither whites nor blacks nor any other race
> necessarily has common political goals in any state.
>
MK. People vote along racial lines often enough for politicians to
compose strategy on that basis.

>
> b) this assumes that as a race, whites are somehow politically unadept
> merely because they have moved to Hawaii. I wonder what happens when
a
> politically adept white moves from New Jersey to Hawaii. Will he
suddenly
> lose political adeptness?
>
MK. Of course. A lot depends on conections, and the possibility that
one would find allies in a dispute.

>
> >The high-school educated Black
> >parents in Hawaii are probably mostly military, therefore full-time
> >employed, and above average even for that population (they did well
> >enough on the military's BTB). So the fact that their mean score is
the
> >lowest in the nation for that group (h.s.-educated Black) is telling.
>
> Oh, this is a funny new twist you throw in. Starting from an
> unsupported "probably",...
>
MK. The Pacific Ocean is probably salty today. I don't feel the need to
check. If you lived here, neither would you feel the need to check
either assertion.
>
> ...you reach a conclusion without looking for comparable data

> elsewhere in the nation. How do blacks elsewhere in the nation do on
the
> BTB (whatever that is)?...
>
MK. BTB = Basic Test Battery, given to all applicants for enlistment.
An aptitude test.
>
> ... How do you know that black military in Hawaii are

> "above average for that population" (what population is "that" one,
and how
> did you determine the average?)?
>
MK. The point was: There are few Blacks (whatever level of education)
in Hawaii. Those who live here are disproportionately military,
therefore employed, therefore, I expect, their children would test -
above- the average (Black, high-school educated parent). They don't.
The Hawaii mean score (NAEP 8th grade Numbers and Operations) is the
lowest in the nation. I'm saying, it's not the students, it's the
wretched school system. Military parents resist assignment to Hawaii,
expressly because our schools are notoriously wretched.
>
> ... And how well do children of the military

> do elsewhere in the country (it would seem to me that mobility might
be a
> factor reducing scores more than the quality of the school system,
given
> parents without a lot of extra resources)?
>
MK. Interesting question. I have no idea where to get those data. NAEP
does give scores for DOD schools.

>
> >Military Blacks in Hawaii are thrice unfortunate, wrt political
> >influence: There are few Blacks, military personnel usually vote
> >elsewhere, and politicians and the bureaucracy know they'll rotate
out
> >of here in two or three years.
>
> And 4th, they might not have any interest in voting as a political
bloc
> based on your racist...
>
MK. You say the sweetest things. Give us a kiss.
>
> ...stereotype, a stereotype that implies that they SHOULD
> form such a bloc. With racists...
>
MK. Later in this post you defend Alan's claim of a genetic basis for
the difference in student test performance. Who's the racist? You are
confused.
>
> ...like you in the state that damn them for

> not fighting for their racial interests, they haven't a chance.
>
MK. Hardly. I observe that the difference between White mean scores and
Black mean scores is smaller in States with numerous small school
districts than in States which compel attendance in large school
districts. My model is that the parent interests oppose the
bureaucracy's interests diverge as districts increase in size. When
districts are large, the bureaucracy wins, except for the children of
the politically adept parents, who shield their children from the
vampires, with GT classes, magnet schools, and tutors to do after
school what the schools were paid to do.

>
> If you haven't gotten the point, you have injected racial politics
into the
> issue and you can expect that stereotyping people based on race as to
their
> political beliefs or clout will draw heavy fire.
>
MK. Actualy, Ron injected racial politics into it, with his insistence
that the "large district, low score" relation meant "urban-minority,
low score". I just attempt to answer that criticism, by looking at
separate data for each race.
>
MK. Discussion deleted...
>
>...> MK. Get serious, or leave.

>
> >> I'm deadly serious; your assuption above is that there are no
> >> demographic differences in white populations from state to state,
nor
> >> in black, nor asian, etc. IS that your view, since you are using
it
> >> to try to support your argument? Surely you are well-aware that
the
> >> white population of DC is largely college-educated, professional,
etc?
> >> You wouldn't have been trying to slip a change-up by me?
>
> Furthermore, by what he has written above, he believes that the black
> population of Hawaii is demographically quite different from the rest
of
> the country.
>
MK. Yes. Black parents in Hawaii are more likely to be in the military,
to have completed high school, to be full-time employed, than are
Blacks elsewhere.
>
>...Presumably he therefore knows that the white population is

> also likely to be demographically different.
>
MK. That doesn't follow at all.
>
>...(I hadn't thought to look for correlations of test scores by

> percentage of the population that is in the military. That might be
> at least as causal as his large district correlation, since he says
> that the military is not politically adept - which itself is news to
> me).
>
> >MK. I do assume that the distribution of genetic endowment is the
> > same, nationwide, within racial groups.
>
> Why would you ASSume that?...
>
MK. Vulgar child. Grow up.
>
> ...Alan L. has been posting about the researchers

> who believe that they have proven otherwise.
>
MK. And...? You want to defend that view? What we have here is a
reflexive critic: MK. is a racist if he asserts Blacks in Hawaii are
more likely than Blacks elsewhere to be enlisted military, and MK. is
wrong if he assumes that the genetic endowment is the same, within
racial groups, nationwide. Btw, that last is not what Alan addresses,
He defends the idea that there are significant differences -between-
racial groups.

>
> >I do not assume that "culture"
> >(whatever that means) is the same, nationwide.
>
> But you assume that "black" is the same, nationwide. Likewise
"white". I
> suspect that in Hawaii, people classify themselves differently
because of
> differences in the politics and culture.
>
MK. I assume that the genetic component of academic test performance
differences (between races) is insignificant for a number of reasons,
one of which is that we cannot manipulate that variable, so there's
nothing we can do about it anyway. We can manipulate institutional
variables.

>
> >Both the mean score of
> >children of White, college-educated parents AND the mean score of
> >children of White, high-school educated parents in DC are the highest
> >in the nation.
>
> By the time you have broken things down that fine, you have probably
> reached the point where the "mean" score has no statistical
> validity. How
> many children of white, high-school-educated parents in DC do you
think
> took the test?
>
MK. Bob has a reasonable point here. This is why I do not use
percentile scores by parent's race and level of education. The NCES
includes the note: "Interpret with caution", when sample sizes are
small.

>
> I looked in the 1996 Math summary report, that does not break things
down
> that far. It appears that around 1700 8th graders took the test in

DC, of
> which 6% were white - maybe 100. A different table says that 19% of
DC
> test takers had no more than high school education,...
>
MK. Are you saying that 81% of the all (Black and White, combined) DC
parents have -more- than a high-school education? I would be surprised.
>
> ...but you should think

> this is probably very high for whites given your assumption above that
> whites in DC are mostly college-educated professionals. So your
> conclusions about white high school parent students in DC is probably
based
> on a sample size of under 10 kids.
>
MK. Just a guess, here. Let's take that estimate of 100 White test
takers. I would assume that the level of education of White DC parents
IS NOT the same as the level of education of Black DC parents (Sound
reasonable to you?), so the distribution of education -overall- is
useless in estimating the distrbution of education for Whites DC
parents of children in the cartel's schools. I would guess it's more
like 50% college and 50% highschool. But that's just a guess. I use the
numbers that NCES supplies.

>
> It should come as no surprise to you that the NAEP does not consider
such a
> small sample size to be adequate to reach any conclusions whatsoever.
> Indeed in Appendix A to the summary report they discuss the
statistical
> issues at length. It becomes clear on even a brief skimming why they
don't
> release raw data to people like you who have no sensitivity to the
proper
> treatment and use of statistical data....
>
MK. Says the man who insisted that I apply correlation to cyclic
(seasonal) data. BTW, can you supply the reference, the name of the
statistics book with "cross-correlation" in the index, and which
discusses "standard outlier techniques". You pretend familiarity with
authorities you do not know.
>
>...The cutoff to report average scores

> at all seems to be somewhere between 2% and 3% of the student
population.
>
MK. It would depend on the absolute sample size, not the percentage. A
ten percent sample of a very small district might be insignificant. A 1
percent sample of the NYC school system would be very significant.

>
> So, unless you have clear reason why I am wrong, however you got an
mean
> score for DC high-school-parented kids, your number is not
statistically
> valid.
>
MK. My argument might be wrong, whether or not your argument above is
wrong. The mean scores by parents race and level of education come from
the NCES "1992 Data on Disk Almanac Viewer".

>
> Of course this won't stop you from trying to overturn the system
based on
> your invalid statistics.
>
MK. The possiblity that one piece of evidence is shaky does not
invalidate the entire argument, since the vast preponderance of
evidence moves in the same direction, and the argument is a sum, not a
sequence.
>
MK. Discussion deleted...

>
> >MY. Your "logical" extension is silly.
>
> So is your entire statistical edifice on which the argument is
> based. That is our point. Your model sounds as ridiculous to us as
> this one sounds to you.
>
MK. I can understand how someone who is "agnostic" on the subject of
Marxism might question an argument against the State monopoly school
system. Most of us didn't sleep through the twentieth century.

>
> >> You don't care to have your analysis technique used for satirical
> >> effect?
> >>
> >MK. You haven't done that. I do not see any tabular data with model
and
> >proposal, from your side.
>
> Maybe we don't think that such tabular data can prove anything useful
and
> would not deign to prepare a proposal based on faulty statsitics.
>
MK. How about some -good- statistics, then? Your proud indifference to
evidence is stunning.
>
MK. This book review comes from a site linked to
"www.schoolchoices.org"...
>
The Language Wars and Other Writings for Homeschoolers
Ruth Beechick (Pollock Pines, CA: Arrow Press, 1995) (ISBN 0-940319-09-
8). US$16.00, 252 pages. Collection of articles by Ruth Beechick first
published elsewhere. Favorite quotation: "A research idea I have never
seen carried out is to study whether this or any other 'scientific'
program can outdo the alternative of a loving mother and her child
together on a couch or at the kitchen table with a few good books and
some simple teaching materials." KMBseen_HCO

>
Take care. Homeschool if you can.
>
www.schoolchoices.org (Massive site. Useful links).
>
www.hslda.org (Very useful links, for prospective homeschoolers)
>
www.rru.com/~meo/hs.minski.html (One page. Marvin Minski comment on
school. Please read this.)
>

Sent via Deja.com http://www.deja.com/
Before you buy.

Ron McDermott

unread,
Oct 26, 1999, 3:00:00 AM10/26/99
to
On Sun, 24 Oct 1999 17:01:21, panther <igon...@my-deja.com> wrote:

> In article <87XdtVqF0GQ7-pn2-UEc6fYSm4qzF@localhost>,
> rmc...@nospam.banet.net (Ron McDermott) wrote:
> > panther wrote:
> >
> > > These are student percentile scores: The score of the 10th
> > > percentile students, 25th percentile students, 50th percentile
> > > students, etc., in the State.
> >
> > Ok, I missed the forest for the trees the first time around; I never
> > noticed that the numbers rose in each row; I was too busy getting
> > them lined up. Averages then?...
> >
> MK. NO! The scores you inspected were student percentile scores.

The percentiles are for the state itself, yes? Then the score listed
as 10th percentile represents a score which is better than 10 percent
of that state's test takers? But the test is national? So what is
the national score which represents these percentiles, I wonder?

> >... Ok, I'll consider the implications of that
> > information. Off the top of my head it occurs to me that the
> > percentage of minorities and inner-city dwellers in the state is going
> > to be telling factor. If this is the case (as I suspect it may be),
> > then I'd expect to see low scores in such states regardless of
> > district size or entering age.
> >
> > > > > I will address only your last point: Whites are a minority in
> > > > > Washington DC, yet the DC White mean score is the highest in
> > > > > the nation. Whites are a minority in Hawaii.
> > > >
> > > > The implication intended is that all WHITES are the same
> > > > everywhere? That must mean all BLACKS are the same
> > > > as well <g>?
> > > >
> > > MK. Get serious, or leave.
> >
> > I'm deadly serious; your assuption above is that there are no
> > demographic differences in white populations from state to state, nor
> > in black, nor asian, etc. IS that your view, since you are using it
> > to try to support your argument? Surely you are well-aware that the
> > white population of DC is largely college-educated, professional, etc?
> > You wouldn't have been trying to slip a change-up by me?
> >
> MK. I do assume that the distribution of genetic endowment is the same,
> nationwide, within racial groups.

I assume the same.

> I do not assume that "culture" (whatever that means) is the same,
> nationwide.

Ok; your statement suggested otherwise.

> Both the mean score of children of White, college-educated parents
> AND the mean score of children of White, high-school educated
> parents in DC are the highest in the nation.

And you account for this... how?

> MK. Discussion deleted...
> >
> > Aw gee, and I especially liked the "falling coconut" correlation..
> > Aren't you the one who believes correlation implies causality? The
> > logical extension of your statement as to the effect of starting age
> > on math results would be that having "kids" enter school at 50 or 60
> > is preferable to having them enter at 5 or 6...
> >
> MK. Why: "logical"? We only have data within a narrow range of starting
> ages. It would be my preference to repeal compulsory attendance laws
> entirely, but that's another argument.

If later is better, at what point does it cease to be?

> > THAT'S pretty inane, but I hadn't said so until now.
> >
> MY. Your "logical" extension is silly.

I consider your leaps of logic equally silly.

> > > MK. Is there a serious critic out there?
> >
> > You don't care to have your analysis technique used for satirical
> > effect?
> >
> MK. You haven't done that. I do not see any tabular data with model and
> proposal, from your side.

The data is immaterial, as I've said repeatedly; it's the conclusions
you draw from them which are flawed. Now I don't happen to have data
on the number of coconuts falling annually in Hawaii, nor the
pineapples grown, but I'm sure that one could get a correlation
suggesting a relationship between these things and the low math scores
in Hawaii, and I could then draw a conclusion (unjustified) on that
basis. That's what you're doing most of the time.


panther

unread,
Oct 26, 1999, 3:00:00 AM10/26/99
to
In article <87XdtVqF0GQ7-pn2-HwUneY1Lwrzr@localhost>,

rmc...@nospam.banet.net (Ron McDermott) wrote:
> panther wrote:
>
MK. Discussion deleted (NAEP test scores, related to district size and
age of compulsory attendance)...

>
> > MK. NO! The scores you inspected were student percentile scores.
>
> The percentiles are for the state itself, yes? Then the score listed
> as 10th percentile represents a score which is better than 10 percent
> of that state's test takers? But the test is national? So what is
> the national score which represents these percentiles, I wonder?
>
MK. It's available. Not useful for this analysis (influence of age and
district size on score).

>
> > >... Ok, I'll consider the implications of that
> > > information. Off the top of my head it occurs to me that the
> > > percentage of minorities and inner-city dwellers in the state is
going
> > > to be telling factor. If this is the case (as I suspect it may
be),
> > > then I'd expect to see low scores in such states regardless of
> > > district size or entering age.
>
>...> I will address only your last point: Whites are a minority in
>...> Washington DC, yet the DC White mean score is the highest in
>...> the nation. Whites are a minority in Hawaii.
>
>...> The implication intended is that all WHITES are the same
>...> everywhere? That must mean all BLACKS are the same
>...> as well?
>
>...> MK. Get serious, or leave.

>
> > > I'm deadly serious; your assuption above is that there are no
> > > demographic differences in white populations from state to state,
nor
> > > in black, nor asian, etc. IS that your view, since you are using
it
> > > to try to support your argument? Surely you are well-aware that
the
> > > white population of DC is largely college-educated, professional,
etc?
> > > You wouldn't have been trying to slip a change-up by me?
> > >
> > MK. I do assume that the distribution of genetic endowment is the
same,
> > nationwide, within racial groups.
>
> I assume the same.

>
> > I do not assume that "culture" (whatever that means) is the same,
> > nationwide.
>
> Ok; your statement suggested otherwise.
>
MK. How so? I don't see that. If "culture" is the environmental part of
human activity (as opposed to the hereditary part, then age of
attendance and district size -are- cultural variables, and I obviously
consider these to vary.

>
> > Both the mean score of children of White, college-educated parents
> > AND the mean score of children of White, high-school educated
> > parents in DC are the highest in the nation.
>
> And you account for this... how?
>
MK. In large districts, the politically adept parents get magnet
schools and GT classes for their kids. Politically adept parents and
the bureaucracy cut a deal: The children of the elite get a decent
education, and the bureaucracy gets to rip everyone else off.
>
> > MK. Discussion deleted...

>
> > >The logical extension of your statement as to the effect of
> > > starting age on math results would be that having "kids" enter
> > > school at 50 or 60 is preferable to having them enter at 5 or 6...
>
> > MK. Why: "logical"? We only have data within a narrow range of
starting
> > ages. It would be my preference to repeal compulsory attendance
laws
> > entirely, but that's another argument.
>
> If later is better, at what point does it cease to be?
>
MK. Probably around 95 years old. But I wouldn't say an empirical
guesstimate is "logical". "Reasonable" probably.
>
MK. Discussion deleted...

>
> The data is immaterial, as I've said repeatedly; it's the conclusions
> you draw from them which are flawed.
>
MK. Maybe, but you haven't done any comparable work.

panther

unread,
Oct 26, 1999, 3:00:00 AM10/26/99
to
In article <381876b6....@newsreader.cais.net>,

loj...@lojban.org (Bob LeChevalier) wrote:
> panther wrote:
>
MK. Discussion deleted (School district size, age of compulsory
attendance, and NAEP math scores)...
>
> >MK.... (New Jersey)...It is entirely

> >consistent with the model. The school districts of Newark and Jersey
> >City get more money than the average for the State. Poor minorities
> > are less adept at political action.
>
> Sounds contradictory to your claim, if a school district with lots of
poor
> minorities gets more money than a district with fewer. Government
money
> tends to be a direct result of political action.
>
MK. In this case, by the education bureaucracy. Have you heard of
Munchausen's Syndrome, where people injure themselves for the sympathy
the injury generates? Then there's Munchausen's Syndrome By Proxy,
where people (e.g. parents) injure others (usually children) for the
sympathy and acclaim as "caregivers". This is the relation between the
NEA/AFT/AFSCME cartel, with its attendant leeches, and poor minority
children.

>
> >Political control of schools does the
> >most harm to the children of the least politically adept parents.
>
> That is a causal claim. You have proven no causality, and have
admitted
> that there are multiple factors correlated with school performance,
some of
> which you have not identified. They are not ALL causes.
>
MK. It's not worth a discussion over English usage.

>
> > The damage done to children is directly related to the enrollment
> > (and so to aggregate budget,...

>
> You have not proven any DIRECT relation. There might be an indirect
> relation (assuming we could agree on the phrase "damage done to
children").
>
> >and so to the incentive for non-parent groups to
> >take school policy away from parents).
>
> Non-parent groups have every incentive to take school policy away.
> It is THEIR tax money too.
>
MK. Good! we agree that these groups find legal justification for their
activity. The cartel's interests are farther fron the interests of
children, or parents, or taxpayers, than these are from each other.
>
MK. Discussion deleted...

>
Take care. Homeschool if you can.

Bob LeChevalier

unread,
Oct 27, 1999, 3:00:00 AM10/27/99
to
panther <igon...@my-deja.com> wrote:
>In article <38197d86....@newsreader.cais.net>,
> loj...@lojban.org (Bob LeChevalier) wrote:
>> >(b) The pattern -does- hold, across the US, for Whites, overall. It
>> >does not hold for children of White, college-educated parents. Even
>> >these, when a few States are subtracted, parallel the trend (not as
>> >strongly).
>>
>> In other words, if you fudge the data enough by selectively omitting
>points
>> that you don't like, you can force your conclusion, sort of.
>>
>MK. You-all were happy to endorse deleting one or two States when Ron
>claimed (falsely) that it reversed the sign of the correlations for
>(test score, district size).

It did? Not my recollection.

The deletion was of extremes, and upon removal of the extremes the
remaining states did not show the correlation (and may have shown the
reverse). The result of the analysis was therefore that no conclusion
could be reached.

There is a much higher standard if you want to delete data and then use
that to argue a conclusion as opposed to arguing against jumping to a
conclusion.

>Odd cases are very instructive: HIV
>positive survivors, good minority urban schools, States with large
>districts and high test performance (Utah),

If they are instructive, they are not instructive statistically. Rather
they are good sources of hypotheses to test rather than conclusions to be
assumed.

>> >Hawaii is confirmation, since whites do not do so well here, and whites
>> >are not the politically adept group.
>>
>> a) this assumes that one can stereotype races as a "group" for purposes of
>> political action when neither whites nor blacks nor any other race
>> necessarily has common political goals in any state.
>>
>MK. People vote along racial lines often enough for politicians to
>compose strategy on that basis.

They may do so on issues where there is racial division. Your argument is
that there is no racial division because all groups are harmed by the
schools.

>MK. The point was: There are few Blacks (whatever level of education)
>in Hawaii. Those who live here are disproportionately military,

No supporting data for this

>therefore employed,

Are you claiming that most blacks not in the military are unemployed?

>therefore, I expect, their children would test -
>above- the average (Black, high-school educated parent).

Are you claiming that most high-school-educated blacks not in the military
are unemployed?

Are you claiming that being employed causes their children to test higher?

Are you claiming that their race has any relevance to any of these
questions?

>The Hawaii mean score (NAEP 8th grade Numbers and Operations) is the
>lowest in the nation.

Why do you focus on a subtest and not the entire math score? You invite
questions, especially when the public usually looks at the comprehensive
score when considering the following:

>I'm saying, it's not the students, it's the
>wretched school system. Military parents resist assignment to Hawaii,
>expressly because our schools are notoriously wretched.

"Expressly". Then you have evidence of this? Do they desire assignment to
ND because the schools are so good? (The black percentage in ND is
probably also largely military, I suspect). I doubt that there is a
groundswell of applications to live in Minot.

>> ... And how well do children of the military
>> do elsewhere in the country (it would seem to me that mobility might be a
>> factor reducing scores more than the quality of the school system, given
>> parents without a lot of extra resources)?
>>
>MK. Interesting question. I have no idea where to get those data. NAEP
>does give scores for DOD schools.

Yep, and for 8th grade, blacks scored as high or higher in the DoD schools
than in any of the state except Nebraska. The DoD schools are of course
predominantly overseas, and are designed around the predictable transience
of their students; it is likely that a student going to a new DoD school
will have the same curriculum and relatively efficient transference of
records.

>> >Military Blacks in Hawaii are thrice unfortunate, wrt political
>> >influence: There are few Blacks, military personnel usually vote
>> >elsewhere, and politicians and the bureaucracy know they'll rotate out
>> >of here in two or three years.
>>
>> And 4th, they might not have any interest in voting as a political bloc
>> based on your racist...
>>
>MK. You say the sweetest things. Give us a kiss.
>>
>> ...stereotype, a stereotype that implies that they SHOULD
>> form such a bloc.

You claim that blacks can be treated as a bloc, and that this is a natural
thing. Yes that is racist.

>MK. Later in this post you defend Alan's claim of a genetic basis for
>the difference in student test performance.

Defend? Not in the least. I merely noted that such a claim existed in
response to your assumption that there were no differences:


>> >MK. I do assume that the distribution of genetic endowment is the
>> > same, nationwide, within racial groups.

Alan points out that this is arguable; I don't know one way or the other.
Therefore I do not accept the validity of your assuming something that is
contested.

>Who's the racist? You are confused.

Someone who assumes that racial differences are important and that race
should be the basis of decisions. Someone who observes an apparent
difference based on race is not a racist, unless he applies that perceived
difference to decisions which are made based significantly on the racial
factor.

You argued for a racist decision (the formation of a political bloc). I
merely oppose that. I do not favor anything that Alan may have posted
about decisions to be made based on his referenced reports regarding
intelligence of blacks. I would not do so, *whether or not* blacks are
statistically more or less intelligent than whites. Intelligence is an
individual thing and statistical intelligence measurements of groups should
not be the basis of any decision making. I decry the formation of
political blocs based on color of skin; actually I decry the formation of
political blocs period, but especially on the basis of such superfluous
criteria as skin color.

>> ...like you in the state that damn them for
>> not fighting for their racial interests, they haven't a chance.
>>
>MK. Hardly. I observe that the difference between White mean scores and
>Black mean scores is smaller in States with numerous small school
>districts than in States which compel attendance in large school
>districts. My model is that the parent interests oppose the
>bureaucracy's interests diverge as districts increase in size. When
>districts are large, the bureaucracy wins, except for the children of
>the politically adept parents, who shield their children from the
>vampires, with GT classes, magnet schools, and tutors to do after
>school what the schools were paid to do.

And you assume (thereby helping make it true, because perceptions can cause
reality in politics) that blacks are politically unadept.

>> Furthermore, by what he has written above, he believes that the black
>> population of Hawaii is demographically quite different from the rest of
>> the country.
>>
>MK. Yes. Black parents in Hawaii are more likely to be in the military,
>to have completed high school, to be full-time employed, than are
>Blacks elsewhere.

And I want evidence for these claims. I am sure the Bureau of Labor
Statistics has them. The Statistical Abstract doesn't break the numbers
out by both state and race alas.

Your claims that blacks are more likely to be educated and employed seems
especially questionable. Blacks are usually educated and employed in most
parts of the country. Hawaii's percentage of population with less than a
high school education is lower than the national average, but not lower
than most of the northern non-urban states. Hawaii's 1995 unemployment
rate was above the national average, and well above the average for those
same northern non-urban states.

It would appear unlikely that your claims are true, in the absence of hard
numbers. Blacks are around 14.4% of the military. Hawaii has around 57K
military and 50K military dependents, which means around 15K blacks in
Hawaii are military. But there are around 40K blacks in Hawaii. So
military-related blacks are maybe 40% of all blacks. ND, SD, MT and WY all
probably have more of their black population among military residents based
on this calculation.

>>...Presumably he therefore knows that the white population is
>> also likely to be demographically different.
>>
>MK. That doesn't follow at all.

Why should only the blacks be demographically different in Hawaii?

>> >MK. I do assume that the distribution of genetic endowment is the
>> > same, nationwide, within racial groups.
>>
>> Why would you ASSume that?...
>>
>MK. Vulgar child. Grow up.

%^o

>> ...Alan L. has been posting about the researchers
>> who believe that they have proven otherwise.
>
>MK. And...? You want to defend that view?

No. Merely that someone says that there is evidence that contradicts your
assumption, and this has been posted on this group within the last week
making it more likely that you have seen it. I have no idea if his book
states truth or whether it manipulates statistics as egregiously as you do,
but I feel it necessary to point out that your assumption is contested.

>What we have here is a
>reflexive critic: MK. is a racist if he asserts Blacks in Hawaii are
>more likely than Blacks elsewhere to be enlisted military,

That actually was not what I felt was racist. It was the claim that blacks
are less politically adept and that they are and should be treated as a
political bloc (especially on matters of education politics, where there is
little reason to believe that blacks have unique interests or problems
based on their race in Hawaii)

>and MK. is
>wrong if he assumes that the genetic endowment is the same, within
>racial groups, nationwide. Btw, that last is not what Alan addresses,
>He defends the idea that there are significant differences -between-
>racial groups.

There may be, or maybe not. Whether those differences are important is
another question.

>> I looked in the 1996 Math summary report, that does not break things down
>> that far. It appears that around 1700 8th graders took the test in DC, of
>> which 6% were white - maybe 100. A different table says that 19% of DC
>> test takers had no more than high school education,...
>>
>MK. Are you saying that 81% of the all (Black and White, combined) DC
>parents have -more- than a high-school education? I would be surprised.

Confirmed in the Statistical Abstract, wherein 26.9% of DC residents over
25 have less than a high school graduation. It is reasonable that a lot of
these will be older people beyond childrearing age from before the time of
universal high school education (most people over 65), and immigrants from
countries without much secondary education. Then, people without much
education are less likely to form stable marriages. Finally, kids are
likely to report the parental education which is highest, so if one parent
has a high school education and the other doesn't, it probably is recorded
as high school graduated parents.

But the truth of the numbers is that, while blacks have a higher dropout
rate than whites, it isn't that much higher. High school dropouts these
days are disproportionately from the Hispanic population, which has close
to 50% of it membership with less than a high school education.

>> It should come as no surprise to you that the NAEP does not consider such a
>> small sample size to be adequate to reach any conclusions whatsoever.
>> Indeed in Appendix A to the summary report they discuss the statistical
>> issues at length. It becomes clear on even a brief skimming why they don't
>> release raw data to people like you who have no sensitivity to the proper
>> treatment and use of statistical data....
>>
>MK. Says the man who insisted that I apply correlation to cyclic
>(seasonal) data. BTW, can you supply the reference, the name of the
>statistics book with "cross-correlation" in the index, and which
>discusses "standard outlier techniques". You pretend familiarity with
>authorities you do not know.

This was answered by someone else who used the terminology (more)
correctly. I do make errors, and I have admitted that I am not a
statistician - I am better trained at knowing what NOT to do than in how to
do things correctly.

>> >MY. Your "logical" extension is silly.
>>
>> So is your entire statistical edifice on which the argument is
>> based. That is our point. Your model sounds as ridiculous to us as
>> this one sounds to you.
>>
>MK. I can understand how someone who is "agnostic" on the subject of
>Marxism might question an argument against the State monopoly school
>system. Most of us didn't sleep through the twentieth century.

I am not the only one who has questioned your logic. Even Mr. Gossman, who
supports a program similar to yours, and is I believe similar to yoyu in
political outlook, found your statistical argumentation unacceptable.

>> >> You don't care to have your analysis technique used for satirical
>> >> effect?
>> >>
>> >MK. You haven't done that. I do not see any tabular data with model and
>> >proposal, from your side.
>>
>> Maybe we don't think that such tabular data can prove anything useful and
>> would not deign to prepare a proposal based on faulty statsitics.
>>
>MK. How about some -good- statistics, then? Your proud indifference to
>evidence is stunning.

That is Alberto, not me. I am not indifferent to evidence, even
statistical evidence if I am convinced that the methodology and analysis
are sound. I will even accept the plausibility something contrary to my
personal beliefs, which is why I don't rule out the racial intelligence
claims that Alan posted; my standards will be high on that (and frankly I
don't consider the claim important enough to any decision-making to bother
right now).

As others have pointed out, it is possible that there CANNOT be "good"
statistics (i.e. statistics that clearly point to a conclusion) in this
area. The complex of relationships may be too complex to sort out.

I have presented no "good statistics" because the job takes a lot of time,
something I don't have. Furthermore, there is less motivation on my part,
since I have no expectation of proving anything useful to the improvement
of education. Rather I only expect to disprove your claims, which doesn't
accomplish much since I doubt that you would accept disproof and suspect
that you would only find some other kooky statistical claims.

Nonetheless, I am slowly assembling some other data to check for
correlations with the numbers you posted. Since I don't have a spreadsheet
program, and don't have the formulas down, this is not something easy for
me to accomplish. But I will report ALL the correlation values if I get
that far, not merely the ones I like.

>MK. This book review comes from a site linked to
>"www.schoolchoices.org"...
>>
>The Language Wars and Other Writings for Homeschoolers
>Ruth Beechick (Pollock Pines, CA: Arrow Press, 1995) (ISBN 0-940319-09-
>8). US$16.00, 252 pages. Collection of articles by Ruth Beechick first
>published elsewhere. Favorite quotation: "A research idea I have never
>seen carried out is to study whether this or any other 'scientific'
>program can outdo the alternative of a loving mother and her child
>together on a couch or at the kitchen table with a few good books and
>some simple teaching materials."

I don't contest that. Homeschooling can work if you have parents of the
right personality (merely "loving" is not enough; patience is a needed
virtue, not to mention either the means to do without any lost income from
the homeschooling parent), who have and are willing to spend the needed
time, and who can and will get the materials needed. One-on-one education
can do wonders, but is incredibly inefficient economically. There is a
reason why most cottage industries have ceased to be cottage industries in
the last century.

On the other hand, being a parent and being a teacher are two different
roles, sometimes conflicting. This becomes more true as the child reaches
puberty and naturally starts to move apart from his parents into
independent adulthood.

The bottom line is that in a world where there was no universal public
education, and far more parents were married and at home than at present,
few got a private education, and homeschooling when it occurred tended to
stop at minimal reading and writing.

panther

unread,
Oct 27, 1999, 3:00:00 AM10/27/99
to
In article <38172448....@newsreader.cais.net>,
loj...@lojban.org (Bob LeChevalier) wrote:
> panther wrote:
>
MK. The topic is NAEP test scores, related to age of compulsory
attendance and enrollment in large district6s. States that compel
attendance at age 7 have higher scores than States that compel
attendance at age 6. States which assign a large fraction of total
enrollment to large districts have lower scores than States with
numerous small districts.

>
> >> >(b) The pattern -does- hold, across the US, for Whites, overall.
It
> >> >does not hold for children of White, college-educated parents.
Even
> >> >these, when a few States are subtracted, parallel the trend (not
as
> >> >strongly).
> >>
> >> In other words, if you fudge the data enough by selectively
> >>omitting points that you don't like, you can force your
> >>conclusion, sort of.
>
> >MK. You-all were happy to endorse deleting one or two States when Ron
> >claimed (falsely) that it reversed the sign of the correlations for
> >(test score, district size).
>
> It did? Not my recollection.
>
MK. Your recollection? You didn't do the work. You just took Ron
McDermott's word. It was juvenile rape arrests per 10,000 population 11-
17, by State, related to age of compulsory attendance. I went back, and
eliminated the two States with the lowest rate which compelled
attendance at 7 (most favorable to my hpothesis), and the two States
with the highest rates that compelled attendance at age 5 or 6 (most
favorable to my hypoyhests). The correlation (age, rate) was still
negative, but not as much. I could post the data all over, or you could
try to recover the thread.

>
> The deletion was of extremes, and upon removal of the extremes the
> remaining states did not show the correlation (and may have shown the
> reverse). The result of the analysis was therefore that no conclusion
> could be reached.
>
MK. The correlation remained negative. The result did not depend on
"one or two outliers" as Ron claimed.
>
MK. Discussion deleted...
>
MK. Generalization: Politically adept groups shield their children from
the malign effects of aggregation into large districts and early
compulsory attendance. "Race" relations:

>
> >> >Hawaii is confirmation, since whites do not do so well here, and
whites
> >> >are not the politically adept group.
> >>
> >> a) this assumes that one can stereotype races as a "group" for
purposes of
> >> political action when neither whites nor blacks nor any other race
> >> necessarily has common political goals in any state.
> >>
> >MK. People vote along racial lines often enough for politicians to
> >compose strategy on that basis.
>
> They may do so on issues where there is racial division. Your
argument is
> that there is no racial division because all groups are harmed by the
> schools.
>
MK. Where do I say that? Not all groups are harmed equally, and some
select few (e.g., State school teachers who send their kids to
parochial school, politically adept parents who get their kids assigned
to GT classes and exclusive magnet schools) benefit. Education and
income are not distributed equally across races. Less educated, poorer
parents lose their contest with the bureaucracy.

>
> >MK. The point was: There are few Blacks (whatever level of education)
> >in Hawaii. Those who live here are disproportionately military,
>
> No supporting data for this
>
Is the Pacific Ocean salty?

>
> >therefore employed,
>
> Are you claiming that most blacks not in the military are unemployed?
>
MK. No. I am saying that 100% of blacks in the military -are- employed,
and whatever the rate of black employment generally, it's lower than
that.

>
> >therefore, I expect, their children would test -
> >above- the average (Black, high-school educated parent).
>
> Are you claiming that most high-school-educated blacks not in the
military are unemployed?
>
MK. See above.

>
> Are you claiming that being employed causes their children to test
higher?
>
MK. "They" being Hawaii's Black children? There is a positive
correlation between family income and test score, so, yes, it -
should-. That it does not is a puzzle, explained by Hawaii's wretched
schools.

>
> Are you claiming that their race has any relevance to any of these
> questions?
>
MK. Look, Ron brought race into it. I answer.

>
> >The Hawaii mean score (NAEP 8th grade Numbers and Operations) is the
> >lowest in the nation.
>
> Why do you focus on a subtest and not the entire math score? You
> invite questions, especially when the public usually looks at the
> comprehensive score when considering the following:
>
MK. A decent question. I use Numbers and Operations (arithmetic) and
Algebra and Fubctions (algebra) subtests because these topics are more
likely to be uniform across the US. Math teachers have wildly different
ideas about geometry, and statistics.

>
> >I'm saying, it's not the students, it's the
> >wretched school system. Military parents resist assignment to Hawaii,
> >expressly because our schools are notoriously wretched.
>
> "Expressly". Then you have evidence of this?
>
MK. Yes. Newspaper reports. A forum on schools that addressed the
question of what the military might do. I attended the forum (spoke at
it, with budget statistics and test data). All branches of the service
attended. There is high-level concern over this. Someone told those Air
Force colonels and Navy omanders to attend. Someone who outranked
them.
>
MK. Discussion deleted...

>
> >> ... And how well do children of the military
> >> do elsewhere in the country (it would seem to me that mobility
might be a
> >> factor reducing scores more than the quality of the school system,
given
> >> parents without a lot of extra resources)?
> >>
> >MK. Interesting question. I have no idea where to get those data.
NAEP
> >does give scores for DOD schools.
>
> Yep, and for 8th grade, blacks scored as high or higher in the DoD
schools
> than in any of the state except Nebraska. The DoD schools are of
course
> predominantly overseas, and are designed around the predictable
transience
> of their students; it is likely that a student going to a new DoD
school
> will have the same curriculum and relatively efficient transference of
> records.
>
MK. Good work. Thanks for the support. Where did you get the data on
Black DOD NAEP performance?

>
> >> >Military Blacks in Hawaii are thrice unfortunate, wrt political
> >> >influence: There are few Blacks, military personnel usually vote
> >> >elsewhere, and politicians and the bureaucracy know they'll
rotate out
> >> >of here in two or three years.
> >>
> >> And 4th, they might not have any interest in voting as a political
bloc
> >> based on your racist...
> >>
> >MK. You say the sweetest things. Give us a kiss.
> >>
> >> ...stereotype, a stereotype that implies that they SHOULD
> >> form such a bloc.
>
> You claim that blacks can be treated as a bloc, and that this is a
> natural thing. Yes that is racist.
>
MK. (a) Unclear. Is it racist to treat a racial group as a block, or is
it racist to observe that some group is treated as a block? (b) This
notion of a "block" is an overstatement. "Differential impact" is
enough to account for the statistical result. (c) Finally, one last
time (I hope) I address Ron's explanation for lower scores in large
districts (he says "large district" = "urban"+"minority"+"poor").

>
> >MK. Later in this post you defend Alan's claim of a genetic basis for
> >the difference in student test performance.
>
> Defend? Not in the least. I merely noted that such a claim existed
> in response to your assumption that there were no differences:
>
MK. I see. So you throw arguments in all directions, whether you
believe them or not. That's been my impression for quite a while.

>
> >> >MK. I do assume that the distribution of genetic endowment is the
> >> > same, nationwide, within racial groups.
>
MK. Discussion deleted (Bob justifies ad hominem argument)...

>
> You argued for a racist decision (the formation of a political bloc).
>
MK. Where? Nowhere do I argue for political decision-making along
racial lines, or the formation of racial voting blocks or pressure
groups. Parents of all races would gain from enhanced control over
education that vouchers, tuition tax credits, or subsidized
homeschooling would provide. Since politically powerless people suffer
most from political control of schools, politically powerless groups
have the most to gain, which is how the racial issue enters our
discussion.
>
>MK. Discussion deleted...

>
> >> ...like you in the state that damn them for
> >> not fighting for their racial interests, they haven't a chance.
> >>
> >MK. Hardly. I observe that the difference between White mean scores
and
> >Black mean scores is smaller in States with numerous small school
> >districts than in States which compel attendance in large school
> >districts. My model is that the parent interests oppose the
> >bureaucracy's interests diverge as districts increase in size. When
> >districts are large, the bureaucracy wins, except for the children of
> >the politically adept parents, who shield their children from the
> >vampires, with GT classes, magnet schools, and tutors to do after
> >school what the schools were paid to do.
>
> And you assume (thereby helping make it true, because perceptions can
cause
> reality in politics) that blacks are politically unadept.
>
MK. Do I assume that education and wealth are not uniformly distributed
along racial lines? Yes. Do I assume that education and wealth
contribute to political influence? Yes. Do I assume that numbers count
in democratic politics? Yes. Do you deny any of this?

>
> >> Furthermore, by what he has written above, he believes that the
black
> >> population of Hawaii is demographically quite different from the
rest of
> >> the country.
> >>
> >MK. Yes. Black parents in Hawaii are more likely to be in the
military,
> >to have completed high school, to be full-time employed, than are
> >Blacks elsewhere.
>
> And I want evidence for these claims. I am sure the Bureau of Labor
> Statistics has them. The Statistical Abstract doesn't break the
numbers
> out by both state and race alas.
>
> Your claims that blacks are more likely to be educated and employed
seems
> especially questionable. Blacks are usually educated and employed in
most
> parts of the country. Hawaii's percentage of population with less
than a
> high school education is lower than the national average, but not
lower
> than most of the northern non-urban states...
>
MK. Talk about selection od data! Why do you compare Hawaii to that
subset?
>
> ...Hawaii's 1995 unemployment rate was above the national average,

> and well above the average for those same northern non-urban states.
>
> It would appear unlikely that your claims are true, in the absence of
hard
> numbers. Blacks are around 14.4% of the military. Hawaii has around
57K
> military and 50K military dependents, which means around 15K blacks in
> Hawaii are military. But there are around 40K blacks in Hawaii. So
> military-related blacks are maybe 40% of all blacks. ND, SD, MT and
WY all
> probably have more of their black population among military residents
based
> on this calculation.
>
MK. Pretty good. No argument. You've answered your own question.
>
MK. Discussion deleted...

>
> >What we have here is a
> >reflexive critic: MK. is a racist if he asserts Blacks in Hawaii are
> >more likely than Blacks elsewhere to be enlisted military,
>
> That actually was not what I felt was racist. It was the claim that
blacks
> are less politically adept and that they are and should be treated as
a
> political bloc (especially on matters of education politics, where
there is
> little reason to believe that blacks have unique interests or problems
> based on their race in Hawaii)
>
MK. Where have I said blacks "are and should be treated as a political
block"? I -have- said that political control of schools harms the
children of the least politically adept parents. Since I have pointed
out that neither whites nor blacks are the politically adept group
(Japanese are significantly overrepresented in the State bureaucracy),
if that's racist, then I must hate whites. I'm caucasian.

>
> >and MK. is
> >wrong if he assumes that the genetic endowment is the same, within
> >racial groups, nationwide. Btw, that last is not what Alan addresses,
> >He defends the idea that there are significant differences -between-
> >racial groups.
>
> There may be, or maybe not. Whether those differences are important
> is another question.
>
MK. My point here is that Alan's work is irrelevant to this issue. It
does not touch the assumptions I make about the regional distribution
of genetic endowment -within- groups.

>
> >> I looked in the 1996 Math summary report, that does not break
things down
> >> that far. It appears that around 1700 8th graders took the test
in DC, of
> >> which 6% were white - maybe 100. A different table says that 19%
of DC test takers...
>
(MK. insert "parents" here, since high school graduates do not take the
8th grade NAEP. It still doesn't make sense.)...
>
> ... had no more than high school education,...
MK. The question was "...more (!)than a high school education", not
"...at least a high school education". There's a big difference.
>
MK. Discussion deleted...
>
> >>... It becomes clear on even a brief skimming why they don't

> >> release raw data to people like you who have no sensitivity to the
> >> proper treatment and use of statistical data....
> >>
> >MK. Says the man who insisted that I apply correlation to cyclic
> >(seasonal) data. BTW, can you supply the reference, the name of the
> >statistics book with "cross-correlation" in the index, and which
> >discusses "standard outlier techniques". You pretend familiarity with
> >authorities you do not know.
>
> This was answered by someone else who used the terminology (more)
> correctly. I do make errors, and I have admitted that I am not a
> statistician - I am better trained at knowing what NOT to do than in
how to
> do things correctly.
>
MK. I see. These trainers recommend calling "racist" people who observe
that politically powerless people suffer at the hands of the political
elite, do they?
>
MK. Discussion deleted...

Bob LeChevalier

unread,
Oct 27, 1999, 3:00:00 AM10/27/99
to
panther <igon...@my-deja.com> wrote:
>MK. In this case, by the education bureaucracy. Have you heard of
>Munchausen's Syndrome, where people injure themselves for the sympathy
>the injury generates? Then there's Munchausen's Syndrome By Proxy,
>where people (e.g. parents) injure others (usually children) for the
>sympathy and acclaim as "caregivers". This is the relation between the
>NEA/AFT/AFSCME cartel, with its attendant leeches, and poor minority
>children.

Curious how you first say "e.g. parents" which makes sense given the rest
of the sentence, and then somehow bring in your phony cartel. Teachers I
know want to be considered "professionals", not "caregivers".

>> >Political control of schools does the
>> >most harm to the children of the least politically adept parents.
>>
>> That is a causal claim. You have proven no causality, and have admitted
>> that there are multiple factors correlated with school performance, some of
>> which you have not identified. They are not ALL causes.
>>

>MK. It's not worth a discussion over English usage.

So you'll probably stick to your semantically sloppy soundbite
sloganeering.

>MK. Good! we agree that these groups find legal justification for their
>activity. The cartel's interests are farther fron the interests of
>children, or parents, or taxpayers, than these are from each other.

Semantically null. There is no cartel.

Bob LeChevalier

unread,
Nov 1, 1999, 3:00:00 AM11/1/99
to
panther <igon...@my-deja.com> wrote:
>In article <38172448....@newsreader.cais.net>,
> loj...@lojban.org (Bob LeChevalier) wrote:
>> panther wrote:
>>
>MK. The topic is NAEP test scores, related to age of compulsory
>attendance and enrollment in large district6s. States that compel
>attendance at age 7 have higher scores than States that compel
>attendance at age 6. States which assign a large fraction of total
>enrollment to large districts have lower scores than States with
>numerous small districts.
>>
>> >MK. You-all were happy to endorse deleting one or two States when Ron
>> >claimed (falsely) that it reversed the sign of the correlations for
>> >(test score, district size).
>>
>> It did? Not my recollection.
>>
>MK. Your recollection? You didn't do the work.

You didn't give us the data. I have no obligation to redig up your data
that I am sure will prove irrelevant.

>You just took Ron McDermott's word.

He has more credibility that you. Thorain has even more on matters
statistical, and mentioned recently why a correlation of even .333 would be
meaningless given no more than the distribution of ages you are correlating
with.

>It was juvenile rape arrests per 10,000 population 11-
>17, by State, related to age of compulsory attendance. I went back, and
>eliminated the two States with the lowest rate which compelled
>attendance at 7 (most favorable to my hpothesis), and the two States
>with the highest rates that compelled attendance at age 5 or 6 (most
>favorable to my hypoyhests). The correlation (age, rate) was still
>negative, but not as much. I could post the data all over, or you could
>try to recover the thread.

I believe that the claim is that you never posted the data, just your
correlations.

>MK. Generalization: Politically adept groups shield their children from
>the malign effects of aggregation into large districts and early
>compulsory attendance.

Unsupported by anything.

>> >MK. The point was: There are few Blacks (whatever level of education)
>> >in Hawaii. Those who live here are disproportionately military,
>>
>> No supporting data for this
>>
>Is the Pacific Ocean salty?
>>
>> >therefore employed,
>>
>> Are you claiming that most blacks not in the military are unemployed?
>>
>MK. No. I am saying that 100% of blacks in the military -are- employed,
>and whatever the rate of black employment generally, it's lower than
>that.
>>
>> >therefore, I expect, their children would test -
>> >above- the average (Black, high-school educated parent).
>>
>> Are you claiming that most high-school-educated blacks not in the
>military are unemployed?

Why would they score better than the average Black, high school educated
person? Your leaps of logic astound me.

>> Are you claiming that being employed causes their children to test
>higher?
>>
>MK. "They" being Hawaii's Black children? There is a positive
>correlation between family income and test score, so, yes, it -
>should-.

Yes, but aremilitary higher paid than most people? The reverse, to my
knowledge.

> That it does not is a puzzle, explained by Hawaii's wretched
>schools.
>>
>> Are you claiming that their race has any relevance to any of these
>> questions?
>>
>MK. Look, Ron brought race into it. I answer.

Are you claiming it is relevant to YOUR claims.

>> >The Hawaii mean score (NAEP 8th grade Numbers and Operations) is the
>> >lowest in the nation.
>>
>> Why do you focus on a subtest and not the entire math score? You
>> invite questions, especially when the public usually looks at the
>> comprehensive score when considering the following:
>>
>MK. A decent question. I use Numbers and Operations (arithmetic) and
>Algebra and Fubctions (algebra) subtests because these topics are more
>likely to be uniform across the US. Math teachers have wildly different
>ideas about geometry, and statistics.

Why do the opinions of math teachers matter? I think that the topics of
math course are determined by state standards. The fact that scores on
your subtests vary so much across the country shows that they are no more
uniform than other portions of the test.

>> >I'm saying, it's not the students, it's the
>> >wretched school system. Military parents resist assignment to Hawaii,
>> >expressly because our schools are notoriously wretched.
>>
>> "Expressly". Then you have evidence of this?
>>
>MK. Yes. Newspaper reports. A forum on schools that addressed the
>question of what the military might do. I attended the forum (spoke at
>it, with budget statistics and test data). All branches of the service
>attended. There is high-level concern over this. Someone told those Air
>Force colonels and Navy omanders to attend. Someone who outranked
>them.

Right. No one of that rank could decide on their own that a forum on
military assignments deserves to have a military person of authority
present.

>> Yep, and for 8th grade, blacks scored as high or higher in the DoD
>schools
>> than in any of the state except Nebraska. The DoD schools are of
>course
>> predominantly overseas, and are designed around the predictable
>transience
>> of their students; it is likely that a student going to a new DoD
>school
>> will have the same curriculum and relatively efficient transference of
>> records.
>>
>MK. Good work. Thanks for the support. Where did you get the data on
>Black DOD NAEP performance?

The NAEP math summary report.

>> >MK. You say the sweetest things. Give us a kiss.
>> >>
>> >> ...stereotype, a stereotype that implies that they SHOULD
>> >> form such a bloc.
>>
>> You claim that blacks can be treated as a bloc, and that this is a
>> natural thing. Yes that is racist.
>>
>MK. (a) Unclear. Is it racist to treat a racial group as a block, or is
>it racist to observe that some group is treated as a block? (b) This
>notion of a "block" is an overstatement. "Differential impact" is
>enough to account for the statistical result. (c) Finally, one last
>time (I hope) I address Ron's explanation for lower scores in large
>districts (he says "large district" = "urban"+"minority"+"poor").

No you don't. You use that mention to launch into irrelevancies about
black military people in Hawaii.

>> >MK. Later in this post you defend Alan's claim of a genetic basis for
>> >the difference in student test performance.
>>
>> Defend? Not in the least. I merely noted that such a claim existed
>> in response to your assumption that there were no differences:
>>
>MK. I see. So you throw arguments in all directions, whether you
>believe them or not. That's been my impression for quite a while.

No. I am saying that you made a claim which Alan has cited a source that
disagrees with that claim. I don't necessarily believe either of you, and
the point of citing Alan is that you cannot assume what you assumed
uncontestably.

>> >> >MK. I do assume that the distribution of genetic endowment is the
>> >> > same, nationwide, within racial groups.
>>
>MK. Discussion deleted (Bob justifies ad hominem argument)...
>>
>> You argued for a racist decision (the formation of a political bloc).
>>
>MK. Where? Nowhere do I argue for political decision-making along
>racial lines, or the formation of racial voting blocks or pressure
>groups. Parents of all races would gain from enhanced control over
>education that vouchers, tuition tax credits, or subsidized
>homeschooling would provide. Since politically powerless people suffer
>most from political control of schools, politically powerless groups
>have the most to gain, which is how the racial issue enters our
>discussion.

You have claimed that they are politically powerless because they form a
political voting bloc, and you have not proven that they do so. There are
blacks who vote on different sides of issues (Reagan appointed a black to
the USSC who is among the most conservative and powerful people in the
country), so the assumption must not be a bloc.

>> >MK. Hardly. I observe that the difference between White mean scores
>and
>> >Black mean scores is smaller in States with numerous small school
>> >districts than in States which compel attendance in large school
>> >districts. My model is that the parent interests oppose the
>> >bureaucracy's interests diverge as districts increase in size. When
>> >districts are large, the bureaucracy wins, except for the children of
>> >the politically adept parents, who shield their children from the
>> >vampires, with GT classes, magnet schools, and tutors to do after
>> >school what the schools were paid to do.
>>
>> And you assume (thereby helping make it true, because perceptions can
>cause
>> reality in politics) that blacks are politically unadept.
>>
>MK. Do I assume that education and wealth are not uniformly distributed
>along racial lines? Yes. Do I assume that education and wealth
>contribute to political influence? Yes. Do I assume that numbers count
>in democratic politics? Yes. Do you deny any of this?

Please state your chain of logic more clearly/

>> >> Furthermore, by what he has written above, he believes that the black
>> >> population of Hawaii is demographically quite different from the rest of
>> >> the country.
>> >>
>> >MK. Yes. Black parents in Hawaii are more likely to be in the military,
>> >to have completed high school, to be full-time employed, than are
>> >Blacks elsewhere.
>>
>> And I want evidence for these claims. I am sure the Bureau of Labor
>> Statistics has them. The Statistical Abstract doesn't break the numbers
>> out by both state and race alas.
>>
>> Your claims that blacks are more likely to be educated and employed seems
>> especially questionable. Blacks are usually educated and employed in most
>> parts of the country. Hawaii's percentage of population with less than a
>> high school education is lower than the national average, but not lower
>> than most of the northern non-urban states...
>>
>MK. Talk about selection od data! Why do you compare Hawaii to that
>subset?

Because you started talking about Hawaii's military blacks and somehow were
comparing them in some way to people with a high school education.

>> ...Hawaii's 1995 unemployment rate was above the national average,
>> and well above the average for those same northern non-urban states.
>>
>> It would appear unlikely that your claims are true, in the absence of hard
>> numbers. Blacks are around 14.4% of the military. Hawaii has around 57K
>> military and 50K military dependents, which means around 15K blacks in
>> Hawaii are military. But there are around 40K blacks in Hawaii. So
>> military-related blacks are maybe 40% of all blacks. ND, SD, MT and WY all
>> probably have more of their black population among military residents based
>> on this calculation.
>>
>MK. Pretty good. No argument. You've answered your own question.

Is there a questin in that paragraph. I said that it appears your claims
are not true.

>> That actually was not what I felt was racist. It was the claim that
>blacks
>> are less politically adept and that they are and should be treated as
>a
>> political bloc (especially on matters of education politics, where
>there is
>> little reason to believe that blacks have unique interests or problems
>> based on their race in Hawaii)
>>
>MK. Where have I said blacks "are and should be treated as a political
>block"? I -have- said that political control of schools harms the
>children of the least politically adept parents. Since I have pointed
>out that neither whites nor blacks are the politically adept group

YOu are treating them as a group for political purposes. You should not be
treating racial groups as political groups, should not be labelling them as
a "political adept" or not "political adept". You cannot label a group
unless you first consider them a group. Blacks are not a group for
purposes of determining political beahvior or political power. To assme
that they are a political group is racist.

>(Japanese are significantly overrepresented in the State bureaucracy),
>if that's racist, then I must hate whites. I'm caucasian.

You are racist in using the racial labels in talking about political
groups.

>> >and MK. is
>> >wrong if he assumes that the genetic endowment is the same, within
>> >racial groups, nationwide. Btw, that last is not what Alan addresses,
>> >He defends the idea that there are significant differences -between-
>> >racial groups.
>>
>> There may be, or maybe not. Whether those differences are important
>> is another question.
>>
>MK. My point here is that Alan's work is irrelevant to this issue.

It contradicts your claim, nothing more, thereby indicating that you need
to support your claim.

>MK. I see. These trainers recommend calling "racist" people who observe
>that politically powerless people suffer at the hands of the political
>elite, do they?

You call them powerless because they are members of a racial group.

Reply all
Reply to author
Forward
0 new messages