October 16, 2009

Latest NAEP math results

Charles Murray blogs:
The narrowing of the black-white gap extends from children born in 1961 through children born in 1973, and it was substantial—from 1.2 standard deviations to about .8. Then the trend goes flat, with a few spikes, for children born over the next 26 years.


My published articles are archived at iSteve.com -- Steve Sailer

48 comments:

Henry Canaday said...

Where's that healthy abortion effect, Dr. Levitt?

Anonymous said...

The gap seems to have shrunk since 1961 because blacks have had more opportunities since 1961. Ultra low birthrates for college educated whites could cull the brightest whites and bring down the white average. Otherwise intelligent whites have over used birth control to commit collective suicide.

Concerned Netizen said...

Perhaps this reflects birth rate trends of the *previous* generation - that is, the more intelligent of the previous generation had more kids relative to the stupid. Look for this "achievement" gap to widen in future as the dummies outbreed the bright.

Just a guess.

Edward said...

The chart starts in 1961, the year the Birth Control Pill was invented. Then in 1973 you had Roe vs Wade - that's around the mark of the first dip on the graph.

The Black-White IQ gap has not fallen further than levels since the 1970s because even the lower class whites (whom, it is assumed, use birth control less and have fewer abortions) have higher baseline IQs than Blacks.

Anonymous said...

I'll probably catch hell for saying this, but if the difference really did decrease to anything less than 1.2 SD, then it's almost certainly due to the [intentional?] dumbing-down of the white kids [not any possible smartening-up of the black kids].

White kids [especially the ones in the public schools] are so poorly educated nowadays that it's simply flabbergasting.

Anonymous said...

The simplest way to subtly narrow the achievement gap is to make the easy test questions easier, but leave unchanged in difficulty the problems requiring involved reasoning. The easy questions tend to be the ones you can actually study for like 8/12 = 2/3. Harder questions requiring deductive reasoning are ones like like "A train one mile in length is traveling 60 mph. How long does it take to completely pass through a tunnel one mile in length?" 100% of iSteve readers can answer this one off the top of their heads, but 100% of NAMs will miss it, unless it's multiple choice.

California has been gaming the difficulty of the API test for years to demonstrate bogus gains for blacks and Hispanics. But you can only water down so much of a test before it becomes obvious that the fix is in. Now that 2/3 of kids in CA kindergartens are Hispanic, the level of white achievement is no longer even relevant.

William1066 said...

Given the spikes in the graph, (1981 and 86) can you really claim there is a trend? The starting point could also have been a spike.

You see a trend - I see a trading range.

albertosaurus said...

I'm sure all of the regular readers of this blog appreciate the implications of the 0.8 standard deviation difference, but let me walk through some of the math for new visitors.

The standard deviation of the Stanford-Binet IQ test is 16. The WAIS has an SD of 15. Let's assume 15 to keep the math easy.

The IQ of college students is around 117 but let's assume 115 ( one SD above the mean) to make the math easy. This means that we should expect a certain percentage of the populace to be at or above that IQ level. Checking the z-score of 1.0 on a table of areas under the normal curve we find the value of .34.

So "college material" would be those 16% above that score (50% plus 34% below). This seems about right. About one in six in appropriate for college.

That of course is assuming a mean IQ of 100. If the population mean were 12 points lower (0.8 times 15) then the college level would be achieved by only 4% of that population (z-score of .46 at 1.8)(.50 below the mean plus .46 equals 96%).

So the expectation would be that whites should be appropriate for college at a ratio of four to one. That means that undergraduates should be about 80/20 white versus black in a situation of no bias.

Many jobs require about college level smarts (plus 1.0 SD). The good jobs often require more yet. For example doctors and lawyers are about st 2 SDs up and Nobel Prize winners (hard sciences) are at about 3 SDs.

The Griggs case gave us the legal theory of "disparate impact". This means that you don't need to actually have to observe racial bias, you can compute it from racial participation rates. One problem with Griggs is that it assumes 50-50 whereas we should expect 80-20 (at about the college level).

Elementary math shows that there is almost no racial bias today in America against blacks even using the Griggs criteria. That is unless you aren't very good at math and expect 50-50. Alas bad math is also what we can expect among blacks.

Anonymous said...

"100% of iSteve readers can answer this one off the top of their heads, but 100% of NAMs will miss it, unless it's multiple choice."

You have poor understanding of statistics.

Anonymous said...

"You have poor understanding of statistics."

Enlighten me.

Hyatt said...

Restrictions on lead in petrol since the 1970s have successfully reduced airborne levels. A good example which shows that environmental controls can be a good thing when they're impact is limited and their goal is specific.

Anonymous said...

Enlighten me.

Steve has NAM readers, thus rendering the statement that "100% of iSteve readers can answer this one off the top of their heads, but 100% of NAMs will miss it, unless it's multiple choice."
logically absurd.

William1066 said...

There goes that anonymous guy arguing with himself again. Reminds me of observing a poor NYC bag lady ranting at some invisible antagonist.

Seriously, for the sake of logical flow, would it be too much trouble to add a unique 'nom de poste' to your replies? Even something like Anonymous1066 would help. It's very easy to add - under name/url.

C. Van Carter said...

The obvious conclusion is that math is racist.

Anonymous said...

"The obvious conclusion is that math is racist."

No, No, No! Math is a "hate fact"!

Anthony said...

The drop at the beginning is likely the improvement of schools for black kids as Brown actually got implemented. The spikes after that probably reflect various demographic changes both in the black and white populations, and in their schools.

Anonymous said...

Just to be clear, the answer is one minute, right? Clap yo´hands!

HyperHormonalMonkey said...

"Restrictions on lead in petrol since the 1970s have successfully reduced airborne levels. A good example which shows that environmental controls can be a good thing when they're impact is limited and their goal is specific."

I've often wondered if environmental contamination added to the 'craziness' of the 1960s. We know that LSD had a huge effect back then but it seems like airborne lead, pesticides and other contaminants could have contributed to making the culture a bit 'off' as well.

JeremiahJohnbalaya said...

"100% of iSteve readers can answer this one off the top of their heads, but 100% of NAMs will miss it, unless it's multiple choice."

You have poor understanding of statistics.


To be fair, couldn't this simply be a matter of rounding error and/or the number of significant digits used in the original post.

Holy Sh-t!! Ben Franklin was a British spy???

Svigor said...

Two minutes. Er, right? Heh, math's never been my thing...

Anonymous said...

We shuda kept dem blacks dummer dan dey are now. Damn, Civil Rights Act of 1964.

Anonymous said...

2 minutes

the engine passes through in 1 minute

the last car of the one mile train passes through 1 minute later

Melykin said...

Albertosaurus,
If we assume that 4% of blacks and 16% of non-blacks have a suitably high IQ to go to college, then that doesn't mean we should expect the ratio of whites to blacks at college to be 4:1, because there are fewer blacks then whites in the general population.

Suppose that 13% of general population is black and 87% is not black. (which is what Wiki claims)

Blacks who go to college would be 4% of 13%, which is about 0.5% of the general population.

Non-blacks who go to college would be 16% of 87%, which is about 14% of the general population.

So the ratio of whites to blacks in college would be 14:.5, or 28:1.

In fact, probably there are a lot of people, black and not, who are in college and who have an IQ less than 116.

Today I saw a college student use a calculator to reckon 4 - 3. I kid you not!!

P.S. The train would take 2 minutes. 60 mph = 1 mile per minute. The engine of the train would take one minute to pass through the tunnel, then it would have to travel 1 mile beyond the tunnel's exit before its tail-end cleared the tunnel.

Anonymous said...

"White kids [especially the ones in the public schools] are so poorly educated nowadays that it's simply flabbergasting"

Not all white kids will sink so low, only the poor white kids. Check out the lower class white kids' score in England, their score is lower than students from any ethnicities. You are looking into the future, and nothing you can do about it.

Anonymous said...

It is so easy to game the system to show improvement in black students score. As in most exam you take, the questions get harder and harder throu the exam. Now, if you change just the medium difficult exam questions to some easy one, the lowest scoring group will benefit the most as the highest scoring group will get it right no matter what. By dumping dowm the right part of an exam, you can do away the gap just like that.
e.g. For 4 grade student
Q1: 5+5=
Q2: 5+4-1=
Q3: 1*5=
Q4: prove E=MC^2
i think most students will get a 80%, don't you think? so everybody is equal, aren't that great!

OneSTDV said...

The gap can be decreased if the tests are made easier and the decrease in difficulty doesn't have to be wildly obvious. An important part of the B-W gap comes from the more difficult questions which are heavily g-loaded. If you reduce the number of g-loaded questions and improve black education (which helps with respect to the less difficult, read/remember/regurgitate type questions), the gap should decrease.

2 minutes. The front of the train has to travel 2 miles at 60 mph or 1 mile/minute which equals 2 minutes total time.

Anonymous said...

(Cliff Arroyo)

It's certainly .... convenient how any data that doesn't fit Steve's readers' biases can be accounted for so handily.

Anonymous said...

Not all white kids will sink so low, only the poor white kids. Check out the lower class white kids' score in England, their score is lower than students from any ethnicities. You are looking into the future, and nothing you can do about it.

Sorry, I gotta disagree with you there.

As you get out towards IQ 130+, the kids become capable of teaching themselves.

But the bulk of the educable population sits in that IQ 100 to 120 sweetspot, and they're the ones who need a strict disciplinarian to hold their noses to the grindstone.

[Although even the IQ 130+ kids will do better if the right structure and discipline is applied to their educations.]

Anonymous said...

But the bulk of the educable population sits in that IQ 100 to 120 sweetspot, and they're the ones who need a strict disciplinarian to hold their noses to the grindstone.

[Although even the IQ 130+ kids will do better if the right structure and discipline is applied to their educations.]

_______________


From my years of teaching in public schools, I can tell you this is absolutely true.

You would not believe how much you can get out of bright middle school aged students who are well disciplined. They are ready for high school curriculum in about 7th grade, but have to wait for their peers to catch up.

Middle school is the biggest waste of time for smart kids. Even average kids could do much more.

read it said...

"Check out the lower class white kids' score in England, their score is lower than students from any ethnicities."


I am not from England so I don't know how they classify lower class whites. Household income? On public assistance?

Anyway, more interesting than the lower class white average would be the whole curve of lower class white, as well as their IQ tests vs. their achievement tests.

meep said...

The problem with the high IQ kids is that school is usually too easy for them... so they don't learn the discipline they need if they are to achieve anything real [whether in academia or the outside world]

Thus my idea of No Child's Ass Left Unkicked [extending Charles Murray's ideas on education]:
http://meep.livejournal.com/1479634.html

Anonymous said...

"Working-class white British boys falling behind everyone else at school"

Read more: http://www.dailymail.co.uk/news/article-511615/Working-class-white-British-boys-falling-school.html

Read it and weep.

Truth said...

"As you get out towards IQ 130+, the kids become capable of teaching themselves."

Great, then why waste resources sending them to school. Teach 'em to read by age five, lock them in their rooms with a stack of books 6 hours a day...presto, 20 years later brain surgeons.

read it said...

"Texas still has a significant "achievement gap" between Anglo and minority students....

Scores for black and Latino fourth-grade students in Texas are now where Anglo students in Texas were in 1992. And Texas scores for black and Latino eighth-graders are where whites were in 1990."

Among fourth graders, Anglos in Texas scored 23 points higher than blacks in Texas and 21 points higher than Latinos in Texas. Among Texas eighth graders, Anglos scored 29 points higher than African-Americans and 24 points higher than Hispanics.

Texas minority students outpaced their peers nationwide. Blacks in fourth and eighth grade finished third and first, respectively, on the math exam. Latinos finished eighth and fourth, respectively. In both cases, their average scores beat the national average for African-American and Hispanic students."

from Dallas Morning News
http://educationfrontblog.dallasnews.com/archives/2009/10/heres-how-to-interpret-texas-n.html


Steve,

Do you know whether the test has got easier since the 1990's?

read it said...

"Great, then why waste resources sending them to school. Teach 'em to read by age five, lock them in their rooms with a stack of books 6 hours a day...presto, 20 years later brain surgeons."


I know you are kidding but in a related story, students in Florida decided to teach themselves AP chemistry when the school didn't offer it:

http://www.miamiherald.com/295/story/1285256.html?pageNum=2&mi_pluck_action=page_nav#Comments_Container

Anonymous said...

I am frankly disappointed that nobody commented on my 'clap yo' hands comment', but instead answered the question. Why are you reading about Ben Franklin or some Austrain dude when the real genius is Jack O'Connell? Shame.

Anonymous said...

Great, then why waste resources sending them to school. Teach 'em to read by age five, lock them in their rooms with a stack of books 6 hours a day...presto, 20 years later brain surgeons.

I think you're trying to be sarcastic, but that describes more or less EXACTLY the childhood & early adulthood of IQ 130+ kids.

Anonymous said...

"Working-class white British boys falling behind everyone else at school"

Read more: http://www.dailymail.co.uk/news/article-511615/Working-class-white-British-boys-falling-school.html

Read it and weep
.

Liberals love these articles, they put on a sad face and pretend to be worried by it. But really they are punching the air and shouting "Yesssss!"

Read it and also notice this:

They show that just 15 per cent of white British boys who are eligible for free school meals because of low household income meet the Government's GCSE achievement target.

We are not so much talking about working class here rather unemployed/underclass. Not a good thing but not the larger group 'working class' tends to imply.

For Indian pupils from similar backgrounds, the figure is 36 per cent, for Pakistanis it is 26 per cent and for Bangladeshis 31 per cent.

Among black Caribbean pupils, the proportion is 17 per cent and black African 25 per cent
.

What the article neglects to mention is the percentage of children of each background who qualify for the free school meals. I rather suspect that, on average, black and Bangladeshi children qualify much more often than white. Im sure, statistically speaking, everyone can see the problem with that right there.

The article then goes on to flat out contradict its own title with this:

The only groups performing worse than low income- family white boys were their counterparts from gipsy families on 5 per cent and those of mixed black Caribbean and white heritage, on 13 per cent.

If those groups perform worse, then how can white boys be falling behind to the extent the article claims. Plus we have the same statistical problems already mentioned.

Truth said...

"but that describes more or less EXACTLY the childhood & early adulthood of IQ 130+ kids."

Not in the dimension I live in; the 3rd.

Anonymous said...

Two quick points:
1- Dumbing down the schools has had a horrendous effect on average students - whatever their race - whose parents can't afford to live in the areas with better schools or send them to private schools. The quality of their education is lower than it would have been a generation or two ago. At the same time this dumbing down didn't help the students who can't do grade level work. The only people whom it seems to have "helped" is the professional liberals who worry about racial learning gaps while sending their own children to private school.

2- I lived in the DC area for years and met a number of well-educated and smart black women. Most of them - probably 75% or more - had one child or none. Their less educated and less intelligent "sisters" usually had two or more. I think this might explain some of the gap and why our attempts to close it have largely failed.

DAJ said...

A train one mile in length is traveling 60 mph. How long does it take to completely pass through a tunnel one mile in length?" 100% of iSteve readers can answer this one off the top of their heads, but 100% of NAMs will miss it, unless it's multiple choice.

The contempt for NAMs displayed in the above quote explains the failure of HBD to go mainstream. Since when does HBD mean "100% of blacks and Hispanics are very stupid?" Whatever happened to the concept of a bell curve? Whatever happened to the phrase, "Yes, blacks and Hispanics are on average considerably less intelligent than whites/NE Asians, but a few are indeed smart?"

Anonymous said...

Good point about middle school being a waste of time for smart students. I didn't really learn anything except basic algebra. Grades 5 and 6 were especially pointless. I had to sit in class with people who were still learning long division and the difference between nouns and verbs. This was in an upper middle class, 85% white school district in one of the wealthiest states where most of the fathers were doctors, lawyers and stockbrokers.

By grade 7, I started cutting school and hanging out at the local thug infested video arcade/drug market. That scene was way more interesting than anything I was doing in school. Finally, in high school they introduced tracking and was able to take AP classes where I learned things I didn't already know. I could have finished HS with the same grades a couple of years sooner had I been given higher level material earlier though.

In the 6th grade, we had to take one of those "achievement tests." I scored in the 99th percentile and the scorecard said that my "grade level" was equal to a sophomore in college. I remember thinking "if that's even close to being accurate, what the HELL am I doing here?"

Anonymous said...

Not in the dimension I live in; the 3rd.


0% indeed...

Anonymous said...

"Great, then why waste resources sending them to school. Teach 'em to read by age five, lock them in their rooms with a stack of books 6 hours a day...presto, 20 years later brain surgeons."

I think you're trying to be sarcastic, but that describes more or less EXACTLY the childhood & early adulthood of IQ 130+ kids.

I agree, or rather, what their childhood should be like.

Anonymous said...

In UK, the illegitimacy rate for whites is 40%
This sub-culture is also called chav culture
The white teenagers in the Chav class in UK behave like black teenagers in USA
Chav teenager girls deliberately get pregnant, so that they are eligible for a new apartment

A very high number of chav girls get pregnant by blacks

kudzu bob said...

"Not in the dimension I live in; the 3rd."

Ah, so this explains the one-dimensionality of Twoof's posts...

Felix said...

As someone stated, “White kids [especially the ones in the public schools] are so poorly educated nowadays that it's simply flabbergasting".

Specifically, most kids suffer from poor teaching methods in math. Which means that they are closing the gap by moving to an equally poor level of performance.

BTW, it’s been argued above that, as you get out towards IQ 130+, the kids become capable of teaching themselves. True but, more significantly, middle class parents intervene and ensure that the kids learn the material, teachers notwithstanding.

(As someone in this category, I was initially alienated at 12-13 by math made “relevant”. But, after failing spectacularly, my mom got me tutored in the summer holidays and after that I could teach myself.)

none of the above said...

Nitpick: The bulk of students isn't between IQ 100-120, if we're talking about the bell curve with mean of 100.

Is there evidence of dysgenic effects being greater for blacks than for whites in the last 50 years? (I wonder how the children of interracial couples affect this.)