The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
Email This Page to Someone

 Remember My Information

 TeasersiSteve Blog

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New Reply
🔊 Listen RSS

Long time readers know I’ve been interested in the question of school test scores in the two biggest states, California and Texas. In the federal National Assessment of Educational Progress scores, Texas routinely beats California across all racial groups. But the NAEP is low stakes to students, which makes it easier for state officials to manipulate results at the margins.

However, looking at an unverified table of high-stakes SAT and ACT college admission average test scores for 2014, white, Hispanic, and black California high schoolers outscore their counterparts in Texas (using a weighted average of SAT and ACT scores). But Texas’s Asians outscore California’s Asians.

All 350,655 1,016 295,583 973 43
AmInd 1,814 982 1,501 992 (9)
Asian 66,385 1,108 18,569 1,126 (18)
Black 20,667 888 37,615 854 33
Hispanic 131,723 905 113,395 891 14
Other 28,357 1,065 12,961 1,003 61
White 101,709 1,113 111,542 1,069 44

Both states are moderately majority SAT: in California, SAT takers outnumber ACT takers 2.1 to 1, and in Texas 1.5 to 1. This appears to be putting everything on the traditional 400 to 1600 scale, rather than the 600 to 2400 scale of the last decade, but that is being phased out soon. The mean was rescaled in 1995 to, ideally, be 1000 with a standard deviation of 200, although both have drifted since then.

So, California’s overall average is 97 points, or a little under a half of a standard deviation below it’s white average, while Texas’s overall average is 96 points below it’s white average.

I’m not going to put too much credence in these numbers: even if the data are valid (which I haven’t checked), my weighted average methodology is crude. On the other hand, the results don’t seem too implausible.

I mostly want to put some numbers out there to provoke somebody interested in this long-running problem of how to synthesize SAT and ACT scores reliably to try to come up with a more sophisticated general model.

🔊 Listen RSS

One of the older, more nagging conundrums for anybody interested in education and demographics is the lack of readily available meaningful data on how high school students do by state and by race on high stakes tests such as the SAT and ACT college admissions tests.

The federal government invests a lot of money in the NAEP test, but that is a low stakes test for students, so it’s more easily manipulable by those states that care about the results. For example, Texas usually manages to have a larger percentage of its less academically inclined students not take the NAEP than does Iowa, which helps contribute to Texas’s sterling NAEP scores.

Or maybe Texas really has figured out an effective, economical system of educating students of all ethnic groups. It’s hard to say, but it’s an important question that deserves study.

A high stakes test, in contrast, is one in which students have motivations for doing their best, which is why I’ve always wanted to look at SAT and ACT scores by state. After all, the NAEP isn’t important in the big picture, while the SAT and ACT are.

But, the percent of 17-year-olds taking one or both college admissions tests vary by state. This, however, is not an insuperable problem since estimates of what nontakers might have scored can be modeled demographically by looking at the variation in usage rates.

Another difficult problem, but one I believe can be modeled, is that the two tests started out regionally, with the ACT dominating states near its headquarters in Iowa City and the SAT near its headquarters in Princeton and on the West Coast.

In the upper Midwest, traditionally, the only students who took the SAT were ambitious one looking for admission to national universities on the East or West coasts. This led to Iowa and Illinois students taking the SAT averaging much higher scores than in the East and West.

In recent years, both tests have become less regional, with ACT-taking spreading to the coasts.

That evolution should help an ambitious analyst come up with a reasonable model for estimating the best guess for the combined SAT/ACT scores by state by race.

An iSteve reader (whose identity I have lost in the shuffle) kindly posted average SAT and ACT scores and number taking by state by race each year from 2006-2014 here. He converted the ACT scores into SAT score equivalents, although I don’t know which methodology he used.

Combine this trove of data with the 2010 Census data on the number of 17 year olds by race in each state and you have the raw materials for building a model that will get around the traditional problems that have bogged everybody down.

Me, personally, I’m not going to do all this work, but if somebody out there has the skills and is looking for a topic, this is an important one.

I don’t have the sources for this data, but if you are interested in working with this, post questions in the comments and the person who posted the numbers might respond.

🔊 Listen RSS
Texas public school students usually score pretty well in the federal government’s NAEP school achievement tests, at least when adjusted for ethnicity. I’ve always wondered how they do it. It would seem like the kind of thing worth checking into.
One way, it turns out, is by excluding more students from having to take the NAEP than other states do. Texas excuses 10% of its 4th graders versus 4% nationwide and only 3% in California. (See p. 5 of this new report on the NAEP performance of the 5 biggest states.) So, Texas has simply made a large fraction of Below Basic scorers vanish. That’s a nice little running start for Texas.
If Texas has figured out how to fiddle with that parameter, I wonder what else they’ve figured out?
(Republished from iSteve by permission of author or representative)
🔊 Listen RSS
The feds’ National Assessment of Educational Progress has a table of 4th and 8th grade vocabulary and reading comprehension scores by state. Sample size issues are of concern for smaller states which tend to bounce around, but we can state with a high degree of statistical confidence that the future of the state of California, the traditional State of the Future, looks dumb. Out of the 50 states, the Golden State ranks 48th, 47th, 48th, and 49th on various measures. Here’s the bottom six of 52 in the four different tests:

In contrast, Massachusetts is 1st, 1st, 1st, and 1st, while the District of Columbia was 52nd, 52nd, 52nd, and 52nd (in case you are wondering why D.C. is the 52nd state, Department of Defense schools rank 2nd, 5th, 2nd, 6th). Obviously, the problem is all those Republicans in California and D.C. If only D.C. would develop enlightened political opinions like Massachusetts, its test scores would soar.

Perhaps more relevantly, Texas is 37th, 36th, 37th, and 36th. Texas always beats California on the NAEP. Has anybody studied this to make sure this is not just a test artifact (e.g., Texas cares about the NAEP and California doesn’t)? If it isn’t, why the consistent difference? Texas is pretty bad, but it’s not as bad as California, and beggars can’t be choosers, so somebody ought to be investigating why Texas beats California.

One obvious objection is that the future isn’t as bad as it looks because Hispanics, as new immigrants, are just being held back by the inevitable biases of testing skills in English.

Indeed, this effect does exist, but how big is it? Here’s national 8th grade vocabulary. The first number is score at the 10th percentile, then 25th, 50th, 75th and 90th.

Let’s first compare whites and Asians. At the 10th percentile, Asians lag whites by 8 points. Presumably, a fair number of these Asian 8th graders just got off the plane from China, so their English vocabulary is limited. At the 25th percentile, the White-Asian gap is down to 5 points. At the median, it’s 3, at the 75th percentile it’s 0, and at the 90th percentile, Asians are out in the lead by a point.
Now, compare Hispanics to blacks, most of whom grow up speaking English, but as we all know from hundreds of articles, African-Americans grow up in conditions that would drive a Trappist Monk crazy for lack of speech. In black homes, nobody every talks, watches TV, or listens to rap music. So, black scores on language are bad, with unfortunate long-term consequences.
At the 10th percentile, where many of the Hispanics are newcomers, blacks lead by 2 points. At the 25th percentile, however, Hispanics are out in front by 1 point, by 2 at the median, 3 at the 75 percentile, and 4 at the 90th.
So, clearly, Hispanics who have all the advantages are, on average, a little smarter than blacks who have all the advantages. In other words, if immigration were shut off for a generation or two, Mexicans would appear, on average, perceptibly more on the ball academically than blacks. Indeed, that was my perception back in the 1970s in L.A., where the Chicanos had mostly been a stable population since WWII.
But, nationally, Hispanics only pick up 6 points on blacks going from the 10th to the 90th percentiles, while Asians pick up 9 points on whites, who are, to be frank, a lot more competition.
Being a little smarter than blacks is, well, good. Or, you could say with equal justice, less bad. On the other hand, Hispanics at the 90th percentile among Hispanics, typically those with all the advantages, are simply not playing in the same league as Asians and whites with all the advantages. They’re down there beating out blacks for third place, not being nationally competitive. There’s not a lot of high end in the Hispanic population.
However you look at it, it’s still not very encouraging considering that our leadership kind of bet the country on Hispanics.
(Republished from iSteve by permission of author or representative)
🔊 Listen RSS

by State, 1960 –
You probably remember the notorious “Democratic
states have higher IQs
” hoax from last May. Well, here, thanks
to Prof. Henry Harpending of the U. of Utah anthropology dept., might be
the closest thing to a national sample of IQ scores ever: the Project
Talent database of 366,000 9th-12th grade students. Unfortunately, it is
44 years years old. Nonetheless, it correlates reasonably with 2003 NAEP
8th grade achievement test scores (here
are the 2003 scores). As you can see, in this list of kids’ IQs back in
the mid-1960s, of the top 10 smartest states, in 2000, Bush and Gore
each won five. So, we’re back to my original conclusion: red states and
blue states are similar in average IQ, as are, on average, Republican
and Democratic voters.

caveats: These IQ scores are set with the national mean of the 366,000
high school students equal to 100 and the standard deviation set to 15.
But, keep in mind that we are only beginning to explore this huge
database, so take everything with a grain of salt.


New Hampshire










New Jersey

New York




North Dakota

















New Mexico






Rhode Island












West Virginia






North Carolina








weren’t adequate sample sizes from Alaska, Washington DC, and South
Carolina, and I excluded South Dakota because the result was too
different from North Dakota. (I think something might be confused about
both South Carolina and South Dakota — I’ll try to find out more.)

also looked at whites only data (unfortunately, the majority of
participants doesn’t have a race recorded) with the smartest whites
(which I suspect is all that white liberals care about — feeling
smarter than white conservatives) were (in descending order):
Connecticut, Montana, Nevada (I bet that’s not true anymore!), Idaho,
Illinois, Maryland, Massachusetts, Minnesota, Missouri, New Hampshire,
New Jersey, New York, Oregon, and Virginia. The dumbest whites were in
(in descending order): Georgia, Louisiana, Mississippi, North Carolina,
Arkansas, Tennessee, West Virginia, and Kentucky. All of these states
voted for Bush in 2000. I suspect, however, that air conditioning and
the abolition of the caste system have some good for the test scores of
whites in the south, especially in North Carolina. Here,
for purposes of comparison, is the 2003 NAEP public school achievement
tests for white 8th graders.

(Republished from iSteve by permission of author or representative)
Steve Sailer
About Steve Sailer

Steve Sailer is a journalist, movie critic for Taki's Magazine, columnist, and founder of the Human Biodiversity discussion group for top scientists and public intellectuals.

The “war hero” candidate buried information about POWs left behind in Vietnam.
The evidence is clear — but often ignored
The unspoken statistical reality of urban crime over the last quarter century.
The major media overlooked Communist spies and Madoff’s fraud. What are they missing today?
What Was John McCain's True Wartime Record in Vietnam?