2006 July 15 Saturday
Private Schools Show No Advantage

The US Department of Education's National Center for Educational Statistics (NCES) has released a study (carried out under contract by the Educational Testing Service) which finds little advantage of private schools as compared to public schools when the family backgrounds of the kids are adjusted for.

In grades 4 and 8 for both reading and mathematics, students in private schools achieved at higher levels than students in public schools. The average difference in school means ranged from almost 8 points for grade 4 mathematics, to about 18 points for grade 8 reading. The average differences were all statistically significant. Adjusting the comparisons for student characteristics resulted in reductions in all four average differences of approximately 11 to 14 points. Based on adjusted school means, the average for public schools was significantly higher than the average for private schools for grade 4 mathematics, while the average for private schools was significantly higher than the average for public schools for grade 8 reading. The average differences in adjusted school means for both grade 4 reading and grade 8 mathematics were not significantly different from zero.

Comparisons were also carried out with subsets of private schools categorized by sectarian affiliation. After adjusting for student characteristics, raw score average differences were reduced by about 11 to 15 points. In grade 4, Catholic and Lutheran schools were each compared to public schools. For both reading and mathematics, the results were generally similar to those based on all private schools. In grade 8, Catholic, Lutheran, and Conservative Christian schools were each compared to public schools. For Catholic and Lutheran schools for both reading and mathematics, the results were again similar to those based on all private schools. For Conservative Christian schools, the average adjusted school mean in reading was not significantly different from that of public schools. In mathematics, the average adjusted school mean for Conservative Christian schools was significantly lower than that of public schools.

Keep in mind that the kids who go to private schools do score higher than the kids who go to public schools. The adjustment they do for student background amounts to an adjustment for genetic influences. But they aren't going to state the obvious: Kids who go to private schools are genetically smarter on average. In order to substantially raise the performance of public school students we'd need to raise the genetic endowment of the average child born.

To properly test the efficacy schools would require widescale IQ testing. Then comparisons could show how much each school accomplishes with the raw intellectual potential it has to work with in the kids it receives. But the political Left in America rejects IQ testing because it shows racial average differences in intelligence. Their continued defense of their secular faith requires undermining social science by preventing the measurement of the most important variables. So social science becomes quackery. What a tremendous waste.

The authors of the study inject plenty of qualifiers and caveats into its interpretation. They realize that plenty of people on both side of the public/private school debate have a lot invested in defending their positions.

When interpreting the results from any of these analyses, it should be borne in mind that private schools constitute a heterogeneous category and may differ from one another as much as they differ from public schools. Public schools also constitute a heterogeneous category. Consequently, an overall comparison of the two types of schools is of modest utility. The more focused comparisons conducted as part of this study may be of greater value. However, interpretations of the results should take into account the variability due to the relatively small sizes of the samples drawn from each category of private school, as well as the possible bias introduced by the differential participation rates across private school categories.

There are a number of other caveats. First, the conclusions pertain to national estimates. Results based on a survey of schools in a particular jurisdiction may differ. Second, the data are obtained from an observational study rather than a randomized experiment, so the estimated effects should not be interpreted in terms of causal relationships. In particular, private schools are “schools of choice.” Without further information, such as measures of prior achievement, there is no way to determine how patterns of self-selection may have affected the estimates presented. That is, the estimates of the average difference in school mean scores are confounded with average differences in the student populations, which are not fully captured by the selected student characteristics employed in this analysis.

We hear a great deal about how schools do not have enough money. Historical comparisons are useful when evaluating this claim. See the NCES page Total and current expenditure per pupil in public elementary and secondary schools: Selected years, 1919-20 to 2001-02.In the 30 years from 1971 to 2001 the total expenditures per student in inflation adjusted dollars doubled from $4884 to $9614. Going back even further the expenditures tripled from 1963's $3228. This availed us of little improvement in outcomes. But faith springs eternal. The demand for more money for education continues unabated.

The comparison was done using NAEP scores and controlled for many factors including factors that effectively serve as proxies for genetic differences - not that they'd ever admit this.

This study compares mean 2003 National Assessment of Educational Progress (NAEP) reading and mathematics scores of public and private schools in 4th and 8th grades, statistically controlling for individual student characteristics (such as gender, race/ethnicity, disability status, identification as an English language learner) and school characteristics (such as school size, location, and the composition of the student body).

A lot of those factors are rough proxies for intelligence.

My guess is that the spending per student in public schools has risen so far that public schools, even if they accomplish less per dollar spent, still spend so much that they are accomplishing close to the best we can hope for. The real educational crisis is that the best we can hope for is not all that much. The average IQ in America is declining due to immigration of low IQ groups. Trying to turn all these kids into college bound students represents the triumph of faith over the evidence of our lying eyes.

With low IQ students we should shift more toward development of job skills that let them perform specific lower skilled jobs. They can have some capacity for learning but require a fair amount of repetition of tasks that they can realistically hope to master.

Schools could be customized for the abilities of the smarter teen kids by providing them with access to recordings of college level lectures and with easier ways to take tests (think internet) for college level material. Let them learn at their own faster pace rather than make them sit in courses that advance at the rate needed for the below average student.

A companion study on charter schools has not yet been released.

The study, along with one of charter schools, was commissioned by the former head of the national Center for Education Statistics, Robert Lerner, an appointee of President Bush, at a time preliminary data suggested that charter schools, which are given public money but are run by private groups, fared no better at educating children than traditional public schools.

Proponents of charter schools had said the data did not take into account the predominance of children in their schools who had already had problems in neighborhood schools.

The two new studies put test scores in context by studying the children’s backgrounds and taking into account factors like race, ethnicity, income and parents’ educational backgrounds to make the comparisons more meaningful. The extended study of charter schools has not been released.

Those who know that environment can not trump genes have to wait for decades as various "solutions" such as more money (and the average spent per student is almost double what the public believes), charter schools, vouchers for private schools, more extensive testing, closing of so-called "failed schools", and still other methods get tried and do not help much. At some point advances in neurobiology combined with the continued decline in DNA sequencing costs will shatter the conventional wisdom of our elites and the truth will enter the public debate on education and social policy.

Share |      By Randall Parker at 2006 July 15 06:50 PM  Education

Half Sigma said at July 15, 2006 9:20 PM:

The purpose of sending kids to private schools isn't so they get higher scores on standardized reading and math tests (which mostly measure g).

People send kids to religious schools in order to indoctrinate them in the religion.

And non-sectarian private schools provide kids with a proper upper class environment that will help them in subtle ways later in life.

John S Bolton said at July 16, 2006 2:27 AM:

With 20% or higher non-genetic influence on IQ, there would seem to be quite enough room for people to develop some expectations for private education.
If their idea were to get disadvantaged minorities into private schools, expecting the major racial divergences to equalize, that would go against the evidence from transracial adoption studies.
I like to promote private education, not because it can do much for the genetically disadvantaged, but because of the bad political and culture-deforming effects of public schools.
There is every tendency built into them to become almost nothing but propaganda organs for despotism.
It is the enslaving education, in that it wants to liberate aggression, especially that of officials.

D Flinchum said at July 16, 2006 4:56 AM:

"it should be borne in mind that private schools constitute a heterogeneous category and may differ from one another as much as they differ from public schools."

And that can be a big difference, indeed. There are private schools that are quite selective, that are as demanding as most universities, and that usually cost per year in the neighborhood of a good four-year college. There are others that have been instituted mainly, it seems, to protect students from the "unpleasant" truths of science and history and the unacceptable topics of literature. I suspect that there are more of the latter than the former. My guess is that this study is a "feel good" nod toward public schools that has very little meaning beyond that.

I know a guy who has lived in both Montgomery County Maryland, where certain of the schools are considered among the best, and in DC, where most schools are among the worst in the US. In both places his only child attended high quality private schools. This fellow isn't rich and he definitely isn't stupid. He and his wife are paying big bucks for a reason.

Omer K said at July 16, 2006 5:01 AM:

It is my view that it is far easier to make High IQ types perform lower than their capabilities than Low IQ groups t exceed their genetic limitations.

Private schools, vouchers etc have the effect of allowing incentives and demands into the education market.
Lower IQ groups wouldnt have to waste 12 years when they reach their limit around 8, freeing 4 years of their lives for themselves.

Higher IQ groups would not be forced to waste 12 years to learn what they are capable of learning in 6. If you can hypothesize that the human brain has a window of time where it is optimally geared to learn, aka what one can pick up (at a given iq) at 18 years of age, can be picked up 30% faster at 14 years of age... then the effects on High Iq groups especially will be one of great liberation.

Randall Parker said at July 16, 2006 6:49 AM:

Omer K,

Fluid intelligence rises until age 16. It is age normalized. So a kid with an IQ of 125 at age 10 is not as smart as the same kid with IQ measured again at 125 at age 16. So you can't teach some kinds of subjects to younger kids that they can handle when they get older.

My argument is that as kids approach their peak intelligence that is the time to accelerate their learning. Hence I argue for video college lectures for teens to watch.

Omer K said at July 16, 2006 6:54 AM:

Probably right. I didnt consider that.
But contrast that with the fact that you can teach a child two, three maybe even five languages up to age 5, but after that age that ability rapidly declines.

Omer K said at July 16, 2006 7:05 AM:

Sorry for the double post.
But on reflection even if my numbers were wrong, my point still stands. Now the magic number is 16 after which decline occurs. Some people study till the age of 25 or even 30, a mild decline should have taken place by then, and allowing them to start learning the meatier stuff earlier, if they have the capacity, can only be of benefit to themselves and auxilliarily to society at large.

gcochran said at July 16, 2006 11:00 AM:

Some people assume that the non-genetic influences on IQ are things like parental SES, going to the right school, number of books in the house, etc.

Mostly they're not. We don't know exactly what they are, but we know that they're not between-family: so they're not those social factors.

They're within-family. So is there reason to think that those expensive private schools do much of anything? No. In fact we know they don't.

So why do people spend those big bucks? Don't they know what they're doing?


Carl Shulman said at July 16, 2006 1:03 PM:

Non-heritable variation can still be biological, e.g. iodine deficiency or mercury poisoning, viruses, and the developmental noise of a protein zigging instead of zagging. Moreover, heritable variation can be non-genetic (if alcoholics never quit and always began drinking at a stable level before their first pregnancy then the developmental effects of alcoholism would show an extremely high heritability).

Think about sexual orientation, which is not very heritable but, despite the claims of those with a religious axe to grind, is clearly not heavily influenced by the social environment. One partial cause seems to be immune system changes induced by previous male children:


Because this effect increases with the number of sons, it explains both a) part of the 'heritable variation,' as 3rd and 4th sons will both have an increased likelihood of being homosexual, establishing a correlation tied to sharing that biological mother and b) part of the non-heritable variation, as the strength of the effect increases with each child.

However, schools can deliver training in specific knowledge and skills even if they cannot increase (often more valuable) g, and there is undoubtedly room for improvement in that area with a more realistic view of the data. I would also note that the teachers' unions get a lot of mileage out of claims that differences in performance by race solely reflect 'bad schools,' i.e. schools where teachers are not paid enough as opposed to schools with bad students. If such rhetoric helps to inflate education costs by maintaining public provision (vs vouchers or privatization) and strengthens socialist-leaning teachers' unions clout in influencing elections results towards irrational economic policy generally, then confusion in the area of student ability is triply pernicious.

gcochran said at July 16, 2006 2:10 PM:

It's not easy to transmit more specific knowledge than we do using current methods: most people have lousy memories for things that don't interest them (and fairly lousy memories for things that do). For example, if you look at a survey of general science knowledge in adults, taking high school science courses has no effect at all. Interestingly, college science courses do.
As for sexual orientation, what you say is true for males but I wouldn't be too sure about females.

This issue gives both liberals and conservativs a chance to demonstrate their foolishness. And they take full advantage of it.

Randall Parker said at July 16, 2006 3:24 PM:


I wonder if memory retention could be improved by retesting. If high school kids took tests of knowledge every month or two that covered a wide range of knowledge whose retention is desired then might they remember more X years later?

I've read claims that memories are strengthened by recall. Could testing software automate the process of longer lasting memory formation?

John S Bolton said at July 16, 2006 3:27 PM:

I thought the environmental portion of variation in IQ was affected by peer influence, and to a large enough extent that it would make sense to pay up to get away from bad influences of that kind.
Some schools teach reading by an anti-phonics approach; responsible parents have to make up for that somehow.
It may have little or no effect on g, to be what public school teachers find altogether too conveneient to call dyslexic; but parents rightly fear such a result.
Another indication that parents who pay heavily for what they consider better schooling, and better peer environment, are doing as they should; is the way the left and moderate right delight in giving admissions advantage to the top X% of every high school in a jurisdiction.
Would destructive malice such as is found in such relation, miss its targets so widely, as to actually help parents by thus dissuading them from wasting their money or time on schools of merely superior prestige?

Carl Shulman said at July 16, 2006 4:18 PM:

Yes, that study was with respect to male homosexuality, and female homosexuality is quite different. For one thing, there seems to be a more continuous spectrum between heterosexuality, bisexuality, and homosexuality among women, while male sexuality seems to be fairly strictly bimodal:


On memory retention, the type of things I had in mind were along the lines of basic literacy and numeracy or occupational training, i.e. life skills that will be constantly exercised if possessed. They are also skills that can be the difference between being employed or not at the margin, and thus worth investing in for low-ability people. People forget science because it is abstract knowledge and there is no incentive for them to exercise it in their daily lives or refamiliarize themselves with it. This is a big problem for a democracy relying on them to vote or serve on juries, but it is much less pronounced for the teaching of practical employment-related skills.

John S Bolton said at July 16, 2006 4:28 PM:

No doubt we could use further clarification of the relative strength of genetic and non-genetic influence on IQ and other results.
I propose that this research be done: through genetic testing of samples from military personnel, find the signature of a sperm donor with thousands or tens of thousands of children, whatever the upper limit may be, and correlate with IQ scores, etc.
This should yield another pure test of genetic influence through the half-siblings shared genes, but unshared environment relative to the reference population averages.

beowulf said at July 16, 2006 8:52 PM:

To crosspost a very wise commentor over at tpmcafe (OK, OK it was me) in response to Matt Yglesias thought that there was nothing to be done to increase test scores---

Its all politics and its all a waste of time, except for one thing.

The first person who figured out the score was the late Benjamin Bloom, who published a research article 22 years ago titled "The Two Sigma Problem: The Search for Methods of Group Instruction as Effective as One-to-One Tutoring". Sadly its not online but if you google Two Sigma Problem, there are summaries here and there. Here's a home schooler who's collected some Bloom links -- http://lionesshomeschool.blogspot.com/2006/05/golden- quote.html

The average student (being average) scores at the 50th percentile on achievement test. But students who are tutored full-time (like, say, child actors are required to be) scored in the 98th percentile-- Two Sigma above the average.

The beginning and end of education reform is simply this-- creating a system that brings students in school at least in the same ballpark as children who are tutored.

It will likely involved computerized learning (so every student learns at their own rate, just as tutored children do) and will use the Bloom's Mastery Learning or the similiar Keller Plan of teaching subject in modular blocks that each student must pass a quiz on before going on to the next section. The military has been working on the subject for years-- here's a power point presentation from a few years ago -- http://tinyurl.com/ewrhb
/crosspost be gone -- of course, its not the "average" kids who are the problem, its the kids who are on the left side of the bell curve that are lost. Bloom's whole theory is that currently, schools have a fixed amount of time to teach a class, the very smartest students nail it before the end of the semester and indeed may get bored waiting for the rest of the class, the average kids do OK (better than OK with grade inflation) and the slow pokes fall overboard and the boat's not coming back to pick them up. The advantage of individualized learning is that the students take as long as necessary to get through each lesson module, so if a kid masters the class by midsemester, he can test out and move on. As for the slow pokes, they may not be able to learn it all in one semester, so they'll go at their own pace until they master it.

Ideally, we should also switch to a year round trimester system. That will give the low intelligence students 33% more time to master the K12 curriculum by the time they're 18 (and the military has found that just switching to an individualized teaching system reduces the necessary learning time by 30%) As for the bright students, they could go on and finish a bachelors or even a masters degree by their senior year of high school-- without a penny of student loans.

gcochran said at July 18, 2006 9:14 PM:

" The average student (being average) scores at the 50th percentile on achievement test. But students who are tutored full-time (like, say, child actors are required to be) scored in the 98th percentile-- Two Sigma above the average. "

I guess Charlie Sheen is hiding his light under a bushel. That's sarcasm: what I'm really saying is that I'd bet serious money that this statement isn't true. If it _were_ true, simply noticing ths fact would let the scions of the rich and powerful would score at the 98th percentile or higher if they wanted to - yet they don't.

Since shared family environments, which are pretty damn close to one-on-one, have almost no effect on adult 'g', I dinna believe it.

crush41 said at July 21, 2006 12:10 AM:

The charter school results will be worth taking a look at. Private/public comparisons that control for SES are inherently flawed because wealthier (and by extension, more intelligent) people send their children to private schools. But these children will, on average, 'suffer' some regression toward the mean. The IQ of parents serves as a rough proxy for the IQ of their children, but if that IQ is above average, the proxy needs to be adjusted downward. The flip, of course, if the IQ is below average. In SES-adjusted comparisons, public schools artificially look better than they should and private schools look worse.

As you mentioned, widespread IQ testing would compensate for this (in addition to solving a host of other questions in the arena of education).

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Web parapundit.com
Go Read More Posts On ParaPundit
Site Traffic Info
The contents of this site are copyright ©