August 28, 2008

The Union's Value-Add

Justin Katz

Congratulations to the National Education Association's Pat Crowley for managing to push his story about Governor Carcieri's Florida condos onto (astonishingly) the front page of the Providence Journal, which used it as a contextual gotcha against the backdrop of the union healthcare story. (Gee, I didn't realize that the governor is rich!)

Normally, I wouldn't have considered this feat worth mentioning, except for a "meanwhile" education story relegated to the Rhode Island section:

College-bound Rhode Island students performed about as well this year as last in the SAT, AP and PSAT tests, it was announced Tuesday by the College Board.

Seniors scored 495 in critical reading on the SAT test, down 1 point from last year; 498 in mathematics, unchanged; and 493 in writing, up 1 point. About 66 percent of seniors took the exam. The scale for scoring ranges from 200 to 800.

Rhode Island students fell below the national average, which is 502 in critical reading, 515 in mathematics, and 494 in writing, the College Board said.

Rhode Island did not fare well when compared with nearby states. Figures from the Associated Press showed that Connecticut’s students scored an average of 507 in math, 506 in writing, and 503 in critical reading. Students in New Hampshire averaged 523 in math and 502 in critical reading. The score for writing was not available.

In Vermont, students scored 519 in critical reading, 523 in math and 507 in writing. The figures for Massachusetts were 525 in math, 514 in critical reading, and 513 in writing.

Scores for Rhode Island seniors who attend public schools were 483 in critical reading, unchanged from last year; 487 in mathematics, down 2 points; and 479 in writing, up 1 point. About 59 percent of the public school seniors in the state took the exam, the 13th-highest participation rate among the states. Figures for private schools were not available.

The College Board said that after two years of declining scores for public school students, the 2008 national scores — 497 in reading, 510 in mathematics, and 488 in writing — stood unchanged. Rhode Island scores, which declined only 1 point in the aggregate, have mirrored that national trend.

In summary, Rhode Island is below the national average in every subject, and our students are even further behind students in the two states by which we are surrounded. Plumbing the data for the public-school/overall distinction for our neighbors, the tale is even more bleak for RI families that can't secure freedom from RI's government/union schools:

So, not only do Rhode Islanders in general perform more poorly on the SATs, but there's a much greater discrepancy between public and private school students, here. Yes, congratulations are certainly due to Patrick Crowley and the NEARI!


I was actually holding off on a trail of analysis that I'd begun because there are more subtle points to be made, and I wanted pondering time, but since Brassband points out in the comments that the above chart shows public school and overall SAT scores (i.e., public school students are included in both), here's a chart comparing public with private schools:

Apart from the fact that there are relatively few non-religious private schools in Rhode Island, one reason I held off — and the reason I break out religious schools separately even though they are also included in the "all private" column — is that they seem more likely to be a refuge for the average family fleeing the public school system. Note that RI religious schools do better than religious schools in either of the other states.


In the comments, Thomas Schmeling calls the two visuals above "gee whiz graphs" because I zoomed in on a score range. Frankly, that sort of point, when made in isolation, strikes me as rhetorical sleight of hand. Examining the axes to discern what's being shown is a critical step in reading any graph, but deciding whether it is deceptive requires consideration; it isn't a given. Schmeling illustrates the validity of my rejoinder by failing to make any argument about whether the tighter distinctions are merited, just implying that they are not.

So here's the second graph starting from zero:

Except to the degree that it is more difficult to read accurately, that doesn't cool my response any, because (probably in common with most Anchor Rising readers) I learned from self-interested experience how to read and compare SAT scores. We all look to the labels, approximate a 200-point difference, and know that to be large (although those of us who took the two-subject version might have a skew). Indeed, Schmeling points out the percentage difference of the public and private categories, but the SATs aren't graded by percentage; they're graded by point, and comparisons of results in practical situations often come down to 50 points per test. People pay for test-specific training to achieve less advantage than that.

Schmeling's second point — that controlling for socioeconomic status makes the difference go away — is even more apt to elide the very point under discussion. Firstly, that private schools are able to educate students at a lower cost per student (and a much lower cost per teacher, to my experience) suggests to me that even an equivalent score brings into question the government/union model.

Secondly, consider the method of calculating "socioeconomic status" used in one study that Schmeling cites:

She determined students’ socioeconomic status (SES) by looking at six factors students were surveyed on in the NAEP assessment: eligibility for free or reduced lunch, eligibility for Title 1 funding, reading material in the student’s home, computer access at home, Internet access at home, and the extent to which a student’s studies are discussed at home.

Four of those six measures relate to parental motivation, not to household wealth. My family, for example, sacrifices hugely in order to claim items three, four, and five (and now to afford private school). And although I haven't seen any studies to this effect, it seems reasonable to me to suggest that the presence of such children has an effect on their peers' performance. If, as I implied above, motivated working- and middle-class families like mine are going that extra sacrificial mile to escape an unsatisfactory public school system, then those children who are left behind have fewer models for better academic behavior among their classmates.

Comments, although monitored, are not necessarily representative of the views Anchor Rising's contributors or approved by them. We reserve the right to delete or modify comments for any reason.

I have done much research on the issue and I have come to a reasonable conclusion that has nothing to do with the unions.

Rhode Islanders are friggin idiots.

Stop for a moment. Look 360 degrees around you. How many of those people do you think are smart enough to come in out of the rain? Nuff said.

Posted by: Greg at August 28, 2008 12:32 PM

Justin -

Anchor Rising has reached a new low when you are defending an elected official who hasn't paid over $12,000 in property taxes or possibly not even disclosed it on required financial disclosure forms.

I guess "open and clean government" are only political attack points for you instead of a central tenet of your political beliefs.

For me, our elected officials need to lead by example. And this is simply inexcusable.


Our Governor should be ashamed.

Posted by: Matt Jerzyk at August 28, 2008 1:27 PM

I would encourage anyone with a 6th or 7th grader to follow the link and read the numbers behind the bar graphs.

The bar graph makes it look as if "non-public" students do about ten points better than their public school colleagues.

Not true. Remember that these are averages, and the private school kids only account for about one-quarter of the total test-takers.

The numbers show that RI high school kids in religious (probably mostly Catholic) and "independent" high schools do about 50 to 100 points better on these tests (depending on the category . . . again, follow the link and look for yourself).

Posted by: brassband at August 28, 2008 1:30 PM

Holy crap! Brassband is right. The scores for non-government schools are much, much higher than the graph illustrates. What a ***king waste of money our public schools are. Thank you very much NEA and AFT!

Posted by: Frank at August 28, 2008 3:31 PM

Matt J, keep us informed on those disclosure forms.

In the meantime, unlike that of the Senate President who failed to disclose five years of [public] income he earned from a municipality which wanted his [public, official] blessing for a lucrative casino, the Governor's situation did not involve public property, public monies or his official capacity.

You'll forgive us, then, if we don't get too excited about a misdirected real estate tax bill and focus instead on situations which involve abuse of the aforementioned public monies and official power.

Posted by: Monique at August 28, 2008 3:50 PM

Justin --

Thanks for confirming my read of the statistics (hey, I went to public school!).

Anyway, I don't mean to say that the public schools are worthless; the point is that for parents who can afford a choice, the differences are worthy of consideration in making that choice.

Someday maybe all parents will be able to partake in that kind of choice.

Posted by: brassband at August 28, 2008 4:01 PM

Geeze Justin, don't you understand the schools need more money... and the greedy rich need to pay more taxes.

The Journal and RI Future make a great tag team grandstanding the progressive propaganda while burying and/or evading the truth.

Posted by: George at August 28, 2008 4:19 PM

Don Carcieri believes in the Leona Helmsley theory of taxation.

Posted by: rhody at August 28, 2008 4:37 PM

Can you compare Senator Obama's SAT scores to John McCain's
Obama 1567
McCain 945

Mrs Obama 1562
Mrs McCain (unknown)

Posted by: Roodie at August 28, 2008 4:53 PM

Jerzyk-I wasn't going to comment on this thread because charts and graphs remind me of math and math scares the sh*t out of me-but I just wondered how you got the time to contribute?Isn't there enough going on in Denver?I am sure you haven't met ALL the other Obama delegates yet.I hope Kim Ahern is having fun-she has quite an opportunity at her age to see politics being made firsthand.It's a good thing all you progressives aren't as nice as her,because I would have a hard time working up a supply of venom(just bustin' your chops)McCain in '08!!.

Posted by: joe bernstein at August 28, 2008 5:12 PM

A couple of comments on the addendum graph.

1. It's pretty sad how much better private schools do better than their public counterparts. Some would say that is indicative of how poorly public educators are doing...I'd argue that the figures suggest a voucher system definitely needs exploring because private school hold teachers accountable, are likely to have engaged parents, and are paid in part based upon their students' performance. You do not have that level of accountability in public education and a voucher system would help foster more accountability in the public sector.

2. RI public schools SAT scores compared to MA and CT is downright atrocious. I'm sure Pat Crowley and Bob Walsh would argue that the SAT score isn't the end all/be all of educational testing. Ok, I agree with that notion. But, what you cannot deny is that MA and CT's public schools are doing a much better job at preparing their students than is RI. You simply do not see the same disparity between RI private schools and MA/CT private schools.

Posted by: donroach at August 28, 2008 5:58 PM

>RI public schools SAT scores compared to MA and CT is downright atrocious. I'm sure Pat Crowley and Bob Walsh would argue that the SAT score isn't the end all/be all of educational testing.

That, and trot out the old excuse about RI's "urban districts" (as if Hartford; Bridgeport; Springfield; Lowell; New Bedford and Fall River bastions of wealth).

If public education is supposed to be the great equalizer, then the public schools should be compensating for the "social ills" coming in the door, adding more value if you will, to offset those "urban challenges."

If this can be done, then obviously, the NEA / AFT controlled schools aren't doing that, by Walsh's / Crowley's tacit admission.

Conversely, if this "value added" equalizing can't be done, then we should stop throwing good money after bad.

Rather, we should redirect existing resources away from the teaching pool that isn't adding value and away from "special needs" and ESL programs that aren't accomplishing their purported missions (instead serving as job security for NEA/ AFT members), and toward those students that can be helped - urban and suburban - expanding gifted and advanced programs ...

And universal vouchers, so concerned parents can get their kids away from the dysfunctional public school systems, which have been this way for so long that it is doubtful that the teachers and administrators / bureaucrats within them would even recognize a high performing system, much less be able to implement and operate one.

Posted by: Tom W at August 28, 2008 6:13 PM

Two comments on the main post:

First, though I'm sure Justin didn't intend it, both graphs are good examples of the kind of "gee whiz" graphs discussed in Darrel Huff's wonderful book "How to Lie With Statistics". Eliminating the origin (0 point) on the vertical scale dramatically exaggerates the actual differences between public and private schools. (example and discussion here:

It appears to me that the difference in public and private total SAT scores for RI is about 16%. That's not insignificant, but it's rather less than the graph suggests.

Second, it's impossible to draw any firm conclusions from this data, since there is no attempt to control for parents' educational achievement and socio-economic status, both of which correlate posiitively with the ability to afford private education and other factors. Several studies have found that, once you control for these effects, there is little or no difference between public and private school children on achievement tests (for instance, here: and here

Posted by: Thomas Schmeling at August 28, 2008 6:46 PM

Professor S--

I don't know whether you followed the link to the detailed RI Profile Report.

You will find there some fairly detailed data showing relationships between scores and financial status, parental education, etc.

There appears to be a very substantial connection between scores and parental income and education (although it should be noted that fully one-third of the students did not respond to the family income query).

We can certainly infer from the data (and from our own experience) that as a general matter, the kids in the private schools come from homes with higher income levels.

As for the answer to the "chicken/egg" question, alas the data do not reveal that.

Posted by: brassband at August 28, 2008 8:46 PM

Justin --

Actually, I wonder whether the Professor was in error when he criticized you for omitting an "origin" of 0?

Isn't the lowest possible score on these tests "200?"

Wouldn't that make 200 the origin?

[Please note, I'm not suggesting that you re-do the graph!]

Posted by: brassband at August 28, 2008 10:15 PM

Boy, I don't know why Crowley calls his website when there is nothing futuristic about it.

Okay, now onto smarter thoughts. I am not surprised at all that parochial schools exceed public schools in any measure except drug trafficking and keeping under the radar if you're an illegal.

I went to parochial schools from the first grade and stayed until the end of my sophomore year. Finances with six kids prevented my parents from sending me through the remainder of high school in a parochial school.

Sadly, I had to go to Central Falls High School when Mr. Garvey was the Principal and Mr. Cruz was the Vice.

Can you imagine my disgust on the first day of real teaching that my junior year algebra teacher handed out of syllabus that barely met the one I had in the eight grade of when I went to Sacred Heart Academy in Central Falls?

Now, there is one positive for going to public schools in RI. Spanish language immersion and that's before class starts.

Posted by: Roland at August 28, 2008 10:40 PM


Darn. I fully intended to mention the 200-point "origin," but it slipped away. That speaks to the whole difference between a percentage-based score and a point-based score. By percentages, the "average" student gets a D, and only a very few get As.

Posted by: Justin Katz at August 28, 2008 10:56 PM

Dear Brass,

Please feel free to skip titles, which I only invoke in areas where I've got some certified expertise. This is not one of them. Thomas will do, if you don't mind. I'd go by Tom, but don't want to be confused with Tom W (not likely!) or Tom Segouros)

Yes, I looked at the detailed data you mention but, without individual-level data there's no way to figure out how much of the apparent difference is do to school effects and how much is due to selection effects. Thus, I'll continue to maintain that the data as presented do not demonstrate the superiority of private schools.

Regarding your last comment, if 200 rather than zero is the appropriate minimum, I'm happy with that. 1400 is definitely not. (Justin-feel free to adjust as you wish).

Justin says, "Schmeling illustrates the validity of my rejoinder by failing to make any argument about whether the tighter distinctions are merited, just implying that they are not."

Justin, I honestly don't know what the above sentence means. However, I repeat that graphics that omit the appropriate origin (whether it's 0 or 200) tend to mislead the reader. I'll confess that I learned the lesson the hard way, after presenting a similar graph in a paper and getting slammed for it. I was wrong, my critics were right, and I think serious consumers of statistics will agree that they were right. I sincerely recommend you pick up Huff's book. It's an easy read, and it will make your presentations on AR more convincing.

best regards,
Thomas Schmeling

Posted by: Thomas Schmeling at August 28, 2008 11:02 PM

Can you compare Senator Obama's SAT scores to John McCain's

But remember, they're not elitists! Please.

Why do I even care what Obamessiah's SAT is? Colleges- the singular reason for the existence of the SATs- are hardly using them anymore.

Posted by: EMT at August 28, 2008 11:19 PM

What I meant was — and you've provided another example, with your latest comment #&151; that take up a stylistic matter not on your way to making a broader point concerning what is being missed, but as the response itself.

You comment on the presentation but don't explain why the point spread isn't the important information to have.

Posted by: Justin Katz at August 28, 2008 11:21 PM

Justin say, "What I meant was — and you've provided another example, with your latest comment #&151; that take up a stylistic matter not on your way to making a broader point concerning what is being missed, but as the response itself."

Maybe I'm dense, but I just can't parse this sentence.

Justin says, "You comment on the presentation but don't explain why the point spread isn't the important information to have."

If you're interested in the point spread rather than the total points, then present the data itself in terms of spread, not in terms of total points.

My point, which still stands, is that your graphic had the effect of visually exaggerating the point spread. I'll say it Huff; its really, really good!

Posted by: Thomas Schmeling at August 28, 2008 11:34 PM


I didn't respond to the points you made below the revised graphic.

My first point was only about your graphic, rather than a substantive point about scores. Your graphic visually exaggerated the differences between public and private school scores in a way that, I believe, was not quite honest. As I said, I do not mean to imply that you intended to be dishonest; I've made the same mistake myself.

Second, I can't imagine why you would say that my point about SES and parental education "elides" the point under discussion, since it addresses it directly.

If you have data that show private education to be less expensive than public, please present it. Please don't forget to include high-tuition private schools (I think Wheeler is closing in on $25K/year) (or else eliminate them from your scores and take into account the fact that, currently, tax dollars pay for textbooks, transportation and other services for private schools in RI.

I would really like to know how it is that you think that "reading material in the student’s home, computer access at home, Internet access at home, and the extent to which a student’s studies are discussed at home" are unrelated to parental SES and parental level of education. I am confident that the correlation is quite high.

Posted by: Thomas Schmeling at August 28, 2008 11:57 PM

Hey Justin,

Reading more closely, I think I may have found one point on we agree. Sorta. Maybe.

You say, "And although I haven't seen any studies to this effect, it seems reasonable to me to suggest that the presence of such children has an effect on their peers' performance.

I think you're right. If I recall correctly, you have a college degree, a stable family and a decent job. That, by itself, puts you ahead of a large percentage of the RI population. I agree that you taking your kids out of school probably leaves the rest with fewer positive role models (which is NOT to criticize those who don't have the advantages you have- they may be trying just as hard, but working with less). The more parents such as you remove their kids, the worse off the rest of the kids are.

I'm sure it will not surprise you to hear that I am a big fan of public education. Even as I spend time trying to convince Providence parents to invest in public education, I never, ever, personally criticize any parent for making the (very personal) choice that their own child will be better off in a private school. Nonetheless, I have been working pretty consistently to convince people to enter or keep their kids in the public schools, precisely because I think we're all better off when kids like yours (and mine) are in the public schools and when parents like you are invested in working to improve those schools. When you leave, and stop working to make things better, you diminish the chance that our schools will improve.

While i won't criticize your choice as a parent, I think that's a shame.

For myself, I'm committed to staying inside the system, working to improve it. I regard it as an obligation. Call me foolish if you like. (if you doubt that commitment, email me and I'll let you know what I've been doing). I'm happy to say that I have NEVER felt that my child has lost any advantage.

Posted by: Thomas Schmeling at August 29, 2008 12:56 AM

Despite all the conversation and concern about the presentation of the graph, Justin's main point is that an NEARI union official is more concerned about how many homes the governor owns and his tax bills than the condition of our public schools. Students return to school this week, and what is Pat Crowley talking about? How embarrassing.

Matt J., do you really think the governor was trying to avoid paying his property taxes? He paid them on his other properties, but on this one he was somehow hoping to screw the taxpayers? The shame belongs to this teachers union which is more interested in bringing down our governor than focusing on the education of our kids.

Posted by: mikeinRI at August 29, 2008 7:53 AM


Not to belabor the point but ...

If the range of scores on each of the three parts of the SAT is 200- 800, shouldn't the low score on the graph of combined scores be 600?

Posted by: Frank at August 29, 2008 8:04 AM

I am seeing the difference between government schools and private schools and it has NOTHING to do with SES. My oldest child is now a third grader. Three houses down the street our neighbor’s kids, including one in the same grade as my daughter, all go to the local catholic school. There is a huge difference in where these kids are academically already. For instance our neighbors child is far ahead in math skills, doing things my daughter has yet to be begin (by the schools that is, I have begun to try to help my daughter keep pace with where I think she should be). Last year, 2nd grade, my neighbor’s child had to do 10 book reports - limited in scope I am sure but at least they are being introduced to the process. When I asked my daughter how many book reports she had done she asked, “Daddy what is a book report?”. Needless to say it is highly upsetting as a parent to see your oldest child already falling behind her peers in second grade because you are utilizing the public school system!

So far I would rank my daughter’s teachers: One good one, one very good one, one who didn’t want to teach.

The school day: There is so much fluff built into the government school day that it is a wonder that these kids learn anything at all. As best as I can tell there is about an hour and a half devoted to academics, the rest is wasted. The proof is in the results. The catholic schools manage to provide superior education and teach religion in the same time allotment as that of the government schools. In talking to friends with kids in private and public schools I am convinced that by 8th grade the private schoolers are a full academic year ahead of the public schoolers.

I cannot afford private schools at this time, if I could my kids would already be there.

Posted by: Frank at August 29, 2008 9:54 AM