April 30, 2006

The Reading Wars--NY Magazine

In today's installment we're going to take to task the article "A is For Apple, B is for Brawl" in New York Magazine. The article is about the ongoing reading wars in NYC between phonics and whole language balanced literacy, but there's much more subtext than that. And, that's what we'll be analyzing in this six part series: When Losers Won't Surrender.
  1. Part I: Severed Ears
  2. Part II: Getting the Teams Wrong
  3. Part III: Fun With Memorization
  4. Part IV: Opinion vs. Fact
  5. Part V: Misinterpreting Research
  6. Part VI: The Homestretch
One of the salient points in the article hinges on the definition of Balanced Literacy. The article makes the point, and I agree, that for all intents and purposes BL is whole language with a different name and hiding behind a supercial veneer of phonics.

In the comments, TMAO defines BL as "balance of phonics instruction, instructional reading, authentic reading at an independent level, vocabulary instruction, comprehension skill development, response and H.O.T. development, and finally writing."

I think many teachers would agree with this definition. But, I think it includes things that are typically included in any good phonics program (vocab, comprehension, reading of authentic texts reading at an independent level, response and H.O.T.) and doesn't sufficiently specify the all important interplay between phonics, intructional reading, and use of authentic texts which determine whether the program is really phonics or whole language.

Anyway, I want to write about this later this week, but for now I'd like to hear how others, especially early reading teachers, define Balanced Literacy.

Anyway, onto Part I of the Reading Wars.

When the Losers Won't Surrender -- Conclusion

(This is part VI of VI. The Last post in this series. Here goes Parts I, II, III, IV, and V.)

We're in the homestretch, gang. If things are going to come unhinged, it'll be now. Let's hope New York Mag doesn't disappoint. We'll pick things up where we left off in Part V.
As a direct result of the NRP, those directing federal educational policy held up phonics as a sort of magic bullet, even though the data, critics say, fell well short of supporting such a blanket conclusion. For example, while the full NRP report acknowledged that “phonics instruction failed to exert a significant impact on the reading performance of low-achieving readers in second through sixth grade” and “there were insufficient data to draw conclusions about the effects of phonics instruction with normally developing readers above first grade,” the more widely distributed NRP summary report endorsed phonics without qualification. “Phonics instruction,” it read, “produces significant benefits for students in K through sixth grade and for students having difficulty learning to read.”
This sounds like a strawman. For most kids, phonics instruction should only last about a year. For lower performers, phonics instruction may last an additional year. So for most kids, phonics should be done by the end of first grade. This is true as well for lower performers who begin phonics instruction in kindergarten and in many of the studies we have, this was the case.
Of course, this is all contigent upon proper phonics instruction in the first place, the kind that typically doesn't get taught in most schools.

Therefore, a proper reading of the research is that proper phonics forms a necessary part of almost all beginning reading instruction and should continue until students have learned how to decode. This should take about a year, less for some students, more for others.

After phonics are over with, good reading instructional programs focus on fluency, vocabulary, and reading comprehension. This is a good research based example of what reading instruction should look like in this phase.

Back to our article.

Next we're treated to a paragraph of unsupported innuendo and then:
“Even the NRP found only a small benefit for systematic phonics instruction—and they could not describe with any specificity what that ‘systematic’ instruction looked like.”
This is because lots of phonics program suck ... almost as much as most whole language and/or balanced literacy prgrams. Of course, the best designed, directly taught phonics programs have effect sizes in the neighborhood of a standard deviation, far more than any WL/BL program. To put this effect size in perspective, a standard deviation increase would take a school performing at the 20th percentile (a typical inner city school) up to the 50th percentile, the performance of an average school.

And, while the NRP failed to describe with specificity what "systematic instruction" looked like, all a concerned educator had to do was to look at the underlying studies to find out exactly what instructional program was used and then go pick up a copy and use it.
One critique of the NRP’s report was that it included mainly studies of struggling readers. The report’s conclusion seemed to be that every child, from a severe dyslexic to the precocious toddler plowing through all the Olivia books—must learn the same five skills in the same sequence to learn to read. But what if that’s a faulty assumption, whole-language advocates ask. If the vast majority of kids read without a problem, they say, then gearing an entire curriculum to the learning-disabled is unnecessary—and may impede other kids’ progress.
This statement indicates a profound misunderstanding of how science works.

Listen WL/BL lovers, if you think you you know a better way to teach reading, here's all you have to do: design your instructional program, find a control group and some kids willing to be experimented upon, and field test it. Until then, shut up.

It's also pretty clear that the vast majority of kids do not read without a problem. Some have great difficulty and many do not learn fast or well enough to perform at grade level. And, as far as the scientifically sound programs not working for all children is concerned, this is patent nonsense too. The studies in which at least one of these programs were conducted included many middle class schools and proved that the programs worked for kids with IQS from at least 71 to 131. (See this article, page 5, figure 5.)
“We need to be centrists,” says Fariña, who has even reached out to Sally Shaywitz to study the effectiveness of a particular intervention program called Fundations.
There you go. When your side gets trounced, don't give up the programs that got beat, dig in and call for centrism in an effort to salvage as much of your failed program as you can. After all, it's all about you, not the kids who your supposed to be teaching how to read.
“Kids come to us in various sizes with varying needs. None of us are reading the Times on Sunday on the same page at the same pace and with the same interest, and neither should kids be doing that in their classrooms.”
I'll leave it as an exercise for the reader to figure out the logical fallacy in that statement.
Fariña still believes that a program front-loaded with phonics can lead to rote teaching, which in turn leads to poorer teachers.
Based on what? Pulling out the "rote" card in education is as odious as pulling out the "race" card in politics.
Most of all, Fariña remains devoted to the proposition that the vast majority of kids just don’t benefit from being drilled.
Based on absolutely nothing. Maybe she should talk to a cognitive scientist.
“I want kids not only to learn how to read, I want them to want to read,” she says. “And I don’t think that all the skill and drill that’s happened over the years will lead to that if we don’t do the other piece of it.”
I'm not quite sure why a research based phonics program necessitates excluding the "other piece of it," whatever that happens to be. A classic false dichotomy.
The fact is, New York is most likely to remain a whole-language town. Federal mandates and MRI scans aside, progressive education is part of the academic culture here.
If you want to know why we have NCLB, this is why.

I told you it'd be worth it if you stuck it out to the bitter end.

When Losers Won't Surrender -- Part V

This is Part V of a multipart post. I'd suggest you go back and start at the beginning (Part I).

Let's pick up where we left off fisking this article in New York Mag:

Where George W. Bush and many red states are phonics supporters, New York is dyed-in-the-wool whole-language country. Influential programs at Columbia and Bank Street College developed variations of the approach before it even had a name. Balanced Literacy, or at least the way it’s practiced in New York, is largely the brainchild of Lucy Calkins, founder of the Teachers College Reading and Writing Project, who is looked upon nationally as a godmother of whole-language learning.

The Bank Street College model! Influential--perhaps. Effective--definitely no. Look:

See how the Bank Street Model (fifth from left) significantly underperformed the performance of the control group (performance = 0) and vastly underformed a well-designed the basic skills model (first from left).

Ms. calkins is fortunate there is no such thing as educational malpractice, because the results of this study would be plaintiff's exhibit 1.

A National Institutes of Health–created commission of Ph.D.’s came down squarely on the side of phonics in a 2000 report, influencing the Bush administration to crack down—some say improperly, perhaps even scandalously—on non-phonics programs.
Is there anything worse than the lazy journalism inherent in the use of the "some say" device.

That argument doesn’t persuade Klein. He’s cultivating mindful, curious readers, he’s said, not vanilla word-decoders. “I’m quite convinced the curriculum we’re using, with inquiry-based learning, will serve our students throughout the city well over time,” he says. In particular, Klein likes that Balanced Literacy looks a lot like the reading approaches in successful school districts on the Upper West Side and the Upper East Side and in most of the city’s elite private schools. In a system where so many great schools coexist with so many horrible ones, Klein is convinced that the solution is not to adopt the practices of the worst schools but to export the best practices of the successful ones and end the educational apartheid.

This is soft-headed thinking at its finest: “I’m quite convinced” based on absolutely nothing.

I've already dealt with the silly unsubstantiated Kozol-esque educational-apartheid arguments in detail.

In the summer of 2002, he hired Klein, a fellow outsider, as his chancellor, and Klein recruited a career superintendent named Diana Lam as his deputy for instruction. It was Lam who brought in Balanced Literacy. Neither Klein nor Bloomberg knew much about the program at the time, except that Lam had used it in cities where test scores went up, like San Antonio, Texas, and Providence, Rhode Island.Nothing more embarrassing than picking an expert that doesn't know what they're doing. And, it would appear that those gains in test scores were largely, if not totally, illusory. But, our hapless journalist accepts the assertion uncritically.
Then the article finally gets to the phonics side of the debate. But, we soon get the following statement:

Parents in the more politically connected parts of town didn’t need to be won over by Balanced Literacy, since more than 200 elementary schools already used it.

"Using" is not the same as "liking." "Tolerating" would be the more appropriate word, and that's only because these skills will almost never have the same rates of reading failure you see in the low SES schools no matter what reading program is in place. This is due to the SES/IQ correlation and the familial support system the high SES kids can fall back on.

Cognition experts like Harvard’s Steven Pinker have argued for some time that while learning to talk is an organic process you can generally learn on your own, like walking, reading is more like riding a bike or driving a car. Someone has to take you through the initial steps and get you over the unfamiliarity of the experience; then you have to spend time on your own perfecting the skills until it becomes second nature. The question at the heart of the Reading Wars is how much direct instruction do children really need.

It's much more than an argument at this point. The weight of scientific authority clearly indicates that reading is not like talking. reading is not hardwired into the brain, like talking is; it is a learned skill.

And, as far as how much direct teaching of reading is needed, again, the weight of scientific authority clearly indicates that direct teaching should be close to 100%. The more non-direct teaching is used the worse students perform.

Take another look at the graph above. The Direct Instruction model (and to a lesser extent the Behavioral model) used almost 100% direct teaching. Look at the results. the remainder of the models used various degrees of less direct teaching. Look at the results. Pathetic.

These results included many low-SES students, but also included quite a few middle-class kids as well.

In 2000, the [National Reading Panel] released its “meta-analysis” and concluded that in order to learn to read, all children must master five separate skills: phonemic awareness (separating words into distinct sounds, like the c, a, and t in cat), phonics (learning the sounds letters and letter combinations make), fluency (the ability to read with speed and accuracy), vocabulary (learning new words), and comprehension (understanding what you’re reading). These basic skills were nothing new to most people who taught elementary-school English. What the NRP added to the debate was the notion that direct instruction of these skills was the only proven method for teaching reading.

The conclusion (that the NPR added the notion that direct instruction of skills is the only proven method) in the last paragraph is contradicted by the fact that that NPR's analysis was merely a meta-analysis. The NPR didn't conduct any new research, it just evaluated old research. (Actually, it threw out 90% of the old research because it didn't conform to even the looser standards we permit in social science research.)

What the old research clearly showed was the direct instruction was the only model proven to be effective. The research had been ignored by our clueless educators. The NPR has merely reminded us that the research existed and that our educators were abiding by it.

Continue reading the final article in this series. Conclusion: Part VI.

When Losers Won't Surrender -- Part IV

This is Part IV of a multipart post. I'd suggest you go back and start at the beginning (Part I).

Let's pick up where we left off fisking this article in New York Mag:

Unlike traditional so-called phonics-based programs, in which kids repeat and memorize basic spelling and pronunciation rules before tackling an actual book, whole language operates on the presumption that breaking down words distracts kids, even discourages them, from growing up to become devoted readers. Instead, students in a Balanced Literacy program get their pick of books almost right away—real books, not Dick and Jane readers, with narratives that are meant to speak to what kids relate to, whether it’s dogs or baseball or friendship or baby sisters.
The Dick and Jane readers weren't phonics-based readers. They were part of the "Look-Say" reading method, a precursor to whole language (which is the predecessor to Balanced Literacy). The underlying theory is the same. Instead of Dick and Jane primers, children now read "authentic texts" which have their reading level contolled by dopey readability formulae.

Yet another major mistake by the author.

The problem with using "authentic texts," besides the fact that many of these so-called authentic texts have been dumbed down to comply with the readability formulae, is that they are frequently not controlled for decodability. This means that children will not be able to decode many of the words contained in these books. This is why in Balanced Literacy children are urged to guess at words they don't know by using context clues--they are unable to decode the words.

Sometimes the teacher will teach the student the phonics rule on the spot. If this sounds like a haphazard method of teaching, that's because it is. Guess what happens if the child doesn't encounter the unknown phonics rule again for a few days? He's most likely going to forget the rule since it hasn't been reinforced and practiced lately. This just makes the memorization process much more difficult than is necessary. Not surprisingly, when phonics rules are taught explicitly and systematically, children tend to learn to learn to read faster and better. (We'll get to this soon.)
Over time, the theory goes, kids learn the technical aspects of reading—like contractions, or tricky letter combinations painlessly—almost by osmosis. The joy of reading is meant to be the spoonful of sugar that helps the medicine of spelling and grammar go down.
But, as we know the theory is no longer just a theory. It is a disproven theory. It was a loser in the scientific process. It should have been discarded long ago. Instead it lives on.

Let's skip around a bit:
The catch is that in the past five years, research has emerged suggesting that phonics, not whole language, is the superior teaching method. Phonics advocates point to the new research as evidence that the Klein reading revolution is badly misguided. WhatÂ’s needed, they say, is a phonics counterrevolution.
The research actually emerged at least 30 years ago, but educators refused to follow it. As this article points out, they still do.
Everyone stands to gain from phonics, advocates say, but no one figures to benefit more than children from low-income families who—unlike, say, the kids at elite private schools, most of which use a whole-language approach—often can’t get extra tutoring in the basics.
The argument goes something like this: kids at elite private schools (and affluent public schools) are getting whole language instruction; therefore, this is the best way to teach children to read and all kids should be taught this way. This argument is not only wrong, but flatly contradicted by research. (In case you were wondering, this happens to be Kozol's argument.)

First of all, the fact that higher performers can succeed in these programs (remember nothing is preventing any student from learning) does not mean that average and low performing kids will suceed in it. And, why should any kid require extra tutoring in the basics? Isn't this an indication that the basics aren't being taught properly or sufficiently in a program that is specifically design to skip teaching those boring basics? And, who's to say that these elite college bound kids haven't been damaged by whole language? SAT verbal scores plummeted by almost a standard deviation (80 points) in the mid sixties after the introduction of look-say. This drop occurred across the board and included kids at the very top as well. Verbal scores have never recovered. In fact, the SAT verbal scores had to be recentered to account for this decline.
Whole-language proponents, in turn, say phonics perpetuates authoritarian, patronizing “drill and kill” strategies that insult the art of teaching and turn kids into fifties-style robots, putting them off learning for life.
As I pointed out already, the drill and kill meme is flat out wrong. It is the whole language kids that get turned off the reading because whole language tends to produce many non and poorly performing readers. Non-readers and poor readers tend not to like reading. Kids who don't like to read tend not to learn much.

And what's all this nonsense about insulting the art of teaching? Please. It is these wacky teaching techniques that insult the art of learning.

This would be a valid point if whole language techniques produced better readers, but as my fancy graph indicates, whole language produces the worst readers. Who's insulting whom?

This is the main problem with these reading wars articles The phonics side has a fair bit of empirical evidence to support its contentions. In contrast, the whole language/balanced literacy side has nothing to offer but its opinion. And, in 2006, that opinion has long since been discredited. Yet, the article offers the position of both sides as if they were entitled to equal weight.

To be continued in Part V.

April 29, 2006

Amusing Memorization Sidebar

This is Part III of VI. Part I and Part II are here.

In Part II we finished off with the memorization meme:
Unlike traditional so-called phonics-based programs, in which kids repeat and memorize basic spelling and pronunciation rules before tackling an actual book...
See, phonics requires lots of memorization. Presumably, this memorization is boring drudgery. To adults perhaps. But, anyone who knows little kids, knows that they don't have the same disdain for memorization that adults have.

And guess what, if you don't memorize the sounds that make up words before "tackling an actual book" you won't be able to decode the words and you'll have to resort to guessing what the words are, which isn't reading.

And, are we supposed to believe that there's no memorization involved in "balanced literacy"? Of course there is. Kids are either memorizing letter sounds, just like in phonics, or they're memorizing whole words, as in whole language. Either way there's lots of memorization going on. It's a necessary evil, but an evil that kids don't seem to mind and might actually enjoy.

Let's set the record straight right now. Reading is a skill, an overlearned skill. Good readers can automatically decode words on the page. This automatic recall means that the good reader has memorized many many words and sounds.

Let's prove the theory by taking a little test. Below are a bunch of words printed in colored text. Your job is to look at each word and say the color of the text aloud. Don't read the word aloud, just say the color of the text. So for the word yellow you say red not yellow.

Got it?

Ok.

Ready .... Go:






Gee that was difficult wasn't it? It's almost like you couldn't stop your brain from reading the words. That's because you are an expert reader and reading is automatic for you. Congratulations, you're a memorizing machine. All good readers have the ability to read connected text rapidly, smoothly, effortlessly, and automatically with little conscious attention to the mechanics of reading, such as decoding. In short, you are a fluent reader. You have over-learned word reading skills to the point where decoding requires little or no mental effort. As a result, you are able to put all your mental energies into reading for meaning. Reading has become so automatic that you struggle not to do it!

Give the same test to a child who hasn't learned to read and she will not get the same brain interference you just experienced.

Let's get back to analyzing the article in Part IV.

When Losers Won't Surrender -- Part II

This is part II of VI. Part I is here.

In part I we learned that "balanced literacy" programs typically teach children how to read by having the students guess at words by looking at illustrations or other context clues. This isn't reading, it's a sham:

While some kids may learn to read from this approach (nothing is preventing them from learning what they're supposed to learn), some higher performers may totally misinterpret the game, and lots of lower performers fail to catch on to what reading is.

We once did a nice demonstration that showed how confusing the approach may be to naive kids. We went into a first grade classroom where a teacher had worked on four different selections. Each had an illustration and the text. The kids could "read" all selections perfectly. We then switched the illustrations and the text (paired them with different texts) and tested the kids. About half of the kids pointed to the words one at a time and, with great fidelity, recited the passage that was appropriate for the picture. In other words, half the kids didn't have the faintest idea what reading was all about.

From War Against the Schools' Academic Child Abuse.

Instead of giving the reader that devastating critique of balanced literacy, we are provided a phonics classroom right out of Dickens:

If throwing Enami into the deep end of the pool like this seems a little intense, that’s pretty much the point. What’s unusual about this lesson—and to its critics, flat wrong about it—is what’s not happening. Enami and her seventeen classmates are not sitting in a row, repeating letter and pronunciation drills. They almost never are. There’s not a textbook in sight, or, for that matter, in the whole school.
Now that's presenting a nice balanced view of the two sides of the Reading Wars.

As we all know, the only way to teach phonics is by "sitting in a row, repeating letter and pronunciation drills." None of those fancy semi-circles for the phonics kids, the only way to learn phonics is by sitting in rows and repeating (ad nausem no doubt) pronunciation drills. This is your classic false dichotomy.

I do like the "throwing the kid into the deep end of the swimming pool" analogy though. Imagine if we really taught kids to swim by throwing them right into the deep end without receiving lessons on the basics of swimming. Some would swim, but most we'd have to pull from the bottom of the pool, just like little Enami.

Instead, they’re learning by immersion, reading books of their own choosing, and when they mess up, which is often, they’re told to keep going.
That's it. Don't correct the students' errors No, that would interfere with their "natural learning" process. Nevermind that all those misrules (i.e., errors) will someday have to be remedied (hopefully by some other teacher), unlearned, and then the proper rules learned.

We're about to learn that this war is not really a war about phonics vs. whole language, it's really war against bad teaching practices vs. good ones. Let's continue.

Balanced Literacy is more of a catchall concept than an actual curriculum, interpreted slightly differently in every school system that uses it, but it is invariably rooted in an education philosophy known as whole language.
I'm going to criticize the author of this article quite a bit, so the upcoming compliment means something. It is an extremely lucky perceptive insight that balanced literacy = whole language. Think of whole language as anti-phonics. You can't balance phonics and anti-phonics. So, what happens is reading gets taught mostly according to whole language principles with a superficial veneer of phonics, taught mostly incorrectly, grudgingly strapped on.

Unlike traditional so-called phonics-based programs, in which kids repeat and memorize basic spelling and pronunciation rules before tackling an actual book, whole language operates on the presumption that breaking down words distracts kids, even discourages them, from growing up to become devoted readers.
First of all, phonics is not about pronunciation rules, phonics is a code-emphasis approach which emphasizes predictable letter-sound correspondences and reading words composed of those correspondences. If you're going to write an article about the reading wars, you should at least try to get your definitions right.

And, "whole language operates on the presumption" should be changed to "whole language operates on the discredited presumption" to make the last sentence accurate because, as it turns out, good readers do in fact break down words as they read them. They just do it extremely quickly. Moreover, there is no evidence that phonics discourages children from reading.

Then there's the memorization issue which we'll look at in Part III.

When Losers Won't Surrender

In the May 1 issue of New York magazine we find an interesting article on the Reading Wars "A Is for Apple, B Is for Brawl." I can tell right off the bat that we're in for trouble because the title confuses letter names for phonics.

Do you remember the opening montage from the movie Blue Velvet? The camera pans through an idyllic small town, replete with picket fences and flower beds, and then burrows into the lawn until we come upon insects devouring the severed ear--a metaphor for evil lying right below the surface of this outwardly perfect town.

This article gives us the educational equivalent of the severed ear.

In the first two paragraphs we are treated to a description of the perfect classroom, but then we burrow in ...

With Kolbeck peering over her shoulder, Enami opens the book and bobs her head, the bright blue beads in her cornrows jostling as she starts reading aloud.

“On a sun day . . . ”

It says sunny day in the book. But Enami’s a little tentative. She hasn’t read this one before.

“A man with a collar . . . ”

The teacher has a suggestion. “Sometimes we look at the picture and figure out if it makes sense.”

Enami eyes the drawing of a man walking a dog. She agrees collar doesn’t seem right. After some discussion, it’s decided that collie works better.

“A dog!” says Enami, satisfied.

She continues—and a page later she trips up on the word disappeared. She takes her best guess: “Stepped.”

“Let’s see if that makes sense,” says Kolbeck.

Again, Enami checks the drawing: a man at the end of a street, turning a corner. Her eyes flash—“Disappeared!” And on she goes.

The severed ear!

This kid isn't reading by learning how to decode the words on the page, but by looking at the pictures and by guessing at the words. That's not reading; that's guessing...

... And no one in the school recognizes this for the horror it is. It's called balanced literacy and is the mandated method in New York City and the preferred method of teaching kids how to read in the U.S.

Let's pick up the analysis in Part II.

Update: This article is getting the treatment over at Kitchen Table Math too. Check it out.

April 27, 2006

Reading Recovery -- On Life Support

Looks like my post on Reading Recovery ruffled some feathers of the pro-Reading Recovery crowd. Even if you shoot the messenger for bearing the bad news, the bad news isn't going to go away. And what is that bad news? I'm glad you asked:

  1. RR only claims to be able to raise the performance of the "lowest students" in the class up to the average performance of the class. This means that in a typical inner city school peforming at the 20th precentile, RR only promises to raise the performance of the "lowest students" up to the 20th percentile not grade level (50th percentile). Gee, that's quite an accomplishment, turning non-readers into slighhtly better non-readers. Pop the champagne!

  2. I put lowest students in scare quotes because RR appears to define lowest students in a way that excludes the actual lowest performers. RR doesn't accept students who lack the prerequisite skills to enter the program and/or students already in special-education. One study (Battelle (1995)) reported that these excluded students numbered 19%. Most likely the bottom 19%.

  3. According to the same study, another 21% of students never complete the course because they fail to make adequate progress and another 9% complete the course but don't get up to the classroom average, which, as I indicated above, can still be considerably below grade level. I'm thinking that these students also inhabit the lower end of the performance curve.

  4. So let's recap: When RR talks about its success rate, they've excluded about half the students who weren't eligible for treatment, who were thrown out of treatment, and who didn't pass treatment. Talk about triage: they toe tag the sickest patients and only "cure" the ones who are only a little sick. One study (Hiebert, 1994) estimated that the average RR student had an entry score of the 34.5th percentile. What about the kids in the bottom quintile? I guess we need another program for them. And, why can't we just use that better program to remediate these higher performing kids? It can't be more expensive.

  5. Then there's the issue of how RR defines "reading." The average RR student completes RR at a level in which they are reading from texts that are still very predictable. This means that children can read words without looking closely at them. The children rely more on the contextual clues, the illustrations, and the repeated sentence patterns in the text. Children who use these contextual strategies to read are more likely to be successful in predictable text than in authentic text. Consequently, children from RR may not read authentic text very well at all when they are returned to the classroom as "successful." The strategies they have learned for reading may not generalize to real reading.

Add, all this up and it's easy to see why there's little recovery in the RR treatment.

I suppose if this were my pet program, I'd have my panties in a bunch too, seeing these results.

I'm thinking if RR were a drug it'd be classified as snake oil.

Edspresso Debate

There is a great debate going on between Michael Petrilli (The Fordham Foundation) and Neal McCluskey (from the Cato Institute) on National Standards at Edspresso.

However, I'm not entirely convinced that either party is on the right track.

As much as I like the free market system, the reality is that we live in a social welfare state and that's not going to change any time soon. A majority of people believe that subsidizing education is a good thing. Once we agree to public subsidies, then the public will want a say in what's going on in education, for better or worse. Now we're stuck in the political process with all its evils.

Then we have the very real problem that we do have somewhat of a free market in education, private schools and our university system; but the market appears not to value increased student achievement or, at least, is incapable of increasing student achievement.

Do we know if a Harvard education is any better than a state college education once you adjust for student ability? Our university system does a fairly good job segregating students by IQ and works as a signaling mechanism for employers, but how well do they actually do in the education department. In this respect education is not like an ipod. An ipod is simply to understand and evaluate, a quality education is not. Education is an extremely complex product that is probably beyond the ability of your average consumer to effectively evaluate. And, third parties haven't exactly stepped up to provide this service. The one thing lacking from all these school ranking services is an identification of the actual education you're going to receive from these institutions.

Then we have private K-12 schools. They seem to be just as infected, if not more so, with the same faddish nonsense that pervades the public schools. They may eke out a tiny marginal advantage, but once you adjust for SES, its pretty small and certainly not sufficient to bring all students to the level we want.

One problem is that we do not have a decent signaling mechanism in place in which the consumer (the parent) is able to readily determine how well a school is actually doing educating students. The devil is in the details (the curriculum) and finding out about the actual curriculum in most schools is close impossible, assuming the consumer even knows what to look for. Most do not.

On the flip side, we have the problem of standards, national, state or otherwise. For the most part they stink at the state level. This is because the standards are written by people who have never successfully taught students anything, let alone the ones on the left half of the curve. Are they effective in weeding out ineffective curricula-- doubtful. Do they prevent successful curricula from being implemented--sometimes.

Then we have the whole regulatory capture issue. Education groups are very adept at getting their "standards" enshrined into national and state standards. The NCTM has been extremely successful getting their "standards" into the NAEP math exam. The test has been revised to accord with these standards, nevermind whether actual math is tested anymore. Or whether the "math" that is tested is actually necessary to succeed in higher level mathematics. It's not like the NCTM has ever successfully raised the student achievement of any student as a result of their standards. They were imposed from the top without any sound grounding in reality. They've also been extensively revised a few times, indicating that they didn't get it right in the past.

No, standards are one of those things that sound good on paper but fail miserably in reality. Kinda like communism.

Solving the education crisis is going to require attending to the smallest of details or putting a mechanism in place in which the providers of education are forced to. Unfortunately, these messy details are not amenable to high level policy debates and sound bites.

April 26, 2006

The Three Little Pigs

It is frequently said that educating children is not like making widgets because every child is different. As a result, the end quality of the widget is highly dependent upon the material with which you start. Some of the starting material is simply defective and will never produce a suitable widget.

There is, of course, a grain of truth in this argument. But, based on NAEP, about two thirds of our educational widgets roll of the assembly line broken. Educators would have us believe that all the broken widgets are due to defective starting material because their underlying assumption is that they presently are doing the best that they can.

I think there is sufficient research out there that convincingly shows that no more than about 10% of the starting material is fatally defective. This means that about 85% of the broken widgets produced by our education system are not the result of defective material, but rather the result of a broken manufacturing process (the instructional delivery system).

With these parameters in mind, let's analyze the starting material.

Sure, the material comes in differing qualities. Most of the material is average quality. Some is below average and some is of exceptional quality. As I've pointed out above, 90% of the material is capable of being crafted into an acceptable finished widget. But, some of the material will take a great deal of skill to fashion, while some of the material almost builds itself. The weakest material will always result in the vast majority of broken end products. And, the best material will have the fewest.

Let's use the story of the three little pigs to illustrate the point (as if I haven't tortured this metaphor enough already).

The Brick House

The pig with the bricks (the best material) doesn't have to build his house that carefully. It will easily stand up to the big bad wolf. Even with sloppy building, the brick house will often be the strongest and will produce the least number of houses unable to withstand the wolf.

The Wood House

The pig with the wood (average material) has to attend to more details in building his house and has to employ a good design or else it won't stand up to the wolf. Even with careful building, the wood house most likely will not be as strong as the the brick house. And, more of the wood houses will fail when the wolf comes.

The Straw House

The pig with the straw (the weakest material) has a very difficult job ahead of him. He has to build his house very carefully and has to monitor the process every step of the way. He has to use efficient and advanced building techniques to make sure the house is up in time before the wolf comes a knockin'. Any errors are potentially fatal, especially errors building the foundation. This pig has to get everything exactly right and work at an accelerated pace or he'll get eaten by the wolf.

Even with careful building, the straw house will most likely be weaker than the wood house. But the well built straw house can withstand the wolf when he comes. But, overall, the greatest number of failures will be straw houses.

In education today, the craftsmanship standard is average to sloppy. Most brick houses stand up, some of the wood ones do too, but almost none of the straw ones do.

And bear in mind that in most states the big bad wolf is emphysemic.

April 25, 2006

In which our hero ...

... takes on the union drones at EDWise on the subject of class size.

Update: Well, that stayed somewhat civil and the teachers made some good points (of course, they were points I had already conceded). No one had a response for my assertion that the current research failed to show any advantage in decreasing class sizes across the board. That's because there is no valid response. If there were any valid evidence, and the UFT surely has access to classroom data, they would have ran it up the flagpole long ago.

So, I'll stick with my present opinion: first you fix the broken instructional delivery system, then you worry about tweaking the class sizes to make sure the implementation can handle the class size, targeting areas that need reductions the most, like the early elementary years in low-SES schools.

April 24, 2006

Kozol -- Part II

As I wrote in Part I, Kozol has jumped the shark. He's no longer content spewing his crackpot theories, he's now actively disparaging successful instructional programs that work in favor of those that don't:
And when children are demonstrably in bad trouble, I would never institute Skinnerian approaches like Success for All. I would spend a lot of money to use remarkably successful and highly enlightened programs such as Reading Recovery, which I'’ve watched being used many times. Everything I criticize I've seen; I could teach SFA, because I've been in so many classrooms where I'’ve seen it done. And I'’ve watched Reading Recovery, and that'’s probably the most effective way of catching children up very quickly. It's usually done only for first graders, but some schools use the same ideas for second and third graders who still need to catch up. ItÂ’s very expensive, but itÂ’s the best way.
Oh really.

"Wasik and Slavin (1993) compared the relative effect sizes achieved by five treatments for reading problems. Reading Recovery was not nearly as effective as two programs that provided explicit systematic phonics with extensive practice reading decodable text (the Success for All and Wallach and Wallach programs)."

If I'm reading this bar graph correctly, SFA is nearly twice as effective as Reading Recovery.

But, this is the part I like the best:

"If a school's goal is to raise the overall level of reading performance, Reading Recovery is not the appropriate intervention to choose. Overall school achievement scores are not improved with the use of Reading Recovery (Hiebert, 1994). Both Reading Recovery advocates and critics agree on this point (Hiebert, 1994; Pinnell & Lyons, 1995)"

Even the Reading Recovery advocates admit it doesn't work. They won't deter Kozol though:
In the Success for All schools the teacher is not allowed to do this. And I'll broaden that statement: This is true in all these heavily scripted, test-driven schools where the principals are in a state of perpetual anxiety because of the threat of sanctions and humiliation. So in these schools the teacher has to cut that little boy off, because he may want to talk of tears of laughter, but tears and laughter won't be tested on the standardized exam. There's nothing in NCLB about the sorrows of a child's heart or the laughter that comes naturally from a child's spirit when he's in a reasonably healthy public school.
What is it about scripted instruction that draws out the over-heated rhetoric?

Here's a script from the third grade Reading Mastery program. Please, someone, tell me what is so objectionable.

Don't teachers make-up their own "scripts" or lesson plans already? So, if teachers make up the scripts, it's OK; but, when the instructional program designer does it, it's not. Nevermind that the instructional designer's scripts have proven to be highly successful with low-performing students whereas teachers' own scripts have not.

Oh, that's right, these scripted programs, like SFA and Direct Instruction, are supposed to be soul killing "drill and kill" drudgery for students. Such charges have been around for a long time:
Critics have often complained that the DI model was a pressure cooker environment that would negatively impact students' social growth and self-esteem. As the Abt Associates' authors note:

Critics of the model have predicted that the emphasis of the model on tightly controlled instruction might discourage children from freely expressing themselves and thus inhibit the development of self-esteem and other affective skills. (Stebbins, St. Pierre & Proper, p. 8)
Because of this expectation, the affective scores are of interest. Three of the five lowest scoring models on the affective domain were models that targeted improving affective behavior; none of the affective models had positive affective scores. In contrast, all Basic Skills models had positive affective scores with the Direct Instruction model achieving the highest scores. The theory that an emphasis on basic skills instruction would have a negative impact on affective behavior is not supported by the data. Instead, it appears that the models that focused on an affective education not only had a negative impact on their students' basic skills and cognitive skills, but also on their affective skills.
You mean to tell me there is a study, which cost hundreds of millions of dollars, out there that conclusively shows that a scripted program had the highest self-esteem scores while Kozol's beloved whole-language (the TEEM program) had the lowest. I'm shocked.

But, Kozol still won't be deterred:
In suburban schools, children are being enabled to ask interesting questions, they're being educated to interrogate reality, they're being educated also to read beautiful children's books, to inherit all of the treasures of this earth. In the drill-and-kill curriculum in the inner-city schools, there's often no real literature.
Compare the cognitive and basic skills scores for Direct Instruction and the scores from the programs from the right of the graph which used all the stuff the Kozol advocates.

Now tell me who would you rather educate your child?

The defense rests.

Update: Unexpected bonus post on Reading Recovery.

April 23, 2006

Jonathan Kozol -- Education's Greatest Monster

Education's lovable crank, Jonathan Kozol, recently surpassed Alfie Kohn as Education's Greatest Monster while on the road promoting his latest jeremiad.

No longer content in rehashing his pessimistic view that poor kids can't learn --wait for it-- as long as they're poor, he is now actively bashing effective instructional programs ostensibly because they are successful in educating poor kids. This is an impossibility under his pet theory.

His blissfully fact-free siren songs are very alluring to our gullible educators. He is one of the most often cited sources for education apologists. This is because he basically just says what he wants with complete disregard to the facts.

Take for example this interview with Voice in Urban Education.

Let's dispense with his main argument right off the bat. Kozol contends, without any proof, that you can easily improve student performance of poor kids by sending them to affluent school districts. Supposedly, the teaching that is going on in these affluent school districts is so superior that the deprived children would naturally just learn everything they need to. If but this were true.

Apparently, no one has told Kozol that NCLB now requires school districts to disaggregate their data by race and SES so it's now pretty easy to show that he is wrong.

Let's use my school district as an example. My school district ranks in the top 5% of school districts in Pennsylvania. It is a very affluent school district, located in the town of a highly selective liberal arts college. We spend more than 95% of all of Pennsylvania's school districts. My district is highly enamored with all the fashionable nonsense that passes for education nowadays. Here is the latest report card. Let's use the white students' score as a proxy for high-SES and compare the white students to the low-SES students using the average scores for reading and math in all grades tested in 2004 and 2005:


Math

White (High SES): 82.7
State Avg: 65.3

Low SES: 52.5
State Avg: 36.8

Black: 40.2
State Avg: 28.5

All students: 80
State Avg: 58.3

Reading

White (High SES): 85.7
State Avg: 71.3

Low SES: 52.7
State Avg: 41.3

Black: 53.7
State Avg: 35

All students: 83
State Avg: 64.3


Now let's look at the achievement gaps for math and reading combined (AG):

AG between High and Low SES students: 31.6
AG State Avg: 29.3

AG between Black and White students: 37.3
AG State Avg: 32.25

There we have it.

Low-SES students only perform marginally better in a high-performing affluent school, but still perform below the state average and have an even larger achievement gap between them and their high-SES peers. And, bear in mind that I haven't controlled for the actual SES of the Low-SES students. In all likelihood the low-SES students in my district have a higher SES than the state average for low-SES students, which would explain most, if not all, of the small gains we see.

So much for that theory. Now let's get on to the meat of the analysis on part II.

Update: As I suspected, the low-SES students in my district have a significantly higher SES than the low-SES students statewide. According to schoolmatters, about 22% of the low-SES students in my district have a family income of less than $15k, compared to 36% statewide. Similarly, 47% in my district have a family income between $15k and $30k, compared to 43% statewide. The remainder fall in to the $30k-$50k bracket, compared to 21% statewide. In PA, 42% of students are classifiable as economically disadvantaged, which is way too high to be an accurate indicator of true poverty. Nonetheless, this higher SES in my district for the students classifiable as low-SES does explain much of the increased student performance compared to the state scores.

April 21, 2006

Undercounted NCLB Subgroups

The eduspere is all aflutter with NCLB's undercounted minority fiasco. The tension is between statistical signficance on one hand and accountability for traditionally underperforming minority groups on the other. I think I have a solution.

When a school doesn't have enough of one or more minority groups to form a statistically significant group -- just aggregate them. We'll call this group the traditionally underperforming subgroup. The group would include african-americans, hispanics, and native-americans.

If you couldn't form a statistically significant group out of these subgroups, you could then disaggregate the non-northeast-asians from the asian subgroup and include them too. And, if you still needed more in the group, you could include the economically disadvantaged kids to capture some of the lower-performing whites too. That should do it.

Actually, why don't we just get rid of these silly racial subgroups altogether and focus on the real problem area -- the low performers -- the kids who fall on the left half of the IQ curve, no matter what subgroup they come from. These are the kids who need to most educational attention and resources.

We don't need to separately track the high performing african-americans, hispanics, and native americans disproportionately located in the more affluent schools, they're going to perform more like their white and asian counterparts. Likewise, shouldn't we be more concerned about the performance of the low performing whites and asians? They're more likely to peform like the lower performing african-americans, hispanics, and native americans, so let's give them the same attention.

Of course, then we'd have to face the ugly reality of IQ disparities and racial differences, but if we really are concerned about not leaving any child behind, we're going to have to face up to this reality sooner rather than later.

April 20, 2006

New NCLB Poll

Let's examine the results from a new NCLB poll as reported in this Duluth News Tribune article.

An AP-AOL Learning Services Poll found nearly eight in 10 parents are confident their local schools will have students up to state standards by the 2013-14 school year target. Yet only half of teachers are confident the kids in their schools will meet that deadline.

The finding underscores a theme in the poll. Parents and teachers often disagree on daily aspects of education, from the state of discipline to the quality of high schools.

A major reason is that adults see children differently. Parents tend to focus on their own children, while teachers work with dozens of students from different backgrounds.

A more likely reason is that teachers are seeing first hand the large scale academic failure taking place in America's classrooms. When you teach lower performing kids, year in and year out, and you see them not learning what you presented (not taught), I can see how it might cloud your perspective.

"I think the standards are being applied to everybody indiscriminately, without regard to their abilities," said Steve Peterson of Knoxville, Ill., who has been teaching for 31 years.

"Schools in general are not going to be able to meet the standards," he said.

Mr Peterson, the standards are being applied indiscriminantly because we don't wan't to discriminate against any student.

Within one sentence, our expert teacher has already managed to "blame the children" for not learning because it is their "abilities" determine whether they are capable of learning. Notice how Mr. Peterson didn't question his own or his school's ability to effectively teach these children. No, it's the students' fault that the schools won't make the standards.

The survey also found:

64 percent of teachers say their state standards in reading and math are about right. Most parents agreed. But parents were twice as likely as teachers -- 31 percent to 15 percent -- to say current standards are too lenient.

That's because the state standards are too lenient. On average, more than twice the number of students pass the state tests that pass the NAEP test. Coincidentally, the number of students who pass the NAEP test is roughly the same amount that successfully complete college (about 30%). And, let's not forget that about 25% of students drop out of school before they take these 11th and 12th grade tests.

Did I mention that the 11th grade math NAEP only includes question at about the 7th or 8th grade level?

Parents with college degrees and higher salaries were more optimistic about their children's chances for success than parents with less money and less education.
And well they should be. Parents with college degrees tend to be smarter than the average person and tend to have kids who are smater than average. Smarter kids tend to do better in school.

Rusty Barker, a parent in Jackson, Mich., said he's pretty optimistic that the local high school can get all students up to math and reading standards. He attended the school, and now his daughter does.

"There's more of what kids need, in a one-on-one basis," Barker said. "While kids have different backgrounds, that shouldn't be an issue for their learning experience."

Actually, it is an issue. Kids from certain backgrounds need much more careful teaching.

President Bush says it is bigotry to expect less of some students, particularly if race is a factor. As he said at the White House in 2004: "We believe every child can learn. We want to know if every child can read and write and add and subtract early, before it's too late."

Many teachers say the sentiment is right, but incomplete, or even naive.

Who's being naive now, Kay?

Some students come to school way behind their peers. They may not have good English skills, or study habits, or parents to reinforce lessons at home. The law says schools must overcome that.

Exactly. And, we know they can overcome those deficiencies.

We know some kids will come into school with these deficiencies, it is just a matter of compensating for them by providing careful teaching and acceleratingtheir learning so they catch up. Most kids should be caught up with careful teaching by the third grade.

Sara Jane Cross, a 75-year-old kindergarten teacher in St. Petersburg, Fla., said she knows that some students come from homes in which education is a priority, but that some of their classmates do not.

"You don't know who you're going to get in a classroom -- what type of child, what kind of home," Cross said. "You can't expect them to keep up with children who come from fine homes."

I think it's time for Sara Jane to retire. It's easy for schools to predict the type of students they are likely to get based on the community demographics. Then it's just a matter of giving the incoming students a pre-test to determine where the teaching should begin and then to get on with it.

We don't expect them to be able to keep up with the affluent kids completely. However, we do expect them though to at least reach an acceptable proficiency level which should be well within their ability with effective teaching.

For now, the law focuses only on the building-block subjects of reading and math. Schools must test children in those subjects in grades three to eight and once in grades 10 to 12.

On that front, many parents and teachers agree, and they aren't happy. In their view, schools have had to narrow their focus, excluding other subjects and creative learning.

"Virtually every parent I know feels the schools are educating to the two subjects they are testing," said Mitchell Stiers, a father of three children in Cedar Falls, Iowa.

The entire narrowing issue is a classic false dilemma.

Schools haven't had to do anything. Some schools have chosen to narrow the curriculum because they can't teach reading and math effectively. The narrowing is caused by school failure. If the teaching were more effective, there'd be no need to narrow anything.

I wonder if the poll included a question on whether parents or teachers believe that it is the school's primary responsibility to ensure that students be able to read and do math proficiently? I have a feeling the yes answers would have been off the board.

April 19, 2006

Reading Mastery III Sample Lesson (Part 2)

In Part 2 we'll analyze the sample lesson from Reading Mastery III I wrote about in Part 1 of this post.

Let's now take a detailed look Lesson 68 (pdf) from Reading Mastery III.

Word Practice

Lesson 68, like all lessons in RMIII starts out with word-attack skills practice (exercises 1 and 2). In exercise 1, the students learn five new words in column one. These words will show up in the story in this lesson and/or subsequent lessons. Basically, this is prelearning of the hard words that the student will soon be reading in stories and other exercises.

In exercise 2, the students review six additional columns of hard words. Most of the words in these columns have already been reviewed in previous lessons, so they are familiar to the student. This is distributed practice. Some of these words will be practiced again in subsequent lessons.

Vocabulary

Five vocabulary words and their definitions are taught. Students know that these words will be used in upcoming stories and exercises and their meaning will have to be known in order to answer reading comprehension and exercise questions. For example the words foul, echo, and outcome are used in the story presented in this lesson and the students must know the definitions of these words to answer the reading comprehension questions presented.

Group Reading

Next comes the first group reading portion-- exercise 5. In this exercise individual students are called upon to read a few sentences from a a brief article on Facts about Coconuts which teaches various facts about the eponymous coconut. Again, these facts must be retained by the students since they will be used in the upcoming story, independent workbook exercises, and repeated in many subsequent lessons. As the students read the teacher calls on individual students to answer the reading comprehension questions listed.

In previous lessons the following selections were taught:

  1. Facts about Ocean Liners (Lesson 62)
  2. Lifeboats (Lesson 63)
  3. Facts about Ocean Water (Lesson 65)
  4. Facts about Islands (Lesson 66)
  5. Facts about palm trees (Lesson 67)
As you'll see from the upcoming story and independent work exercises, the facts taught in these articles will be used by the students and must be known by them

Reading: Decoding

In exercise six, the students read aloud the story
Linda and Kathy Find Some Food. The purpose of this exercise is two-fold: the students are to demonstrate accurate decoding and answer reading comprehension questions.

For the decoding portion, students are selected to read 2-3 sentences aloud until they reach the
* 2 Errors * signal. If the students haven't made more than two decoding errors by this point, they continue reading the rest of the story. If they've exceeded the two error threshold, they read the story again from the beginning until they can. This stresses to the students that their reading is expected to be careful and accurate. By the end of the story, the students must finish with less than 12 errors. If the error limit is exceeded the story is re-read.

Once the story has been read successfully within the decoding error limit, the teacher re-reads the story and asks the reading comprehension questions A-U presented. The questions relate to inferences drawn from the story and knowledge of taught vocabulary words and fact articles, among other things.

These first six lessons should take about 35 minutes to complete.

Independent Workbook Work

After the teacher directed exercises 1-6, the students begin independent workbook items in exercise 7.

Items 1-4 review facts from the coconut article presented in exercise 5.

Items 5-9 review items learned in the story read in exercise 6.

Items 10-17 review items learned in previous lessons.

  • Item 10 reviews facts about palm trees taught in lesson 67.
  • Item 11 reviews estimating weights taught and reviewed in lessons 39-66.
  • Item 12 reviews the word families for the words walk and jump taught in previous lessons.
  • Item 13 reviews lifeboat fats taught in lesson 63.
  • Item 14 reviews facts about palm trees taught in lesson 67.
  • Item 15 reviews globe and direction facts developed in lessons 20-64.
  • item 16 reviews spider facts taught in lesson 62.
  • item 17 reviews facts about warm-blooded animals taught in lesson 57 and will be review again in lesson 69.
Clearly there is a lot of distributed practice taking place in the independent exercises. Students are permitted to review today's stories for items 1-9, but are not permitted to review previous lessons for the answers to review items 1-17. They are expected to have mastered these.

The independent work exercise should take students 20-30 minutes to complete.

Workcheck

In exercise 8, the teacher reviews the workbook exercises and the students correct their errors. If any student exceeds the error limit, additional time is scheduled later that day to firm the student up on the items he missed.

The workcheck should take 8-10 minutes to complete.

The entire lesson should take about an hour and fifteen minutes to complete.

There are 140 lessons in RMIII. Every five lessons the students are tested on their decoding accuracy and fluency. Students are required to read a passage from the previous lesson in under a predetermined time limit with less than a predetermined number of errors.

That's it. The entire program in a nutshell.

After the lessons are completed, the teacher can teach a series of lessons based on two novels the students read using a similar but less structured format. Or, the class can proceed right into RMIV.

Reading Mastery III Sample Lesson (Part 1)

Here's a link to Lesson 68 from Reading Mastery III.

The Lesson is from the Teacher's presentation book and includes the relevant pages from the students' textbook and workbook. This lesson would be covered in December of third grade for most students who began learning to read in first grade; however if the students went through the reading Mastery I/II fast Cycle, which covers two years of reading instruction in one year, this lesson would be covered in December of second grade. Most kids of above-average intelligence or better who received a typical pre-K language exposure are capable of passing the Fast Cycle course.

Lesson 68 is typical of the lessons in RMIII and is representative of what transpires in a typical direct instruction language class in 2nd or 3rd grade.

RMIII lessons are designed to be taught in three parts:
  1. Exercises 1-6 are taught directly by the teacher within a 35 minute time period.
  2. Exercise 7 is an independent workbook exercise that the students complete at their seats after the teacher directed portion. Students are given between 20-30 minutes to complete the exercises.
  3. Exercise 8 is the work check portion in which the teacher and students check the workbook exercises and the teacher provides corrections to students who made errors. The workcheck portion should take 8-10 minutes.
Additional time may be necessary after the lesson to firm-up students who made too many errors during the lesson. Students should be firm on all the material presented in the lesson by the end of the school day.

I'm going to analyze the lesson in detail later today, but wanted to make it available for anyone who is interested in taking a look. The link to part II will be here.

April 18, 2006

Let's Cut to the Chase

Here's yet another article in the Seattle PI about the math wars. It raises the same old questions, gets the same old responses from the players, and finishes up with the same lack of answers. What's the point?

Isn't it about time for some real investigative journalism?

Intead of churning out the same hackneyed stories, why can't our intrepid journalist attempt to get some real answers?

Let's see how it might play out. I'll play the intrepid journalist who wants to write a piece on the elementary mathematics war.

Both sides claim to teach math better. But, who's right? Talk is cheap.

One way to find out would be to test the outputs -- the fifth or sixth grade students who just finished the elementary math programs. The good thing about elementary math is that it's cumulative and its main purpose (besides learning basic math skills) is to prepare kids to take algebra. Certain basic math skills--manipulation of fractions, understaning of decimals and place value, problem solving skills, and the like--need to be mastered by students and are indisputable prerequisites to algebra. We just need a quick and easy exam that tests these skills.

According to international tests, Singapore has the best prepared students in math. We could just use one of their readily available placement tests, fifth or sixth grade (take your pick), and use it as a guide to what good students need to have learned in elementary math.

Another option would be the sixth grade posttest in Connecting Math concepts--another good measure of what kids need to know at the end of elementary math.

Personally, I'd go with the CMC posttest with a sprinkling of problems from the Singapore Math tests to round out the skill set.

With my customized test in hand I'd go out to interview teachers from both sides of the debate. Let them blather on about why they think what they do is best. Let them talk all they want. Then I'd spring the trap.

Show me all the fifth or sixth graders that went through your program from start to end, and who didn't get any outside tutoring or help, we're going to test them and see what they really learned. What they learned is a reflection of what you taught and how well you did it.

That puts the end to the debate right there. We'd have a winner and a loser.

Actually, it probably wouldn't. This is education, so the loser wouldn't just admit defeat and grudging adopt the winner's program. There'd be talk about integrating the features of both programs because you "need both." Then the loser would continue what he was doing all along with a few minor adjustments and call it a "balanced" program.

Too bad for kids.

April 17, 2006

Compare and Contrast

From the article Math comes with its own problems in The Seattle Times we learn about a math student with shaky skills:

Jeremiah Pilkington knows the drill: First the butterflies in the stomach. Then the frustration.

"I sometimes would feel some people are smarter than me," said Pilkington, an eighth-grader at Bellevue's Tillicum Middle School.

The student didn't learn math. He feels anxious, frustrated, and dumb. Clearly, at this point the student doesn't like math much. Feeling dumb, leads to a lack of self esteem and eventually causes a student to act bad instead of look dumb in front of his classmates.

But then ...

But Pilkington, who until recently was behind in math, pulled a near-perfect score on his most recent test. His secret? Breaking problems into pieces to see what was really going on behind the numbers.

"Then I understand," Pilkington said. "It's actually a really cool feeling, because then you know what you are actually doing."

The student then gets taught how to do math properly. Lo and behold he now has a "really cool feeling" about hmself. He likes math now. He is motivated to learn now. And, to boot, he knows more math.

Complete turn around in student motivation. The only difference is better instruction.

So what did we learn:

Students taught math poorly don't like math, feel anxious and frustrated, feel dumb, lack self esteem, and are unmotivated to learn.

Students taught math well like math, feel really cool and smart, have self esteem, and are motivated to learn.

I'm shocked; shocked I tell you, because all I ever hear from teachers is that the reason kids don't learn is because they lack motivation. It's the student's fault she didn't learn.

It's never the instructor's fault for failing to teach. And yet, they want us to treat them like professionals.

April 13, 2006

Education Journalism 101

In today's St. Paul Pioneer Press we have a good example of a gullible journalist uncritically buying the educrat party line. The article is entitled "Pawlenty's math instruction goal might not add up" followed ominously by "Educators unsure if algebra proposal can work in practice." Hopefully, you see where there is going already. Let's take it after the introductory blather.

This is Gov. Tim Pawlenty's dream classroom. In a plan first touted in last month's State of the State address, Pawlenty said he wants all eighth-graders to take Algebra 1 and high school students to pass Algebra 2 before graduation, a move he hopes will strengthen student math skills across Minnesota.

Seems like the governor has a reasonable enough plan. Taking algebra in eighth grade means finishing elementary mathematics by seventh grade. Easy enough. I know at least one elementary math program that can be completed by the end of sixth grade that has an excellent track record preparing almost any student for algebra who attends class regularly. This gives even the slacker educator an extra year to prepare students for algebra.

But educators say it's easier said than done.
Especially when you don't know how to do it in the first place.

More rigorous math instruction in middle school and high school years means boosting math instruction in elementary school. And that may mean more training for elementary teachers, whose education often is more focused on language arts than math.
A tacit admission that elementary math instruction isn't very good as it stands. And, apparently, it's being taught by teachers who aren't qualified to teach math. This explains the next paragraph.

Raising math expectations also means tackling math-anxiety issues with students — and possibly their parents.
You know who has has math anxiety issues? kids who haven't learned math very well. You know why kids don't learn math very well? Because they weren't taught math very well by their school.

And, parents seem to have math anxiety too. I'm not sure why parents' anxiety is even relevant. Parents send kids to school so math teachers, who presumably know math, can teach their kids math. I bet this is a set-up for soemthing ...

Inside her Apple Valley classroom, Hoffman said she doesn't know if it's feasible. To make it work, math teachers would need longer class periods, smaller class sizes and aides to help kids who are struggling. And parents would play a role, she said.
I knew it. The parents are somehow implicated in all this.

The underlying assumption is that they are doing the best they can already. And, the only way to improve is to 1. increase the time of math classes (at the expense of what?), 2. decrease class sizes (more money), 3. hire more aides (more money) , and 4. have parents play a role (foolish).

Of course, in practice none of these ideas has proven to reliably increase student achievement in the absence of effective math instruction. But, if you had effective math instruction in the first place, you wouldn't need any of these ideas to make it work (with the possible exception of more time for the lowest performers). Chicken and egg.

"The rigor has to start in the elementary schools and carry through to the middle schools or they will not be ready for algebra by eighth grade," Hoffman said. "And rigor includes studying at home and parental support. If this is going to fly, we need parents involved at home."

Students across the nation are falling behind when it comes to math and science, and many argue something has to be done to increase the rigor.
They have the kids over six hours a day, 180 days a year, and that's not enough to teach elementary math. Yet, parental support and involvement is still necessary. Sure, great if you can get it. But, the kids who need the most parental support are least likely to have parents who are able to provide effective support. That's why we send these kids to school in the first place, isn't it? Because their parents do not have the ability and/or are unwilling to teach them. The assumption going in should be that no parental support will be forthcoming. The instructional design should be premised on that.

Finally, where is the research that shows that parental support is a prerequisite to learning math? In fact, the research shows the opposite. Parental support isn't necessary. We spent hundreds of millions of dollars on Project Follow Through to find out that the best performing instructional program required no parental support and one of the worst peforming ones actually wasted good money trying to educate parents so they could provided better parental support. Look at the math scores in Table 2.

About one-third of students who graduate from Minnesota high schools and go on to public colleges and universities in the state must take some sort of remedial course in college. At the U, 99 percent of those students were enrolled in a developmental math class. At state colleges and universities, about 82 percent of those students were in math as well.
OMG. These are the college bound kids. Imagine how little the non-college bound kids learned.

"I don't want to paint a doom-and-gloom picture," Seagren said. "I think our kids have enough fundamental strength to bring them to the next level."

Sounds like "blame the student" to me. The kids choose whether they learn or not. It couldn't possibly be that the schools could somehow affect learning and lack thereof.

"It sounds like a middle school and high school issue, but it's really an elementary problem," said Hagelberger, a member of the board of directors for the National Council of Teachers of Mathematics. "But in order for this to work, for students to take algebra in eighth grade, they need a rich foundation in math to be successful."
Yikes. I think I just agreed with the NCTM. I'm sure the NCTM means something a little different than what their saying. I'm thinking their definition of "rich foundation in math" probably means a something other than what I consider a rich foundation in math -- like learning math instead of learning about math.

"I think it's typical for elementary teachers to be language-oriented rather than math-oriented," said Rochester Superintendent Jerry Williams. "A big part of elementary education is teaching kids how to read, so that's natural."
I thought one of the three Rs included math. So, a big part of elementary education (a third, if the 3Rs are any guide) is supposed to have included math already. I wonder why it didn't? Could it be that they've been wasting a lot of time with equally bad reading programs and that not teaching kids how to read tends to be much more visible?

State officials and educators also say parents and community members need to get revved up about math and science. Seagren pointed out that despite the evidence that kids are falling behind in these subjects, many parents believe their kids are doing just fine.
Where'd they go and get an idea like that? Rampant grade inflation? Social promotion? Redefining math and reading to be something other than math and reading?

"I think, as a culture, we're math averse," Seagren said. "But now our world is changing.''
One of the down sides of not teaching math very well is that people start hating math.

Then the journalist asked a few kids what they thought:

Cory Patel was torn. He said if he has a good teacher and some additional help, he does well in math. But if he doesn't, it can be more difficult.

"I think it's fine where we are, but I guess upping the standards wouldn't be a bad thing," Patel said. "For us to take algebra, I think they need to condition us earlier so students are ready."

Little Corey got it right. Math isn't so hard as long as you're taught well.

Therese Harrah is in algebra this year and thinks about half her classmates could take the class. She also thinks there's a big fear of math out there that students could overcome by taking good notes, following directions and taking the problems one step at a time.

"People think it's really hard because of all the equations, variables and exponents," Therese said. "But as soon as you learn how to break down the solutions in different parts, it's not that hard."

Poor Therese has already bought the "blame the student" meme. It's the kids' fault they didn't do well. They need to take better notes, follow direction better and solve the problems systematically. Isn't this the teacher's job? To, you know, teach how to do this stuff?

Algebra really isn't very hard. So long as you've mastered elementary math. Then it's just a matter of "learn[ing] how to break down the solutions in different parts" according to little Therese. Of course, you learn how to do this by being taught how to do it. It's only hard if you're not taught how to do it well. And, if you're not taught it well, you can forget about calculus. And, if you can't do calculus well, then you can forget about a career in math, science, or engineering -- i.e., the high-paying careers.

This is one fine piece of journalism. Only interview players from one side for a nice biased self-serving view. (Though, kudos for throwing in the remediation rate stats.) You can always provide balance by interviewing a few kids too and see what they think.

April 12, 2006

Out of the Stratosphere

I like the professional policy blogs, such as EduWonk, This Week in Education, NCLBlog, and The Quick and The Ed. They excel at analyzing and reporting on general education policy issues like charter schools, vouchers, NCLB, drop-out rates, and the like. Though, occasionally, they seem to value each other's (and the other think tanks') opinions a wee bit more than is healthy. (They should heed the valuable advice of Mr. Wolf).

They're generally reliable when it comes to education policy that's up in the stratosphere, but tend to be just as unreliable as the rest of us when it comes to street level issues. Let's take a recent post on History is Not Math at The Quick and the Ed:

Although there are still some folks--primarily in ed schools--who haven't bought into it yet, there's actually a pretty clear, well-defined research consensus about how children learn to read, and what educators need to do when for kids to become proficient readers.

Cetainly there is some truth to this statement, but it doesn't probe deep enough to the real problem we have in early reading instruction. Everyone pays lip service to the National Reading Panel's findings as to what well designed reading programs must include. To wit: phoneme awareness, phonics, fluency, vocabulary, reading comprehension. Almost every commercially available early reading instructional program claims to have all these critical components. Problem is most of them stink.

The NRP was a naive observer when they looked at the valid research on early reading instruction. It focused on the superficial attributes of the successful reading programs and failed to attend to the critical details of what makes good instructional programs effective. This is because no one on the NRP team ever designed an effective reading program or knew what to look for.

Dead reckoning.

So when The Quick and the Ed concludes:

In both reading and math, we can study what and how kids need to learn in order to become proficient because--calculator debates aside--there's some general agreement about what, ultimately, kids need to know. Everyone agrees that kids need to learn how to read, and there's not a lot of debate about what that means.
they seem not to recognize the very real problems that still exist today in early reading instruction. Sure, everybody can agree on what "kids need to learn how to read" because the NRP made its recommendations so vague and meaningless that even ineffective programs can claim to comply.

Effective education requires close attention to detail and adherence to strict quality control measures.

If the professional education policy blogs want to have any impact on education they need to stop worrying about meaningless generalities and stratospheric policies and start attending to the street level details where all the problems are occurring. Maybe, Alexander Russo is on to something when he's written that most education blogs aren't very good.