Tough Nuts

Can someone explain to me what all the hub-bub is over Paul Tough's NY mag article?

The generally reliable Eduwonk calls it the "[m]ost important education article written this year."

I must be missing something.

The Tough article is split into two parts:

[T]wo parallel debates about the achievement gap have emerged. The first is about causes; the second is about cures. The first has been taking place in academia, among economists and anthropologists and sociologists who are trying to figure out exactly where the gap comes from, why it exists and why it persists. The second is happening among and around a loose coalition of schools, all of them quite new, all established with the goal of wiping out the achievement gap altogether.

Tough gets the first part, the causes of the achievement gap, almost entirely wrong --no doubt because he relied upon the views of "economists and anthropologists and sociologists" from "academia."

Then he commits the sin of omission in the second part by focusing on one type of middle school school reform model, KIPP, which, though successful, requires students and teachers to put in long hours and a long a school year to make up for the poor teaching these kids received in elementary school.

But I'm getting ahead of myself.

Tough is 30 years late to the game with his stunning revelation that low-SES children are exposed to far less words at home as children than their higher-SES peers with the result being a vocabulary gap that continues to widen. Becker identified this phenomenon back in 1977 in "Teaching Reading and Language to the Disadvantaged—What We Have Learned from Research" published in the Harvard Educational Review.

In the analysis that follows, we conclude that schools systematically fail to provide instruction in the building blocks crucial to intelligent functioning, namely, words and their referents. Children from homes where there is strong adult support for refining the use of language are more likely to succeed in school than those from homes with less adult-child contact and adults with less education (Coleman, 1975; Freeberg & Payne, 1967; Glass, 1973).

Becker concludes:

The hypothesis that vocabulary-concept knowledge plays a major role in reading comprehension is supported in the research literature. In his review of research Carroll (Note 9) concludes that

much of the failure of individuals to understand speech or writing beyond an elementary level is due to deficiency in vocabulary knowledge. It is not merely the knowledge of single words and their meanings that is important, but also the knowledge of the multiple meanings of words and their grammatical functions. (p. 175)

Carroll also argues that vocabulary-concept knowledge is the key area of concern for improving reading comprehension for the economically disadvantaged, but he is quite aware that there is no easy way to make gains in this area. (emphasis mine)

Becker understood that it was the lower-IQ of these children and their parents that's at the root of the problem. Due to this IQ deficiency, these children picked up vocabulary knowledge at a slower rate than their higher-IQ peers and, unfortunately, "there is no easy way to make gains in this area" because it is extremely difficult to accelerate the learning of vocabulary in these kids due to the nature of learning vocabulary acquisition. (I wrote about the problem in detail here.)

In contrast, Tough draws this spurious conclusion.

Hart and Risley showed that language exposure in early childhood correlated strongly with I.Q. and academic success later on in a child’s life. Hearing fewer words, and a lot of prohibitions and discouragements, had a negative effect on I.Q.; hearing lots of words, and more affirmations and complex sentences, had a positive effect on I.Q. The professional parents were giving their children an advantage with every word they spoke, and the advantage just kept building up.

It's not just the exposure to vocabulary that is problematic; it's also that these kids are acquiring vocabulary at a slower rate than their middle class peers. This is a function of IQ which Tough is too quick to dismiss. After reviewing some more correlative "research" Tough draws this final conclusion as the summation of why the achievement gap exists.

Taken together, the conclusions of these researchers can be a little unsettling. Their work seems to reduce a child’s upbringing, which to a parent can feel something like magic, to a simple algorithm: give a child X, and you get Y. Their work also suggests that the disadvantages that poverty imposes on children aren't primarily about material goods. True, every poor child would benefit from having more books in his home and more nutritious food to eat (and money certainly makes it easier to carry out a program of concerted cultivation). But the real advantages that middle-class children gain come from more elusive processes: the language that their parents use, the attitudes toward life that they convey. However you measure child-rearing, middle-class parents tend to do it differently than poor parents — and the path they follow in turn tends to give their children an array of advantages. As Lareau points out, kids from poor families might be nicer, they might be happier, they might be more polite — but in countless ways, the manner in which they are raised puts them at a disadvantage in the measures that count in contemporary American society.

There is no evidence that if you give a low-IQ kid X, then you're going to get the same Y back that you get back that you get from giving X to a higher-IQ kid. Not to mention the sticky problem that lower-IQ parents have never been shown to be capable of providing X to their children and for that matter neither have schools. Again, Becker, nails the problem much better than the naive Tough:

These logical and empirical analyses clearly point to a problem for educators who strive to teach reading comprehension to all children. The data suggest that school programs do not systematically build vocabulary-concept knowledge. Current programs are structured to teach middle-class children or children who, to a large extent, are taught oral-language comprehension at home. We assume that this form of language learning is then transferred to reading comprehension at school. (emphasis mine)

At best, Tough has partially identified the problem facing low-IQ kids (vocabulary acquisition deficiencies), but fails to recognize that merely identifying the problem does not imply an instructional remedy to solve the problem. Tough never does attempt to find an instructional remedy to the vocabulary acquisition problem. Instead, he races off to find and analyze a school that has gotten some results with at-risk kids. The model he picks is KIPP.

Now I like KIPP and think the model has accomplished some impressive results. But, the KIPP model is not without its problems. KIPP is a middle school reform model. As a result, when kids enter KIPP they are far behind. Part of KIPP's solution is to the extend the instructional time offered to students by extending the school day, week, and year. This is not an ideal solution. Why wait until fifth grade to intervene? Why allow these students to be mistaught and not taught for their first five school years before intervening? Perhaps it becomes necessary to teach for long hours following such a delay, but perhaps KIPP is not teaching as efficiently as it could. Maybe if KIPP started in K, they could get away without having to increase the instructional time.

Other instructional reform models manage to increase the academic performance of at-risk children without extending the instructional time offered. For example, the DI model only requires about 3 hours of instructional time per day in reading, math, writing, spelling, and language to get kids performing at grade level by first grade and keeps them there until the end of elementary school- fifth grade. Kids coming out of the DI program in fifth grade can then be taught in a more traditional manner in middle school and high school without the need for increasing their instructional time burden.

In short, it may be that the KIPP model is not scalable as Tough suggests, but that does not mean that other reform models have the same infirmities. In this respect, the second part of Tough's article sets-up a false dilemma and a strawman, which he allows Rothstein to take a whack at, by suggesting that the KIPP model is the only successful reform model and by failing to note that some of the scalability issues inherent in the current KIPP model might be mitigated if KIPP started in elementary school like other successful reform models.

Then Tough concludes with a series of half-baked inferences from his discussion of the KIPP model. Here's the most egregious:

the reality is that even the best, most motivated educator, given just six hours a day and 10 months a year and nothing more than the typical resources provided to a public-school teacher, would find it near impossible to educate an average classroom of poor minority students up to the level of their middle-class peers.

There is absolutely no evidentiary support for such a conclusion. Counterexamples abound in the research literature. There are well over a hundred DI implementations that achieve KIPP-like results in a normal school day and year with typical teachers found in poor minority neighborhood under typical funding levels. Tough was either too lazy or too dishonest to go out and fund these schools or schools like them. No one wants to admit that most schools are adequately funded today. Now, that story would be worthy of being story of the year.