February 12, 2007

Closing the Achievement Gap

Bad news.

It's not going to happen. Not anytime soon. Not ever.

To raise the achievement gap, we need to raise the performance of the lower performing groups. (Another way is to drag down the higher performing groups. But that's just silly.)

There exist two schools of thought.

The first school believes that the achievement gap is caused by factors external to schools, such as poverty, racism, discrimination, and the like. The theory goes that if you eliminate or ameliorate these factors, you can raise the performance of the low performing groups up to the level of the higher performing groups. Sounds good in theory, kinda like communism sounded good in theory. In practice, however, results have been elusive and by elusive I mean non-existent. Unfortunately, no amount of wishful thinking in editorial pages seems to be able to get this theory past the the rainbows and magic stage of development. Needless to say, this is the dominant view of most educators.

The other school of thought is that the poor performance of the lower performing groups is the result of inferior instruction. Improve the instruction and performance will also improve. One problem with this school of thought is that improving instruction has also been an elusive task. Most efforts to improve the performance of low performers via improved instruction have failed miserably. You can count the number of successful programs that achieve consistent, reliable, and educationally significant results on your fingers of one hand. You have DI, SfA, and a few others. The "problem" with this improved instruction approach is that improved instruction won't just raise the performance of low performers it will raise performance across the board, so the achievement gap isn't going to be reduced in real terms any time soon. But, don't take my word for it take Zig Engelmann's, the creator of one of the few instructional programs that has gotten positive results:


Following the remodeling [in 1972], we opened a learning center, which was designed to serve hard-to-teach children and school failures. One of the earliest groups, however, was not low, but was composed of six preschoolers whose parents were professionals or professors. One student was Wes' youngest son. We worked with these children as four-year-olds and five-year-olds. They went through the reading and math programs as fast as they could go at mastery, which was frighteningly fast. (There was no need for the language programs because these children were very bright.) Even though they worked for only a little more than an hour a day, they went through all four levels of the programs we had for Follow Through classes. Before they
entered first grade, they performed on the fourth-grade level in both subjects. And they loved school.

We never published anything about the performance of these children, largely because the group had only six children, which meant that experimental purists would "question" the results. (Usually, at least 15 experimental children are needed for establishing outcomes that are recognized as valid.)

Although these children were awesome, their performance showed a critical difference between their potential and that of the at-risk child. When these advantaged children came to the third and fourth levels of the reading program, where the material becomes entrenched in and decorated with sophisticated language, they did not slow down. The profile for the at-risk child is different. Performance slows considerably when they reach the vocabulary-rich transition. They have parallel problems with math when the word problems become more substantive than a few pared-down sentences that present necessary information in a "familiar" format.

... [W]e don't have to worry as much about the performance of higher performers. They will tend to learn from teaching that is hideous, as many programs for the talented and gifted demonstrate. These programs provide instruction that appears to be purposely designed to teach, explain, and develop skills in the most circuitous and confusing manner possible. Certainly it's cruel to subject students of any skill level to such instruction, but in the larger scheme, far less cruel than subjecting at-risk children to certain failure. Although it would have been possible for us to work with both populations, we reconfirmed the decision not to work with higher performers but rather to show the degree to which at-risk children would catch up to higher performers with careful instruction. We figured that teaching higher performers effectively is so easy that in time, those who educate them would learn how to do it effectively. It certainly hasn't happened yet. But we felt that we needed to work with the lower performers simply because it is not easy and teachers don't know how to do it. In fact, we believed that if we didn't do it, it wouldn't happen because nobody in or out of Follow Through (with the exception of the University of Kansas) was close.

(p. 2-4)

This finding was later confirmed in a subsequent study;

To show the degree of acceleration that was possible with students at or above average, Doug Carnine and I did a formal study, which appeared in our 1978 Technical Report to National Follow Through. Thirty children in a Springfield, Oregon elementary school went through our reading program at an accelerated rate. These were not children with extremely high IQs, but all but two met the district’s criteria for "higher performer," which was that they performed at or above the district average when they began the first grade. (The district had no kindergarten.) The two exceptions were low performers who were added because the teacher felt they could benefit from the program. During grades 1 and 2, the children had daily reading lessons of about one-half hour per day and devoted another half-hour to independent work. The teaching was conducted by trainees in our practica. The classroom teacher was a star, but she did very little of the teaching. She made sure, however, that the trainees performed very well.

At the end of their second grade, children read at the middle fourth grade level according to the Stanford Achievement Test. They performed on the fifth-grade level of an oral reading test. The top ten children received a fourth-grade test that measured speed and accuracy. (We could find no test for the second or third grade.) Students performed on the seventh-grade level.

The children were not taught science as a subject, but level 3 of our reading program has stories that are heavy in science content. The class performed at the fourth-grade level in science.

(pp. 6-7)

These results shouldn't be surprising to anyone. If the quality of instruction is held constant, smart kids will always learn more and faster than their dull peers. And yet, you can't go a day without reading how the goal of education is to eliminate the achievement gap. The allure of rainbow and lollipop solutions is strong. but, based on the evidence at hand, it's not going to happen. So, get over it.

27 comments:

Parentalcation said...

Shhhh... they will hear you.

You underestimate the ability of the system to create enough enrichment programs to slow down the fast learners.

Perhaps I should open up a DI tutoring program. 3 to 4 levels of acceleration should be worth at least a couple of hundred dollars per child.

Anonymous said...

Even if "they hear you", it may not slow down the population of fast learners all that much. Private school and home school may just become even more attractive to the parents of these children.

Which might *increase* the achievement gap if fast learning minority kids don't have parents who can afford to leave the public school system.

-Mark R.

NYC Math Teacher said...

I teach many "fast learning minority kids" who are stuck in heterogeneous (with respect to math skills) classes. While they continue to outperform, what of the opportunity cost? What of the lost advancement opportunities while I dwell on multiplying simple decimals? This ridiculous ongoing experiment in egalitarianism is maddening to me.

SteveH said...

"... teach many "fast learning minority kids" who are stuck in heterogeneous (with respect to math skills) classes."

What do these kids and their parents say to you, if anything? Does the "system" keep these kids from better opportunities because that would not be "fair"?

Tracy said...

Well it sounds like we can decrease the size of the performance gap.

Just give the low-performing kids a good curriculum and effective instruction system and give the high-performing kids the current normal system.

I don't regard this option as ethical, but it's certainly doable.

MassParent said...

The data I've seen suggests that the gap did close between the 1950s and 1970s, by perhaps one-third of a standard deviation. Since then has stagnated.

Ken, I thought you argued against constructivist curriculum because it is not well suited to slower learners. But here you're saying the curriculum does not matter as far as the gap is concerned. I would imagine that different curriculum and teaching methods could result in differences of the distribution.

NYC Math Teacher said...

steveh,

The kids and parents say nothing. The kids are either blissfully unaware or too young (6th grade) to speak up, and the parents are either disengaged or not fully aware.

Here in NYC, the ed school philosophies hold sway in the public schools so, yes, I believe that the system is at fault. I think the busybodies honestly believe that equity is king and that the high performers both teach and learn from the struggling students. They're just too blind to realize they're wrong (or too stupid to realize that there is no empirical evidence top support their "theories")>

Parentalcation said...

Massparent,

Actually I think Ken argues against constructivism because it's inefficient and does a poor job of educating slow learners to grade level.

Actually I think Ken should clarify himself a bit here. If achievement gap means the percentage gap between students scoring proficient on some arbitrary measure of performance, then the gap could conceiveably be closed. If the gap is defined as the average test scores of one group to another group, then the gap will never be closed.

KDeRosa said...

Massparent, that was a one time gain caused by increased nutrition when blacks were lifted out of poverty. Nowadays pretty much everyone is the US, including the "poor," is super-nutriated.

Improving the curriculum will increase student performance across the board. So this is a good thing and a good reason why you'd want to switch to a better curriculum. But, improving the curriculum won't improve the lot of low performers more than higher performers. In fact, it seems just the opposite--higher performers are better capable of taking advantage of the improved curriculum. So, if the distribution changes, it seems like it'll change in a way that widens the gap.

Michael said...

Nowadays pretty much everyone is the US, including the "poor," is super-nutriated.

And part of the reason for that is the meals provided by public schools. Of course, it's not as good a situation as you think, particularly in parts of Texas.

SteveH said...

"I think the busybodies honestly believe that equity is king and that the high performers both teach and learn from the struggling students."

"equity is king"

Yes, even in our affluent, full-inclusion public schools. They talk of individuals, but that's not primary.

When "they" talk of closing the achievement gap, they are talking about statistical groups, not individuals. They should be talking about individual educational potential gaps, not group gaps.

Schools could help individuals right now, based on merit or hard work, but they think that would damage the group. They have to pretend that this is good for the the better students.

There is also the problem of teaching environment - separating the kids who care and try from those who don't. Many shcools won't do this. This is not so much of a problem in our affluent schools. Perhaps NYC Math Teacher can comment on his neck of the woods. It seems to me that one could focus on an environmental gap as a separate issue from an academic gap.

My focus has always been on individuals as opposed to groups. Schools have a hard time with this. How can you get the best out of individuals if you don't treat them that way academically? This is much more than learning styles.

Dr. P. said...

“The data I've seen suggests that the gap did close between the 1950s and 1970s, by perhaps one-third of a standard deviation. Since then has stagnated.”

There is a very detailed debate on this over at AEI.

From the AEI promo: “For decades, the difference in the test scores of blacks and whites on the SAT, National Assessment of Educational Progress test, Armed Forces Qualification Test, and traditional IQ tests has been a vexed issue for American educational policy. Two of the leading scholars of this controversial topic, James R. Flynn of the University of Otago (New Zealand) and Charles Murray of AEI, will debate the causes of the difference, its implications, and recent trends. New studies of the subject by Professor Flynn and by Mr. Murray will be available for distribution.”

Basically Flynn claims that the gap will continue to decrease, while Murray says no. Both make good arguments, but I think Murray has the better of it. This is a refreshingly intellectual debate free of the usual emotional baggage associated with this topic. Watch the whole video if you have the time.

MassParent said...

Ken, I suspect the gap in measured IQ for blacks is a bit more durable than the measurements at Ellis Island. But forever is quite a long time.

The IQ gap narrowed during the the period of desegregation and the war on poverty; and since then, during the period of diverging incomes and racial laissez faire, has stagnated. I don't believe you can reach an informed conclusion that the gain during the war on poverty was a one-time thing. You might be able to disagregate data since the war on poverty for selected individuals, to make an argument like that, but the credibility of s study like that would not be high.

KDeRosa said...

and since then, during the period of diverging incomes and racial laissez faire,

Racial laissez faire? Since when?

The fact remains that traditional lower performing groups perform about as poorly in affluent schools as they do in inner city schools nowadays. And, there's no more gains, with the exception of some smal isolated groups, you can suck out of increased nutrition at this point. You might be abale to eke out a few more IQ points if more mothers in this group breast fed. But, that's pretty much it. Where are the extra IQ points going to come from, unless the intermarriage rate increases?

tc said...

How do we know that nutrition was responsible for the b-w gap narrowing in the 70s? It seems to me we still don't know what's caused any of the seeming changes in IQ over time. Flynn's latest work (which he talks about in the AEI link above) has found that the b-w gap among children has continued narrowing throughout the 90s (but not for adults).

KDeRosa said...

See here.

I believe the nutritiona theory is the consensus view at this point because malnutrition, especially prenatal, has been shown to cause lower IQs.

Ari said...

Children may be getting enough calories and macronutrientsbut there still remain problems with the micro. For example not enough omega 3 and too much transfat may not be too good for the brain.
Even the pregnancy diet may have an impact
http://news.bbc.co.uk/2/hi/health/4631006.stm

4trogan said...

Say both lower and higher performers were given Cadillac DI right from the start. Would there ever be a time when the gap would close, say in high school? In other words, is there some benchmark the students would reach whereby their solid foundational skills changed from being the goal, to the means by which each student now had the tools to achieve... whatever the student wanted to move on to?

Instructivist said...

Some minority kids thrive beyond imagination with a good curriculum, good instruction and a calm environment. I see this first-hand since I teach these kids. What defeats many kids are cultural (no culture of learning) and behavioral factors. The behavioral problems then spill over and interfere with the learning of the willing.

In addition to a good curriculum and instruction, what is needed is getting rid of the heterogeneous classroom.

Parentalcation said...

4trogan asked:

"Say both lower and higher performers were given Cadillac DI right from the start. Would there ever be a time when the gap would close, say in high school? In other words, is there some benchmark the students would reach whereby their solid foundational skills changed from being the goal, to the means by which each student now had the tools to achieve... whatever the student wanted to move on to?"

Short answer: NO

High IQ learners will always be able to learn new concepts at a greater rate. I also suspect there is a peak that low IQ individuals would reach where the complexity of further knowledge would slow down the learning.

Additionally, once adequate content knowledge is acquired, the next step is the "creative" relating of different diciplines of work and eventually creating new content knowledge.

NYC Math Teacher said...

steveh,

There is also the problem of teaching environment - separating the kids who care and try from those who don't. Many schools won't do this.

In short, my school does this to a very limited extent. I teach classes containing children with extreme behavioral problems who don't care and who infringe on those who do care.

Instructivist said it best above regarding behavioral problems and the need to eliminate heterogeneous classes.

4trogan said...

There is only one problem with eliminating heterogeneous classes in my opinion, however, and that is that kids with the milder behavioral problems will not improve being around the kids with the moderate to severe behavior problems. Maybe it would work if there was some fantastic, systematic way of grouping the kids by behavior, and then all these kids got behavioral interventions based upon Applied Behavioral Analysis by truly skilled professionals... but that's not bloody likely in the public school system.

I will also say that my own autistic son behaves much better around more normal acting kids. When he's around kids that have their own issues, he tends to let himself "go" more rather than the checking of his own behaviors he does around the normal kids. ("neurotypical" is the PC word to use; but since I'm a parent of one of these kids I like to just say "normal"... Is that wrong?)

NYC Math Teacher said...

4trogan:

For the record, when I referred to heterogeneous classes, I mean as measured by ability and without behavior as a factor.

I believe that behavior problems should be handled differently, with schools making it easier to remove students who impede the learning process, whether they are in advanced classes or slower-moving classes.

Ari said...

Instructivist said...

What defeats many kids are cultural (no culture of learning) and behavioral factors.

I know for myself it is very hard even at my age to learn when the rest of my family is only interested in watching tv. Even the best instruction can't compete with tv, video games, movies, pop music and everything else out there.

Also, the kid who gets whatever he wants, (fast food, toys, etc) is simply not going to stand for the hard work required for mastery. Neither will the parents, who only want to make their kids "happy."

TMAO said...

The problem with Kozol and his ilk is not that they are somehow "wrong" in their documentation of the horrid conditions of many urban schools, or the way in which ethnicity correlates to the quality of your school's physical plant -- they aren't -- the problem is that they conflate the notion of "closing the achievement gap" with closing all differences between rich and poor. Their approaches that demand we do more on the governmental/ economic level are not without merit, but they set far too high a bar for success.

I can live with an achievement gap as long as the achievement of the poor is acceptable. It isn't. Part of the analysis of achievement gap closing that is often ignored is the fact that performance is going up, however you splice the numbers, for pretty much everyone over the last few decades. That's good. Not good enough, not nearly good enough, but in itself a positive trend. I cannot imagine a point where the children of the wealthy perform equally to the children of the desperate poor, but I can envision a scenario, where the performance of the poor reaches an acceptable standard, at a minimum.

And that day, contrary to the Brothers K., can be made possible by the efforts of those who work in schools.

KDeRosa said...

TMAO, I couldn't agree more.

TMAO said...

Is there any one human described as having "ilk" as frequently as Jonathon Kozol? I want to know.