"We Don't Care About Data."
January, 2005

The huge gap in school achievement and later quality of life between minority and white students is usually explained by things that are largely irrelevant--culture, "race," family structure, the percentage of minority children in a school, socioeconomic status, students' self-perceptions and teachers' expectations.   I’d like to think that focusing on the wrong things is simple laziness--a person looking for keys under a streetlamp; it’s the easiest place.  But now I think it’s more because “social reformers” are happiest dealing with vague abstractions—their “big picture” of how “society oppresses the poor.”  This enables them to conjure up gaudy schemes (which elicit hormonal secretions.  The bigger the scheme the more important they feel.), to get long-running grants and high prestige positions, hire friends, write articles, and end up with nice retirement annuities.   

Let’s get serious about improving achievement.  You aren’t going to change anyone’s “race” or culture.  No “program” is going to raise children’s self-esteem and children’s and teachers’ expectations—for very long.  And we aren’t going to “make the distribution of wealth more equitable or equal”—even if we knew what that meant.  These sorts of efforts to create a “new man” and to revolutionize society almost always yield disaster.  [Think “Soviet Union.”]  But by then the reformers are long gone.  Besides, their kids are in private schools.   But don’t tell that to the reformers.  They’ll get testy.  After all, you're taking away their stock in trade—the false promise of edutopia—if we’d only give them power, money, our kids, and all the time in the world.

Reformers almost never consider the obvious.  What is closest to student learning is not race, social class, culture, school size, and all the other factors the reformers tout, but communication with the teacher—organized as instruction within a curriculum.   The reason poor kids don’t learn much in school is that they come to school less prepared and because most schools use curricula that are horrible (superficial coverage, illogical sequences, little built-in practice) and teaching methods that miscommunicate information. And there are tons of good data showing that well designed curricula and logically clear instruction can override the effects of social class, minority group status, and family background.

Follow Through
In the mid 1960’s, President Lyndon Johnson’s administration created Head Start—a large number of preschool programs primarily for disadvantaged children.  After a few years he also funded Follow Through, to see which Head Start models (curricula) yielded the most beneficial change.  Pretty rational.  Find out what works best and promote it.  Find out what fails and dump it.  That’s how they do it in medicine, engineering, and other serious professions.

That’s NOT how they do it in education.

Follow Through ran from 1967 to 1995.  It tested nine curricula--many of which are still used.  Follow Through involved about 75,000 children per year in about 180 schools.  Each model school was compared with control schools. 

Here’s a summary description of the models.

Cognitive/Conceptual Skills Models
Cognitively-Oriented Curriculum (High Scope Foundation).
This program (STILL widely used) was based on Piaget's theory of stages of cognitive development, and his assertion that teachers should be more like guides on the side rather than communicators of information.   

Florida Parent Education Model (University of Florida).
This program taught parents of disadvantaged children to teach their children. At the same time, students were taught in the classroom using a Piagetian approach.

Tucson Early Education Model (University of Arizona).
TEEM used a language-experience approach (much like whole language). It was based on the notion that children have different learning styles.

Affective Skills Models
Bank Street College Model (Bank Street College of Education).
This model emphasized learning centers that gave children many options, such as counting blocks and quiet areas for reading.  Much of the teaching was incidental as the teacher tried to follow children’s lead. 

Open Education Model (Education Development Center).
This model was derived from the British Infant School model. Reading and writing were not taught directly, but through stimulating a desire to communicate.

Responsive Education Model (Far West Laboratory).
This eclectic model used learning centers and students’ interests to determine when and where each child would be stationed. The development of self-esteem was considered essential to the acquisition of academic skills.

Basic Skills Models
Behavior Analysis Model (University of Kansas).
Developed by Donald Bushell, this model used a behavioral (reinforcement) approach for teaching reading, arithmetic, handwriting, and spelling.  Children received praise and tokens for correct responses. Teachers used programmed reading materials that presented tasks in small steps.

Language Development (Bilingual) Model (Southwest Educational Developmental Laboratory).
This model used an eclectic approach based on language development. When needed, material was presented first in Spanish and then in English.

Direct Instruction Model (University of Oregon).
Developed by Siegfried Engelmann and Wes Becker, this model used the DISTAR (Direct Instruction System for Teaching, Achievement, and Remediation) reading, arithmetic, and language programs. The model assumes that the teacher is responsible for what the children learn.

Here are some of the main features of Direct Instruction.
1. Direct Instruction focuses on cognitive learning--concepts, propositions, cognitive strategies.  It is not rote learning.

2. Brief (5 minute) placement tests are given to ensure that each child begins with lessons for which he or she is prepared. 

3. Children are taught in small groups.

4. The children sit in front of the teacher--close enough that she can see and hear each one.

5. Lessons move at a brisk pace.  This sustains children's attention and results in a high rate of learning opportunities per minute.

6. Instruction is organized in a logical-developmental sequence.  All of the concepts, rules, and strategies that students need in any lesson have already been taught.  In addition, what they learn in any lesson is used in later lessons.  There is no inert knowledge.

7. Knowledge (e.g., how to solve 4 + X = 12; how to sound out words) is taught directly and explicitly.  For example, the teacher verbalizes her reasoning process while demonstrating the strategy for solving an arithmetic problem.  This enables students to internalize the teacher's knowledge and become independent.

8. Instruction is aimed at mastery.  The group and each child is always "firm" before the teacher moves to the next exercise.   

9. Teacher-student communication has a common format from lesson to lesson.  This means that students need to attend only to the content of the communication, and do not have to figure out how the teacher is communicating.  The general format is Model, Lead, Test:
a. Model:  For example, the teacher says, "I can read this word the slow way.  Listen.  wh e n."

b. Lead:  This step is guided practice; teacher and students work problems, sound out new words, or read passages together.  For example, the teacher says, "Read this word with me. 
Get ready.  wh e n."

c. Test: Children now do the exercise on their own.  "Your turn to read this word the slow way.  Get ready..."

[More on this highly effective format later.]

10. Gradually, instruction moves from a teacher-guided to a more student-guided format. 

11. Direct Instruction would most likely be used at the beginning of some class periods.  The rest of a class period would be individual or small group work on generalizing or adapting what was learned to new material or problems.

[From G. L. Adams (1995).  Project Follow Through: In-depth and Beyond.  Effective  School Practices, 15 , 1, Winter. ]

Findings.  Which Curricula Did Good Things for Kids?  Which Curricula Made it Worse for Kids?
A major source of data was scores on the Metropolitan Achievement Test, the Coopersmith Self-Esteem Inventory, and the Intellectual Achievement Responsibility Scale.  The main results were as follows.

1. Children who were taught reading, spelling, and math with Direct Instruction were far superior in achievement to children taught with any other method in both basic and higher-order conceptual skills (e.g., problem solving).  Most of the other "innovative" models did far worse even than non-DI control schools.

2. Disadvantaged children taught with Direct Instruction moved from the 20th percentile on nationally- standardized tests to the 50th percentile.  In other words, Direct Instruction made them regular students in achievement.  However, the standing of disadvantaged children receiving some of the other (still used) non-DI curricula decreased relative to the rest of the country.

3. Children taught with Direct Instruction developed higher self-esteem and a stronger sense of control of their learning than did children receiving the other forms of instruction; this, despite the fact that some of the other curricula focused on self-esteem.

4. Follow-up studies showed that children (predominantly Black or Hispanic) who had been taught reading and math using Direct Instruction in elementary school were, at the end of the 9th grade, still one year ahead of children who had been in control (non-Direct Instruction) schools in reading, and 7 months ahead of control children in math.

Also, in contrast to comparison groups of children who had not received Direct Instruction in earlier years, former Direct Instruction students had higher rates of graduating high school on time, lower rates of dropping out, and higher rates of applying and being accepted into college.

Here’s a graph from the Washington Times. 

Notice that DI and Behavior Analysis—the two models that had clear objectives, taught in a logically progressive sequence, involved teachers focusing on exactly what they wanted kids to learn, communicated as clearly as possible, and provided practice to the point of mastery—did the best in all areas—how much kids learned, how they felt about themselves, and how much control they felt they had over their learning. 

Ironic.  The MOST teacher-directed approaches produced kids who felt that THEY were in control of their learning.  I suspect this is because they learned SO MUCH and so easily!

So, you think schools, districts, and states adopted Direct Instruction and Behavior Analysis?  WRONG. Instead, the Ford Foundation hired another team of statisticians to analyze the data that HAD been analyzed by ABT Associates in Cambridge, MA.  Apparently, the Ford Foundation, long a supporter of so-called progressive causes and programs, was not happy that the “progressive” ed programs (whole language, child-directed, self-esteem first, constructivist) not only were beaten by their self-created enemy (Direct Instruction and Behavior Analysis) but (as the graph shows) actually SUPPRESSED children’s growth.

The new statisticians made the claim that no model did any better than the others.

And THIS was the news sent throughout Edland.  “Do whatever you want.  They are all good.  And don’t listen to the people who say DI was the best.”

Result?  DI and Behavior Analysis were shunned for decades.  And the eduquacks kept training new teachers to use the models that Follow Through data had shown were next to useless and often destructive. 

You see, just as the grand social reformers presume that OUR society belongs to THEM (because they assume that they are much smarter than the rest of us) and is an object for them to experiment with, so the edureformers consider kids and their futures to be their “responsibility” (for they are SO much smarter than parents and teachers) and also their property.  And THIS is why the past 100 years in education is largely the history of experimenting with kids.

This has begun to change—as states have passed accountability legislation making districts raise achievement or else.  Also, No Child Left Behind and Reading First put pressure on schools to use curricula and methods that are shown to work—which narrows the field to Direct Instruction and programs that share its design features.

But make no mistake, the progressive eduquacks are alive and well.  This is their “hudna.”  They are doing what they have always done.  Waiting for a change in administration.  Then they will say, “WE’RE BAAAAACK!”