Return to conference
I'd like to start by thanking Dave for his gracious invitation
to speak at the conference. And I'd also like to thank Dave, Aida and Stan for putting
this conference on. Having done conferences a number of times myself, I have an
appreciation for the incredible amount of work that went into it, thank you (Applause).
I'd also like to say that anyone who thinks a single person can't make a
difference probably hasn't met Marion Joseph, and I would like to publicly acknowledge
her efforts on behalf of the children of the State of California (Applause). (Standing
As I was saying last night, I've certainly been a lot more tired since I
met Marion (Laughter). Although Mr. Colvin said yesterday that the term is medieval,
I train teachers for a living. And I want to say that I'm gravely concerned about
how we are going to meet the shortage for qualified teachers and maintain high standards
all at the same time. I am coordinator of the special education credential program
at cal state Stanislaus, and I am under pressure to increase the number of credentials
we award. I've suggested that we set up a card table out in the quad and as people
write checks, we can hand them credentials. I don't know how to get more qualified
people out there. That's not my topic today, but I would like discussion about that
maybe later on.
Dave asked me to come and talk about the implications of Project Follow
Through for language arts, and he did not have to ask twice. I learned about Project
Follow Through early in my career and it has had an enormous impact on what I do
on a daily basis. While there are many implications and important things to talk
about, I'm going to talk about 3 or 4 thing today as time permits and allow time
for your questions.
The first thing that I would like to say, it's a very obvious point but
it sometimes gets lost, is that learning depends on language. And certainly learning
to read depends fundamentally on vocabulary knowledge. Wes Becker was probably the
first to assert that vocabulary deficiencies are the primary cause of academic failure
in disadvantaged students. I recommended a book by Hart and Risley called "meaningful
differences." These authors have undertaken over the course of their careers,
the incredible task of examining language practices in families. And this longitudinal
worked revealed significant differences in the experiences of young children. What
this figure shows (Overhead) is the estimated number of words spoken to children
from families who are on welfare, working class families and what they called professional
families. By the time the children are 4 years old, the difference in numbers of
words spoken to children between professional families and low S.E.S. households
is more than 40 million words. That gives you a glimpse of, perhaps, the magnitude
of the problem for any intervention program in an attempt to equalize the experiences
You saw this graph yesterday when Dr. Grossen spoke. I wanted to put it
up here again and talk a little about language. On this graph, language is purple,
it doesn't show up well, it is the last bar in each four -bar set. As Bonnie told
you, the 20th percentile is where we would expect children in this income bracket
to be performing without any intervention. Any time the bar moves up from the 20th
percentile would be considered a positive effect of participation in the follow through
model. When the bar falls below the 20th percentile, that in fact is a negative
outcome. In other words, that means that the children would be performing at the
end of third grade, at a lower level than we would have predicted they would have
performed without any intervention. In medicine this is called an iatrogenic effect,
which means that a bad situation was made worse by the intervention. The patient
was sick, he came to the hospital, we treated him and now he is dead. If you look
at this bar, you can see without exception there is only one model that made significant
improvement in the language skills of children -- the direct instruction model.
I think this is important. In California, because of our incredibly diverse population,
we tend to focus on students acquiring English as a second language. There's another
problem, and that is native English speaking children who do not speak English proficiently.
I've had the pleasure of looking at a lot of textbooks in the last few months
(Laughter). And I keep seeing something in concept development or vocabulary programs
referred to as a naturalistic approach. All of the programs in Project Follow Through
addressed language in some way, and I think what the data would suggest is that a
naturalistic approach is not an effective approach.
I want to mention briefly the programs used in the direct instruction follow
through model. I know that Mr. Honig had questions for Bonnie last night at dinner
and I want to make it clear, when we're talking about the direct instruction model,
we are not talking about generic, teacher delivered instruction. These are very specific
programs based on extremely sophisticated principles of instructional design. The
programs have been revised a number of times since follow through. In their current
incarnation, the programs, published by SRA McGraw Hill, include language for learning,
reading mastery, spelling mastery, and connecting math concepts.
I don't have time today to go into the design of the programs, but it is
important for you to understand that it is essential to have excellent delivery,
but you must be delivering a well-designed program, and those things cannot exist
independently of one another. I wanted to make sure you understand that.
So language is essential because, sometimes we forget and we try to teach
reading to students with language deficits, and we have to remember every day of
our lives that reading sits on top of language. Reading is parasitic with language
and you can't teach reading to kids who don't have language.
Finally I would say that the follow through data suggest that, for at least
some children, they will require explicit instruction and massive amounts of practice,
if they are to master the language of instruction that is essential through their
success in school.
I think Bonnie also showed this graph -- (Overhead) I know she did because
she loaned it to me this morning. I didn't bring mine with me. I think a second
major implication of follow through is that self-esteem is a by product of achievement.
If you look at the data, again the red bars are the self-esteem measures -- you
will see that the models that emphasized self-esteem in their curriculum -- down
at this end, didn't do so well on measures of self-esteem. The direct instruction
model not only had high measures of self-esteem, but also high measures of academic
This is really important to me. (Overhead) you may have noticed reading
my bio that my background is in behavior analysis. I'm interested in behavior and
in studying the relationship between behavior and the environment.
This is a model (Overhead) that's not one that I designed. I have to tell
you've used it for so long now, I don't even know who to give credit to any more
and I apologize for that. It is called the frustration self-esteem model of problem
behavior. The two most common reasons that kids are referred for special education
are they can't read, they have behavior problems. I need people to understand that
the two are inextricably linked. This model suggests that through deficient school
practices, kids have a series of unsuccessful outcomes, and over time you cannot
sustain an attribution that does not involve yourself. So that over time, students
develop an impaired sense of self and poor self-esteem and this contributes to the
development of escape and avoidance behaviors. So they learn to do things that get
them out of tasks, that delay the onset of instruction. And these things often are
at least initially reinforced by peers, who are sometimes glad that instruction is
delayed as well. And then this problem behavior feeds back and contributes to their
on going unsuccessful out comes in school.
What I think is important to understand is that it is real easy for us to
be focused on the problem behavior. We try to put contingencies in place to either
try to set up a system to reward the child for not engaging in problem behavior or
set up a punishment contingency to suppress the behavior. This model suggests that
in our efforts to suppress behaviors, focusing on this end (Overhead) we took our
eye off the problem. And the genesis of the problem is deficient school practices.
We need to provide practices that will be effective and cause students to be relentlessly
successful and prevent those problems from occurring.
You are kidding? Out of time? I do have to get this out. Sandra Horn mentioned
this earlier and I think it is a critical finding. IQ scores cannot predict achievement
gain. That's a very important finding of the follow through data. What you see
in this graph is work done by Gersten, Becker and colleagues, who analyzed data from
kids in direct instruction follow through groups. The figure shows gains on the
wide range achievement test over time by IQ level. You are looking at kids with
IQ's under 71, 71 to 90, 90 to 100 and so forth. Three conclusions can be made from
this analysis. First, the rate of gain was similar overall in all groups. Not surprisingly,
if you look at the beginning point, kids with IQ scores below 71 have achievement
scores that are lower than their peers, and also the ending point. But looking at
the slope of change, you can see the slopes are very consistent.
A second finding is that the students who have below average IQ had above
average achievement scores. The average -- the mean is represented by the dashed
line. So you can see even kids with IQ's under 71 are performing above the mean
on those measures.
Now I want to talk for a second about how that happens, because it's important
for you to understand that it doesn't happen by using heterogeneous grouping. It
didn't happen by teaching the same thing to all children at the same time for the
same amount of time. It happened by grouping children according to their needs,
and teaching them, and watching very carefully on a day-to-day basis what their progress
was and being able to make adjustments based on information about the students' performance.
If you teach everybody the same thing in the same way at the same time, you
are going to get a normal curve. You won't get something that looks like this (Overhead).
I want to share with you because my time is close to being up, share a story. I
think that if you remembered one thing I said, I would hope that it's this. Millie
Schrader is here and she is a friend of mine and the best principal in the united
states and I have I had the incredible honor of working with her in Elk Grove for
about six years, implementing direct instruction programs. And one of Millie's teachers,
Rhinda Miller, was teaching a lesson one day. And I observed it and said, how do
you think the kids did? Rhinda said, "pretty good for the green group".
Here's what I want you to know. There is no such thing as pretty good for the low
group, okay? In a mastery based program, when the low group finishes lesson 20,
their performance will look exactly like the performance of the high group when the
high group was on lesson 20. The high group may have been on lesson 20 nine months
ago, but it is really important for you to understand that standards are like end
of year goals. And I think it was Robert Mager who said, we know where we start and
then our goals tell us where we are going. And if at any time you don't know where
you are between those two, you are lost.
What we have to know is, are we, on a daily basis, making progress with every
child for whom we are responsible toward those standards? If not, we will get to
the end of the year and find out the kids didn't meet the standards, we will recommend
retention, and that becomes a punitive system. And I don't think that's what anybody
has in mind in standards-based education.
Because I want to leave time for your questions, let me summarize. Language
is essential and needs to be taught. Self-esteem is an outcome of achievement.
And learning is a function of instruction more than any other variable, including
Every child can learn. They are not responsible for discovering important
information themselves, such as the alphabetic principle, it is our responsibility
to teach them. And I want to end with what has become my favorite quote to give
you some idea of how long these problems have been going on. Quintilian, a Greek
philosopher and rhetorician, said "a child was born to learn. If a child turns
away from education, it is a sin upon the elders and not a flaw in the character
of the child". Thank you.