Blue Ribbons and Special
To the Editor:
Unfortunately, the U.S. Department of Education's new
Blue Ribbon Schools identification procedures are not
likely to be any more effective at identifying schools
where achievement is improving than were the old
Revamps Blue Ribbons, Basing Awards on Testing," Aug.
7, 2002). The basic problem is that neither set of
identification procedures pays any attention to the rates
of identification of pupils with special educational needs
or rates of retention in grade.
Virtually all states exclude the achievement of special
education students in state ratings of school performance.
But there is solid scientific evidence that schools with
similar student demographics often exhibit very different
special education participation rates. Some schools, that
is, identify twice as many kids as having special
educational needs as do schools with similar student
Identifying low-achieving students as pupils with
special educational needs removes those students from the
accountability rolls and from the new Blue Ribbon School
achievement estimation. The federally funded research my
colleague Anne McGill-Franzen and I conducted, as well as
a similar study by David Potter and individuals of the
South Carolina Department of Education, demonstrated that
as high-stakes accountability pressures have increased, so
too have the rates of identification of students as
needing special education services.
We documented the use of special education placements
as a strategy for creating the impression that achievement
was rising. This strategy was successful, in that the
principal who told us of how such a strategy worked had
his school named a Blue Ribbon School the year after his
interview with us. By tripling the number of low achievers
identified as pupils with special educational needs, he
manufactured what looked like improving achievement.
That practice is increasingly common today, and the
continually increasing special education enrollments
buttress that claim—a phenomenon that Jay P. Greene
neglected to include in his recent Commentary on rising
special education enrollments ("The
Myth of the Special Education Burden," June 12, 2002).
Flunking low achievers has a similar effect on reported
achievement levels. That is, flunking the lowest-achieving
quintile of students at the end of grade 3 means that 4th
grade achievement will seem to be rising, at least in the
year following the flunking. With an extra year of
schooling, and with the expenditure of the extra thousands
of dollars that extra year costs, those flunked 3rd grade
students should score better when they finally take the
4th grade test a year late. But this does not indicate
that the school is a more effective school, just a more
Blue ribbons should be awarded to those schools that
reliably serve poor children well. A focus on improving
test scores is one strategy for identifying those schools,
just as examining a company's financial record is a
reasonable strategy for estimating improvements in its
earnings. But Enronitis accounting procedures have
undermined confidence in corporate statements. Educational
Enronitis, of the sort encouraged by the Blue Ribbon
School identification procedures, has long been documented
and long ignored by policymakers.
The failure of the Department of Education to develop
rules and procedures that incorporate the scientific
evidence on estimating educational quality allows
Enron-like cooking of the achievement numbers and
undermines the credibility of the Blue Ribbon awards.
Irving and Rose Fien Distinguished Professor of
Elementary and Special Education
University of Florida
To the Editor:
I read with interest your article on the U.S.
Department of Education's Blue Ribbon Schools program. All
programs need to be reviewed over time or they become
The focus on results is clearly what the public and
parents are seeking, so the Blue Ribbon program's changes
are on target. Often, many reports from schools are rich
in processes or "things done," but poor in reported
positive results. Yet I regret that the award has dropped
processes from its focus. It may be difficult to determine
if the results are connected to the design process or to
some other factor, like a particular teacher or principal.
If your readers are interested in looking at both
processes and results, I recommend the Malcolm Baldrige
National Quality Award Program. This is an award that
rigorously connects what a district says it does in
leadership, planning, data collection and management,
stakeholder needs, curriculum design, and instructional
delivery processes to results. The award gives the
district affordable feedback from experts on how to
improve the district's processes to gain the results
desired. It is based on a model of continuous improvement,
or the belief that "good enough is not enough."
As a member of one of the first school districts to
gain this award, I can attest that involvement with the
Baldrige standards over time will dramatically improve
student achievement. We found that by applying the
Baldrige standards, all our processes improved. We were
able to link this improvement to the improvement in
I urge readers who wish to become involved in working
toward an award and improved results to learn more at the
Baldrige Web site,
Pearl River Public Schools
Pearl River, N.Y.
Educator to Public: 'Put
Up or Shut Up'
To the Editor:
One of your recent installments of Close Up: The Public
Education Network-Education Week Poll, focused on
the poll's findings concerning teaching quality ("Teaching
Quality Viewed as Crucial," July 10, 2002), shows just
how far out of touch with reality the American public
The problem is not the quality of teaching, it
is the quality of learning. These are American
public schools we are talking about. They are structurally
inadequate and highly underfunded, if grade-level literacy
and numeracy for all students is the goal.
I've worked in the public schools for decades. Many of
these students can't do the work. Many others won't do the
work. Most have other priorities: their friends, clubs,
activities, sports, and occasionally drugs and sex.
The public schools reflect society as a whole, and we
are in real trouble. The IQ of public school students is
average or below average. Many schools want to force
difficult subjects like algebra down into middle school,
when most of these kids can't even count. Hello! Can't be
Many of the parents of these students don't understand,
don't care, or both. Let me give you a typical example: I
was in a conference with a parent about the failing grades
of her child when she told me we needed to end the
conference so she could get to a baseball game. Typical.
There is a saying in the West: If you plant yellow corn,
you get yellow corn.
You want to "fix" schools? Try creating an institution
where learning is the priority, not babysitting and
recreation. If you want to "fix" schools, spend the money
necessary to achieve full literacy and numeracy for all
It's time for the American public to wake up and smell
the coffee. Wishing doesn't cut it. Fantasies won't cut
it. No administrative fad will do it: No "high
expectations," no "block scheduling," no innovation will
improve schools when it is the structure and funding of
schools that are failing. It's time to get a clue and put
up or shut up.
Making a Virtue of
Teaching to the Test
To the Editor:
Since the grade of A in a subject does not indicate the
same learning in Atlanta as in New York City, or anywhere
else for that matter, the SAT is a necessary student
measurement for colleges in their admissions process. It
gives them one indicator through which to compare the
applicants on a level playing field ("Overhauled
SAT Could Shake Up School Curricula," July 10, 2002;
"On Changing the SAT," June 5, 2002).
I applaud the College Board's plan to revise and, as
the board hopes, improve the SAT as a measurement of
curricular information, thereby making it a better
indicator of a student's success in college.
If we are going to assess students by testing, there
will be, without a doubt, teaching to the test. Instead of
complaining about this and the time it takes away from the
curriculum, we should let the test drive meaningful
We could take the soon-to-be-added essay portion of the
SAT, for example, and announce its general topic a year in
advance. Since recent headlines bemoan how little our
students know about history or science, we could make the
schools prepare students to write essays in one of these
areas. Have them research and write about cloning,
alternative fuel sources, World War II, the Civil War, and
other such topics. In this way, students would be studying
relevant subject matter, learning to write, and preparing
for college and the SAT all at the same time.
I know this may sound too simple but, in teaching an
Advanced Placement course, I taught to the test because it
was the material the students needed to know, as suggested
by the colleges and universities. Why would it not work
Gordon R. Rode
Head of School
'High-Threat Tests' Upset
To the Editor:
I've been thinking about the Princeton Review's
conclusion that North Carolina has the best testing
program in the nation, as reported in your recent Testing
the Testers," July 10, 2002). Perhaps we do, but I
think we misuse it. I call it "high-threat testing."
Threatening all of our students, teachers, and
principals every year with dire consequences based on test
results is a horrendous practice. Children are threatened
with retention. Teachers are threatened with less pay, and
everyone is threatened with publication of the results. I
would like to see North Carolina give the tests, but
instead of threatening us, let us use the scores as
feedback for future instruction.
Decisions on retention should be made, using multiple
assessment tools, by the people who know each student. And
a school should not be labeled a "school of distinction"
or "exemplary" or anything else based on standardized-test
results. My school, which is a Title I school (meaning
many students receive free lunches) made significant gains
last year but, according to North Carolina's mysterious
formula, missed earning bonuses for its teachers by
one-tenth of a point.
Giving a financial bonus to the teachers of students
who do well and not to other teachers is demoralizing. It
is common knowledge that the scores on standardized tests
are more reflections of socioeconomic status than anything
else. And all teachers know that teachers work hardest
with the children who score in the lower ranges. In my
room, to a child, the students who scored at the lower
levels of 1 and 2 were English-as-a-second-language
students or students who were not being raised by their
natural parents or were being raised by only one parent.
Schools that have public-housing developments in their
districts have much catching up to do for their students
and work the hardest, even though their scores almost
always average out lower. For those teachers to watch
teachers of easier populations earn an extra $1,800 is
demoralizing. Doing this to teachers year after year is
causing more teachers to retire as soon as they can.
Younger teachers leave the profession entirely after a few
years. What a loss for the children and for the future of
Why Not 'Flag'
SAT, ACT Exams?
To the Editor:
I was saddened and puzzled by your Aug. 7, 2002 article
"Test Companies Lower 'Flags' on College-Entrance Exams."
One company, the College Board (owners of the SAT),
blinked in the face of a lawsuit; the other, ACT Inc.,
quickly followed suit. Henceforth, they will not notify
test-score recipients (such as colleges and universities)
when the scores were achieved with extra time—a
nonstandard condition. The College Board's decision was
apparently made on the basis of a 4-2 vote by a panel of
experts the parties had convened.
Thus will innocent third-party score recipients receive
flawed reports. Although these reported actions by private
entities do not have the force of law, I worry that they
open a Pandora's box and will have a chilling effect on
our nation's various (and nascent) efforts for meaningful
reform, tests with integrity, and accountability.
As of the fall of 2003, no one will ever know that some
ACT and SAT tests were taken under nonstandard conditions.
You report that the College Board will "stop flagging for
accommodations such as extra time."
"Such as" invites more. Extra time is but the tip of
the accommodation iceberg. What about scribes,
calculators, readers, extra explanations, and so on? And
what about other worthy groups: English-language learners,
"at risk" students, students we all know who just don't
test well (though not labeled as disabled)? Where is the
fairness and equal protection for them? Who will be the
next defendant? It was a sad day for valid and meaningful
Moreover, these are norm- referenced tests, which
obtain their validity and reliability by having all
students take them under standard (similar) conditions.
It's puzzling why these companies squandered the very
credibility of their products, especially since it is
not illegal to report that a test is given in a
nonstandard way. In fact, some laws—and certainly testing
protocol—require it. However, henceforth, no one will know
how these tests were taken. And, apparently, time doesn't
really matter, though these testing companies intend to
keep timing most students.
What now is the norm and what, if anything, do such
results mean? Whither test integrity? The companies
appeared to defend their action as promoting fairness for
all students. I'm all for fairness, but they overshot the
mark. The decision to stop "flagging" is neither fair nor
sensible. Who benefits from test results whose very
essence (validity and reliability) is jeopardized?
A better solution: Either (a) don't implement this
policy and go to court to defend the test(s), or (b) give
all students (not just students with disabilities) the
option to take these tests with accommodations that will
be flagged, or, if time doesn't matter after all, (c)
simply let everyone take the tests untimed. Then, fairness
and validity would be re-established.
At the very time that accountants, from Enron to
WorldCom, are under scrutiny because their numbers may not
tell the true story, the College Board and ACT Inc. choose
to head in the same ill-conceived direction—their numbers
will not tell the true story. Sad and puzzling. I just
don't get it.
Miriam Kurtzig Freedman
Advanced High School:
Should College-Level Programs Be Offered to All?
To the Editor:
The most telling anecdote in Jay Mathews' essay
"Advanced Placement" (Commentary, Aug. 7, 2002) is
that of the high school in East Los Angeles where all
students are admitted to AP courses because "even [those]
who flunk an AP course are better prepared for college
than those kept out of the courses altogether."
Mr. Mathews is dead on, both about the value of AP and
International Baccalaureate courses and about the need to
make those courses available to more students.
The best way to democratize AP and IB, to ensure that
every student has a crack at excellence, is to transform
the whole school. The Bard High School Early College in
New York City, created a year ago, offers one model. When
I visited the Bard school, I saw 11th graders discussing
Plato in a way that would make me proud of college
seniors. I saw a cadre of city kids who will become
leaders, kids who otherwise might have been lost to us.
And I also saw the liberal arts where they ought to be,
where—as Earl Shorris has so eloquently pointed out—the
social action is most consequential.
Bard's experiment, along with several other intriguing
efforts, spurred the Bill and Melinda Gates Foundation to
announce its Early College High School Initiative this
past spring. In this initiative—also supported by the
Carnegie Corporation of New York, the Ford Foundation, and
the W.T. Kellogg Foundation—each of seven partner
organizations will create academically rigorous small
schools where disadvantaged students get the support and
challenge they need to excel. For example, we at the
Woodrow Wilson National Fellowship Foundation are working
with four-year colleges and research universities to
develop nine new early-college high schools for urban
Whatever the specific model, all of us in the Early
College High School Initiative are working to give a
broader range of students, as Mr. Mathews says, "a hard
taste of college academic life," and indeed to stimulate a
lifelong appetite. It's high time that we make excellence,
opportunity, and challenge for high school students not
just an AP-style taste, but a schoolwide diet.
Woodrow Wilson National Fellowship Foundation
To the Editor:
Jay Mathews' Commentary on the rising numbers of
students taking the Advanced Placement exams is right on
target. Indeed, if prep schools like the Fieldston School
are dropping AP courses (although students will still take
AP exams), this is not a sign that AP exams are of little
value. Rather, it is a sign that if standards are high and
courses intellectually challenging, we don't have to teach
to a tough test.
What Mr. Mathews did not say is that the growth in the
number of AP exam-takers is not an isolated phenomenon. As
a result of the K-16 movement, reports about the waste of
the senior year, and state performance assessments for
high school exit that are given as early as 10th grade,
students—and not just privileged and gifted students—are
taking advantage of a number of ways to get college credit
in high school.
Across the country, dual-enrollment programs are
growing. In these, high school students take courses on
college campuses, or teachers certified by postsecondary
institutions give regular college courses in high schools.
A number of community colleges have successful programs
for high school students who not only flourish in the more
adult, less regimented atmosphere of college, but also
complete both the high school diploma and the Associate of
Arts degree on an accelerated schedule.
And there is an even more radical way for students to
earn the A.A. or two years of college credit in high
school, especially those for whom there are financial and
other barriers to college-going. With Jobs for the Future
as coordinator, the Bill and Melinda Gates Foundation, in
partnership with the Carnegie Corporation of New York, the
Ford Foundation, and the W.K. Kellogg Foundation, has
provided funding for an effort to establish 70
early-college high schools over the next five years.
Partnering with both two-year and four-year postsecondary
institutions, early-college high schools combine high
school and college in a single school, and allow students
to take college-level courses as soon as they are able.
Early-college high schools share the attributes of
effective small schools, but they also unify academic work
from grade 9 through grade 14. Designed to increase the
number of first-generation, low-income, English-language
learners, and students of color having the preparation to
attain a bachelor's degree and highly skilled work,
early-college high schools demand more of students
earlier, while eliminating obstacles to college entrance.
Our young people deserve such options, and the good
news is that they are taking up these various
opportunities for challenging intellectual work.
Jobs for the Future
Early College High School Initiative
To the Editor:
Jay Mathews presents a clear and compelling argument in
his Commentary. Unfortunately, he fails to address the
most important issue facing the use of the Advanced
Placement program in today's college-admission scene. The
AP program was originally designed as a means of
acceleration and enrichment. Most students took AP classes
in their senior years of high school, and because the
program was meant to provide an opportunity for
advancement on the college level, the scores were not
available until July of the summer before the freshman
year of college. Now that colleges are using AP scores in
the admission process, students are taking AP classes
earlier and earlier as a means of enhancing their profiles
for admission, not simply as an opportunity for
As a past AP grader, I often encountered essays written
far below the level one would expect of an AP- quality
student. I became aware that many high schools allow any
student who is interested in the AP program to enroll,
even if the student is far below average in ability. With
this being the case, judging a high school's quality of
education or curriculum by the number of AP courses
offered is ludicrous. If it is important to provide this
sort of ranking to the public, only a ranking based on the
outcomes of those students' exams would truly indicate the
quality of the education they are receiving at their
particular high schools.
A case in point: My nephew took AP calculus his senior
year. This, of course, appeared on his transcript. He
earned an A in the course. In reality, his class covered
four chapters over the course of the entire school year.
No one in his class took the AP exam. My nephew was
admitted to engineering programs at several colleges. I
would assume that those colleges believed this young man
had successfully completed AP calculus.
I wish I had a solution to this problem, but I don't. I
only wish that there were some way to return the AP
program to its original intent. Possibly dropping the AP
designation from transcripts would help. It would then be
necessary to see the results of the tests to determine a
student's true level of ability. Until the purpose of the
AP program matches its application and use, any
conclusions drawn will be in error.
Anne Macleod Weeks
Director of College Guidance
AP English Literature Teacher
Online AP Consultant
To the Editor:
Bravo to The Washington Post education writer
Jay Mathews for his fine essay on the Advanced Placement
program. I too have been concerned by what he calls "a
season for bashing ... the college-level courses and tests
... in American high schools." I too am a proponent of the
As the "No Child Left Behind" Act of 2001 sets the
direction for schools nationwide, one must ask whether its
goals are even compatible with a successful AP program. I
worry they are not.
The AP program has two distinct goals: to help students
be admitted to good colleges prepared to succeed there;
and to give students college credit for courses taken in
high school. Each is a worthy goal unto itself.
The mandate for America's high schools is to enable all
students to earn a diploma. That requires completion of
prescribed courses and success on standardized tests. It
does not require proper college preparation.
The federal and state governments now impose heavy
burdens on schools to meet the "everyone earns a diploma"
standard. Districts are evaluated based on factors that
focus on middle- and lower-performing students—that they
show progress on tests of minimal competency, that they
attend school regularly, that they graduate. Districts
earn no extra credit for students who achieve
excellence, as judged by taking honors and AP courses or
being admitted to colleges and succeeding there.
School districts face diminishing resources, like good
teachers and administrators as well as funding for
programs and supplies. One strong criticism of the AP
program is that it diverts scarce resources from needier
students. Rigorous high school coursework facilitates
success in college, but not all students go on to college.
So why not direct all efforts and resources towards what
everyone can achieve?
Mr. Mathews argues that more students should be
encouraged to take AP courses. But those courses, like
actual college courses, often have prerequisites. Unless
students master algebra in grade 8, they may not be ready
for AP math and science courses in high school. Unless
they take honors-level English or social studies, they
will not be prepared for AP courses in those disciplines.
Rigorous coursework is a cumulative process.
Preparation begins in middle school. Yet, in most
districts, there is little if any communication between
middle school and high school teachers or counselors. You
cannot simply drop kids into AP courses for which they are
unprepared and expect success. That can only lead to
frustration for all.
Along with the issues of scarce resources and solid
preparation, there is the issue of grades. Local schools
devise their own grading systems. Grade point average and
class rank are important to college-bound students. In my
district, an A in regular English equals an A in honors
English or AP English. An A in basic math equals an A in
AP calculus. Students need an overall 3.5 GPA to qualify
for National Honor Society, and sometimes the students
taking the honors and AP classes don't make it.
Kids figure that out and "play" the grading system.
Taking all regular courses and lots of nonacademics can
lead to a 4.0 GPA and high class rank. Taking AP courses
and earning some B grades can be costly to both a
student's GPA and class rank. Some of our brightest kids,
with no understanding of real college preparation, opt for
the easy A grades. They ask me: "Why should I work so hard
in AP courses and maybe earn a B and drop in class rank?"
I believe there should be some standardized
grade-weighting system so that B's earned in AP courses
equal A's earned in regular courses. Many schools do this;
others believe it is unfair or unacceptable to colleges.
Finally, there is the "triple E" issue: excellence,
equality, and elitism. Is it elitist to encourage the
pursuit of academic excellence by some students? Or must
all students receive equal coursework? In an era of
shrinking resources and growing demands for
accountability, how do we balance equality and elitism in
our public schools? That is the great unspoken issue.
Betty Raskoff Kazmin
Board of Education Member
MCAS Results and Missing
To the Editor:
In their letters to Education Week ("Test
Questions," Aug. 7, 2002), David P. Driscoll and
Jeffrey M. Nellhaus of the Massachusetts Department of
Education take exception to some of the views expressed in
"Ensuring Failure," (July 10, 2002). Both claim that
the Massachusetts Comprehensive Assessment System, or mcas,
tests are criterion- referenced, but neither explains the
basis on which they make this claim.
At the same time, Mr. Nellhaus admits that "questions
for MCAS tests are selected and rejected for their
usefulness in differentiating performance." This means, I
submit, that items answered correctly by most students
during pilot testing will be discarded from the
operational version of the MCAS. This was one
of my essential points. Selecting questions in terms of
item discrimination is characteristic of norm-referenced
testing, not criterion-referenced testing.
Also noteworthy is that about the only validity
evidence presented in the 1998 Technical Report on the
MCAS is that MCAS scores correlate strongly with scores on
the Stanford Achievement Test-9th Edition and the
Metropolitan Achievement Test—both of which, of course,
are norm-referenced tests.
Mr. Driscoll asserts the MCAS tests are "high quality."
On this issue, I would simply ask Mr. Driscoll if he knows
of any other state test on which two items were shown to
be so obviously defective that the state withdrew them
from scoring—as happened with the MCAS in 2001.
Mr. Driscoll also touts the fact that "[o]n the 2001
MCAS test for grade 10, virtually every district in the
state showed significant improvement over previous years'
results in both English and math." What this ignores is
one of the unfortunate causes for apparent improvement in
grade 10 scores: Huge numbers of students are being failed
to repeat grade 9, or are dropping out before the MCAS
test in the spring of grade 10. To illustrate this fact,
the following table shows the numbers of students in
Massachusetts enrolled in the fall in grades 9 and 10 for
the years 1995 to 2001:
Cohort difference -4441
Percent missing -6.5%
Cohort difference -5018
Percent missing -7.1%
Cohort difference -5800
Percent missing -8.0%
Cohort difference -6091
Percent missing -8.2%
Cohort difference -8286
Percent missing -10.7%
Cohort difference -9878
Percent missing -12.4%
The enrollment numbers come straight from the
Massachusetts Department of Education. All I have added
are the differences between grade 10 enrollments in one
year and grade 9 enrollments the previous year—or, in
other words, the numbers and percentages of grade 9
students who are "missing" from grade 10 the following
As is apparent, in the years before MCAS, which was
introduced in 1997, only around 6 percent or 7 percent of
students turned up missing between grades 9 and 10.
However, by 2001, the rate at which students were missing
from grade 10 had nearly doubled, to 12.4 percent.
For white students, the "missing" rate increased from a
little over 5 percent in the pre- MCAS years to 9 percent
to 10 percent in 2000 and 2001. But for African- American
and Hispanic students, the missing rates skyrocketed even
more. For African-American students, the rate more than
tripled, from a little over 7 percent to almost 24 percent
in just five years. And the rate at which Hispanics were
missing from grade 10 nearly doubled, from 17 percent to
I will not discuss here the extent to which increases
in missing grade 10 students is caused by higher dropout
rates, inflated rates of failing students to repeat grade
9, or some combination of the two. But I do note that
these figures, based on fall enrollments, almost surely
underestimate the extent to which students are missing by
spring of grade 10, when the MCAS tests are administered.
For Mr. Driscoll to tout with pride increases in grade
10 pass rates, while ignoring the fact that lower
proportions of students— especially minority students—are
even getting to grade 10 is, to be charitable, more than a
mite myopic. Evidence indicates that the vast majority of
students who are failed to repeat grade 9 will not persist
in school to high school graduation (see, for example,
Part 7 of "The Myth of the Texas Miracle in Education" at
Lynch School of Education
Chestnut Hill, Mass.
Wrong on N.Y. Tests
To the Editor:
Walt Haney's Commentary on state achievement tests
cites New York state's tests among those that "have been
constructed using norm- referenced test-construction
procedures" and "are designed so that all students
can never succeed." He is wrong about the New York tests.
New York tests are criterion-referenced and are
designed so all students, given reasonable instruction,
can and should pass them. We do not prescreen or exclude
test questions on the basis of difficulty or P values, but
rather evaluate item-difficulty statistics only as flags
for potential flaws in the questions. In addition, our
process for setting cut scores and our conversion tables
from raw to scale scores are not norm-referenced. In fact,
New York state tests are calibrated using an
item-response-theory model in which questions are
evaluated for fairness, fit to the state learning
standards and ability to measure the content taught to
students in New York schools.
All of the information on New York tests is available
in over 50 technical reports that we have made accessible
to the public. It is critical that any claims about New
York state's tests be based on the facts.
James A. Kadamus
New York State Education Department
Deputy Commissioner, Elementary,
Middle, Secondary, and Continuing Education