PERvERTING TESTS
Munson Elementary changing tactics on proficiency tests
By TONYA SHIPLEY / Zanesville
(OH) Times Recorder Staff Writer
Originally posted March 8, 2003
This
is an excellent overview of what is happening in education–and what’s wrong
with it.
ZANESVILLE — There is a positive belief in the halls of Munson Elementary
School that this year will be different.
The students, teachers and the administration believe this year their school
will do better on the proficiency tests.
First
problem: The goal is to improve learning, skills and knowledge. Those
improvements should show up in properly written proficiency tests as higher
scores. But educators are short circuiting the process, as demonstrated below,
by going for improvements in scores.
Example: The goal of an unborn chick is to get out of the egg. The chick pecks
away at the egg, getting stronger in the process. If someone decides to help the
chick by breaking the shell open so the chick can get out sooner without doing
the work needed to build strength, the chick dies soon after birth. It’s not
just accomplishing the objective that matters, it’s also the process and the
journey–like taking the scenic route instead of an interstate highway. You have
to have theories of
learning, learning
styles and motivational
styles, with a strong curriculum incorporating core
learning standands, and follow that route. (See, also, Funderstanding).
TODAY’S
BEST OF MYSHORTPENCIL.COM
• SEE
A LIST OF THIS WEEK’S COMMENTARIES
• More
Stories on Pedagogy
Testing began Wednesday and will continue until the final test is given on
Tuesday, March 18. This year the whole school has become more targeted on the
tests since the school, along with McKinley Elementary School, was among 161
Ohio schools which failed to meet federal achievement requirements based on the
No Child Left Behind Act.
Second
problem: Test preparation. If there is one thing students are experts at, it’s
taking tests. They take hundreds of them. Some school districts spend
a month or more preparing for state tests. It’s inconsistent with sound
teaching practices. Tests
measure educational progress, they don’t create it.
"It’s been a real intensive effort to get ready for the test,"
Principal Bob Grayson said.
The process by which the school prepares students for testing has changed in an
effort to improve scores.
Some of the changes are:
• Developing a list of students ranking them into groups of who passed the
tests, those who almost passed the test, and those who did very poorly.
Grayson said a students who were in the group of almost passing, were highly
targeted to get them over the margin into passing, while intervention tactics
were used with the group which tested poorly.
Third
problem: Teachers are using standards not to raise all boats, but to concentrate
on marginal students. Students who "naturally" score in the passing
range are essentially ignored. Minimal, if any effort, is expended in moving
them into the "exceeds standards" range. There’s no payoff for
teachers or school districts. This is not the widely derided practice of teaching
to the middle, this is teaching to the bottom.
Scotia-Glenville is solidly in this mode. At the December 16 S-G Board meeting, BOCES
Data Analyst Kathleen Maxwell told the staff to focus on the "almost
passing" students.
"We don’t care if [students] get a high pass. If they’re just at the
cutoff, that’s all we care about," Kathleen said.
From the school’s perspective and goal of creating politically acceptable
results on state tests, even if it is on a bogus
performance index, Kathleen is right. But what about academic excellence?
Shouldn’t schools be just as concerned with moving students from passing to
exceptional performance?
They aren’t. If students pass state tests, that’s the standard, regardless of
their potential to do better, and regardless of the consequences for later
learning, notably, the lowering of effort, doing enough to get by, and the
failure to adopt and exercise a focused academic discipline to excel. It’s all
being lost because the students can sense, if they don’t know it already, there
is almost no reward for doing the work to get into the "high pass"
category of exceptional performance. It is not unlikely that this approach to
learning and concentration on the marginal students makes it more difficult for
students to pass state exams.
• The school does off-year testing in grades first, second, third and fifth to
assist them in listing the students.
Scotia-Glenville
does this, too, with Terra
Nova testing. The idea is to detect weaknesses in student learning and fix
them before state exams. The Terra Nova closely tracks NY’s state exams.
This is the Fourth Problem: No single test can tell you what the problems or
weaknesses are with a particular student’s learning. From listening to teachers,
they tend to over-specify the learning problems based on the most recent test
result. It doesn’t work. Tests like the Terra Nova, although labeled as student
assessments, are really teacher and curriculum assessments. The most valid use
of these tests is the cumulative average of the students in a class or school to
discover gaps in the curriculum or teacher effectiveness. Former S-G Director of
Curriculum had it exactly backwards when she said, "Terra Novas should
never be used as a measure of Scotia-Glenville–they are diagnostic tests for
students."
Even BOCES analyst Kathleen Maxwell implicitly acknowledged the problems of
using test results to evaluate student weaknesses when she said that state exams
are like taking a student’s temperature on one day. You can’t take a patient’s
temperature on one day and know what treatment to pursue for the rest of the
year.
Why not? Part of the reason is the Terra Nova and state exams aren’t that
precise because they often ask only one question pertaining to a skill or
specific learning standard. You can’t know from one or two questions whether a
student needs to spend a little, a lot, or no additional time in mastering the
standard. The student may have missed the question because of vocabulary, a
reading problem, carelessness, confusion with an earlier question, a temporary
distraction, fatigue, or any number of other problems that would make it foolish
to take off on a course of remediation based on the content or standard the
question purportedly measured.
Despite Kathleen’s statement that you can’t know what you should do to help a
student based on a single test result, the S-G school board repeatedly
complained that teachers do not provide parents with enough specific information
about how their children are performing on Terra Novas so they can know how to
help them. Board member Margaret Smith, a teacher by training, seemed least
capable of all of understanding what Kathleen told them. She repeatedly asked
questions that demonstrated she believed that if she knew what questions her
children missed she would know how to help them. It’s not that easy.
In general, when using the Terra Nova test, the most a parent or educator can
hope to learn from a series of three or more tests is something about strengths
and weaknesses in broad subject or skill areas, and the student’s rate of
learning, both compared to the student’s past rate of learning and compared to
students nationally. Beyond that, the attempt to extract more specific
information is just as likely to detract a student from more profitable learning
experiences as it is to fix a problem.
It’s true, I
relied on IOWA scores to detect and fix problems with my daughter’s math
performance. But, I used a series of data over 3 to 6 years, not to diagnose
specific math problems, but to evaluate the rate of her general progress in math
relative to her demonstrated ability to achieve, both in math and English, and
relative to the achievement of other students, nationally. I did not teach to
the test. I used comprehensive math texts from Saxon
Publishers, and I taught all the material, cover to cover–about 40
hours per text, including "homework" and tests. (Public schools take
at least 120 hours in class plus homework to cover between 70 and 85% of a
year’s math). In other words, I
taught to the standards that are rich and challenging, and I let the tests take
care of themselves.
Moreover, I did not analyze the test or specific problems. It was unnecessary. I
could tell from her performance on practice problems and unit tests where she
needed help. Which brings up an important point–the results on standardized
practice and state exams don’t reveal anything about the student’s weaknesses
that the classroom teacher shouldn’t already know from grading the student’s
homework and tests, provided the teacher has integrated the content and skills
of state standards into the curriculum.
• The school purchased several published programs which are aligned with the
proficiency test outcomes and state standards.
Ditto,
S-G. I call it teaching by the numbers, like painting by the numbers. At the
end, you get a picture, but students are no better educated by the process than
they are better artists. Take away the numbers and the outline and students are
lost, in both education and art.
• Downloaded several of the previous proficiency tests and used those as
practice tests.
Drill
and kill. This is not education. The time would be more productively spent
working through a rich and challenging curriculum.
• The Ohio Department of Education provided practice books.
NY
recently released access to multiple choice questions from prior exams.
Doubtless, scores will increase, especially if the content and format of
questions does not substantially change from year to year, but I doubt learning
will.
• Extended day program and summer school.
One of the problems students had with the tests were the extended answer
questions. These types of questions aren’t just asking for an answer, but for
the process the students went through to get the answer.
These questions offer multiple points depending on how much of the question a
student completed correctly, but Grayson said students didn’t know how to go
through the logic steps to answer them. Now the teachers are working on working
the steps of problems with the students.
This
kind of work should be integrated with the curriculum. What I’ve seen happening
at S-G is that teachers pretty much teach the way they always have, making the
same kinds of assignments and tests, and then spending a couple of weeks on the
special skills needed for state exams. This is wrong, wrong, wrong. The DBQs,
structured answers, and other skills and content need to be sprinkled in
throughout the course to provide repeated practice over a long period. Cramming
for the exam is not only bad and ineffective practice, it also teaches students
they can succeed by adopting this bad habit.
The way the school thinks about the tests is changing, too.
"We feel that to improve scores in our building it’s a K-6
responsibility," Grayson said.
That
much is true. It’s consistent with learning theories. Avoiding problems
achieving excellence in the 6th grade requires the execution of a plan that
starts in kindergarten.
By doing the off-year testing all students are taking tests at the same time.
Mark Burrier, a sixth grade teacher, noticed a change in attitude among the
students and the school. He said the students are given two and half hours to do
each test and most of the students use all that time, often checking their work
to make sure they are doing it right.
"These kids are working their hearts out. They want to do their best,"
Burrier said.
So far the students feel good about their tests.
"It’s kind of easy. It’s really simple if you think about the questions,"
said Kymm Chandler, 12, a sixth grader.
Out
of the mouths of babes. . . The truth is the learning standards aren’t that
tough. They
can’t be and have everyone pass them. For example, S-G’s special education
students are scoring as "almost passing"–high level 2–even though,
by definition, they are at least 1.5 years behind general education students.
State exams and learning standards are only capable of preventing schools from
graduating students with the deplorable reading, writing and math skills of the
’80s and ’90s. They are not capable of producing academic excellence.
None-the-less, they are a necessary evil in today’s professional-controlled
public schools. At least all students graduate knowing something, being capable
of performing at a moderate degree of competence. But academic excellence is now
governed more by game theory, and the pursuit of passing scores, than by sound
education practices. In my opinion, that’s because a majority of today’s
teachers weren’t all that good at learning when they were in school and college.
Education is still a long way away from being able to teach The
21st Century Student.
* * * *