|
II. We know:
Our state tests are demanding and anchored to the
National Assessment of Educational Progress (NAEP).
STATE ASSESSMENT DATA -
view and/or download chart in PDF (20 KB) RI faces its
performance deficits squarely. Until the reauthorization of the ESEA (now NCLB),
individual states’ testing programs had no mandatory
relationship to the national tests, or NAEP, given in
reading, writing and mathematics. NCLB requires all
states to participate in the NAEP, so while states will
still design and run their own assessment programs, the
federal government has a barometer by which to gauge the
effectiveness – and honesty – of each state’s efforts.
The NAEP tests do not reach every child in a state, just
a representative sample of 2,500 students in the tested
grade. RI has been participating, so this new
requirement will present no change, nor unpleasant
surprises. Indeed, from the beginning RI anchored its own testing
program in the same challenging content reflected in the
federal tests. This means we are not kidding ourselves
about how we will fare when compared with other states.
Some states chose either to give easier tests or to set
the cut scores so low that their students appear to be
more proficient than they are. Those states announce
proudly that 80% of their students meet the state’s
standards for reading or math, when the highest NAEPs
proficiency percentages for any state are in the low
40s. By the same token, RI accounts for every student
eligible to take a given test. Eligible students who,
for whatever reason, do not complete an assessment,
receive a ‘no score,’ which is counted as part of the
school’s overall proficiency rate. Some states have run
into problems with schools exempting large numbers of
students, presumably academically weak students, from
taking the tests. As a result, in those states no one
knows precisely what a school’s overall proficiency rate
is. Again, RI’s ‘all kids’ testing gives an honest
baseline. As schools re-organize to become more
personalized, absenteeism will drop and the ‘no-score’
rate will drop as well.
We recommend: |
‘Personalization’ – getting to know the students – will
help reduce ‘no scores.’ Research has demonstrated that every student should have
at least one adult in the building who knows him or her
well. Teacher teams at the middle school have been shown
to be very effective personalization tools, besides
aiding cross-disciplinary curricula and mutual support
among teachers. At the high schools, advisories – in
which each professional in the building takes a group
of, say, 12 - 15 students, for the course of their stay
in that building – are one reasonably simple way to make
sure that each child is known. High schools are
especially anonymous and their scheduling is notoriously
difficult because high school students have varying
interests and abilities, which is why advisories are
easiest to implement. Learning environments of no more
than, say, 400, are probably small enough that children
are known. But most secondary schools in RI are
considerably larger and need an internal strategy to
look after its students. Knowing the students well can greatly improve student
achievement in general, as well as improving the
students’ cooperation with specific demands like state
tests. Furthermore, since personalization closes the
cracks through which students slip, ‘no scores’ will
drop as each child has an adult who makes sure the
eligible student takes the tests, makes up the tests or
has an excellent, exemptible reason for not being in
school during testing. Also see this document’s Section IV for more suggestions
for how to help children become less distracted from the
education process. Today’s society has a great deal to
do with the difficulties schools experience, but schools
cannot afford to let themselves believe that the
children’s needs are someone else’s job. |
While elementary reading is a bright spot, over-all
performance over the past three years is not heartening.
The chart showing achievement
over the past three years, by characteristics is a quick and simple
representation of where RI stands in relation to its
overall goals in reading, writing and mathematics. Each
of the three areas uses all of the test data in that
area – for example, writing includes the two New
Standards writing subtests and RI Writing – and takes
the average scores over the last three years. Using
three years of data evens out the peaks and valleys
resulting from particularly inspired or challenged
classes as well as from the different forms of the test.
The result, then, is a visual description of our
accomplishments – the lighter-colored bars placed above
the lines – and our remaining challenges – the large
gaps between where we are and 100% proficiency.
Because our tests are consistent with the NAEP, we can
assume that though the state and national tests are very
different from one another, the graph roughly
approximates how our performance would appear in
national comparisons. Again, we are facing our challenge
head on. Note the pattern of declining proficiency in all areas
from elementary through secondary. Especially
disheartening is student performance at the high school
level, because the academic achievement of the nearly
19% of the students who have dropped out by grade 12 –
whose academic achievement we conjecture to be poor – is
not even included in the chart. However, our 4th grade reading scores are a bright spot
made possible by a state-wide effort to target funds,
training and attention to elementary reading. Concerted
efforts, whether local or statewide, do pay off in
improved academic performance.
The Office of
Assessment recommends: |
|
1. Don’t think that you can dramatically improve test
scores by focusing on the test itself. Focus instead on
improving instructional practices to meet the needs of
all learners. Become expert in your content and the
pedagogy needed to support your students. If your goal
is to support children for success, you don’t need to
worry about the test. Students who are able to think,
read, write, and solve problems are well prepared for
any testing demand.
2. Don’t give practice tests a few weeks (days) before
the state assessments begin. Look at the types of
responses required on the assessments. Most of them ask
students to read critically, write clearly, and approach
problem solving strategically. These skills can’t be
introduced in a couple of weeks. They should be
integrated into each day’s work. See #1.
3. Don’t introduce accommodations on the day of testing.
Know your students’ challenges well enough to
incorporate necessary strategies or accommodations
throughout the school year. Remember, the accommodations
allowable on state assessments are not limited to
children with IEPs; all students may have access to
them. The only caveat is that the accommodations used in
testing are the same as those typically used in the
classroom. For instance, don’t introduce scribing on the
day of the writing assessment. If a child needs this
support for a state assessment, he would probably
benefit from this help every time he writes.
4. Make all teachers responsible for a school’s success
or failure. We all hear about teachers fleeing from
grades 4, 8 and 10 because they feel the pressure of
state testing. In reality, students begin learning what
they need for state assessments in kindergarten. Even
the most talented teacher can not prepare a student for
four years’ worth of standards in any one year. When
scores arrive, celebration or concern should be shared
among everyone in the building.
5. Use more than rubrics to capture all we that want our
children to learn. The over-use, mis-use, and crazy-use
of rubrics (preset criteria for quality) has gotten out
of control, as though we’re rubricizing American
schools. Rubrics can articulate the features of quality
work. However, their real magic is when they are paired
with student work that has teacher comments. This
pairing helps students and teachers understand what it
is about the work that meets standard or what can be
done to improve it so that it can meet standard. Simply
labeling work as a “6” as described on the rubric does
little more than giving something an “A”.
6. Don’t use state assessments as a primary tool to
diagnose individual students’ needs and strengths. The
state assessment results should confirm what you already
know about a student’s ability to perform “on demand.”
Daily, teachers gain insights into student learning by
assessing student work and exhibitions, using their
knowledge of content and understanding of performance
standards to make informed judgments.
7. Convey via teacher attitudes how you want students to
respond to the state assessments. There’s a balance
between paralyzing students with fear and conveying that
the tests needn’t be taken seriously because they “don’t
count”. We know this balance exists because we’ve
observed it in many classrooms, at all grade levels and
with different populations of children.
8. Teach to the standards on which the test is based.
Many educators believe that if they had the test
blueprint for each year’s assessment, they could do a
much better job improving test performance. Perhaps they
are right, but such test preparation would narrow the
curriculum and expose students to only a fraction of the
things they need to know and be able to do for success
in the modern world. No state test is broad enough to
address all the content standards that should have been
cumulatively taught by the time of the tested grades.
For example, if teachers knew that Japanese folk tales
would be the genre used in a state reading test, the
majority of the year would be spent reading Japanese
folk tales and all other genres would get little
attention. A well-articulated curriculum that spans
grades k-12 should prepare students for any focus a test
might have.
9. The entire school community needs to cultivate
assessment literacy. Assessment is complex and
confusing. Different measures tell us different things
about student performance. If a student scores well on
one type of writing test but less so on another, one of
the tests is not “wrong, “ but probably tests different
types of writing, different intellectual skills, and/or
different process skills. It doesn’t matter which is
more accurate; it matters that we talk about how to
support high student performance under different
conditions for many writing purposes.
10. Don’t worry about testing! Testing will never go
away; indeed, we’ll be doing more of it to comply with
federal mandates. That’s okay if we understand
assessment and work together to design systems that help
us measure what we think is important for all of our
children to learn! |
|
