ML 4.0 Transitional//EN">
Section
III
EMERGING MODES OF
ASSESSMENT
The number of standardized tests being used to assess student learning in
California’s public schools is increasing rapidly. Moreover, the surge in
student testing is likely to continue unabated for the foreseeable future.
President Bush’s agenda for reforming the public schools, known popularly
as “Leave No Child Behind,” will require California to annually
administer more standardized tests, at more grade levels grades, in more subject
areas than ever before. If the increase in state and federal assessment
activities is to be useful in informing the design of educational programs
intended to improve student learning, their results must be made available to
teachers, students, and parents on a timely basis. Such is not the situation
today.
In many instances, the time lag between administration of a
standardized test and release of the results is greater than six months. Few
teachers ever get back test results on their students early enough to modify
their teaching strategies. This delay, particularly in light of the availability
of relatively inexpensive networked database technologies, is inexplicable.
Today, standardized tests are auditing devices, not instruments of educational
practice and improvement.
The problems afflicting standardized testing
programs stem not only from the tardiness of their results, but also from their
design. Many tests continue to focus on those aspects of knowledge that can be
captured in the multiple-choice format. The National Research Council’s
Committee on the Foundations of Assessment has recommended that the current
generation of standardized tests be abandoned. The Committee recommends that the
next generation of these tests provide the information needed by teachers to
track the learning progress of their students at a higher level of detail than
is currently provided by today’s assessment instruments. Such detailed
information would allow teachers to give their students information and guidance
on what they need to do to improve their academic performance. The Committee
also argued that such assessment instruments would better serve the paired goals
of educational equity and excellence.
Three trends show promise of being
able to significantly improve current assessment practices. The first is new
research on cognition and learning, particularly findings on the critical role
timely feedback plays in fostering productive learning. The second is
development of computer-mediated instructional materials that incorporate
sophisticated embedded assessment capabilities, and the findings that these
capabilities foster adaptive learning activities. The third trend is the
emerging generation of distributed database technologies, which can used to
gather, analyze, sort, and disseminate assessment results quickly.
The
Working Group agrees with the findings and recommendations of the National
Research Council. It is therefore recommended that the State take action to make
its assessment programs more flexible, accessible, and responsive.
Collectively, these changes would produce assessment policies and practices that would be more student-focused, learning centered, and supportive of school improvement.
Commentary: Practice-oriented research and documentation can serve as valuable tools and can support decisions to continue or discontinue current practices. This information may also be used as a basis for allocating funding and/or State incentives. Schools often are not provided with specific resources for this activity, but it should be a priority. A priority should also be placed on disseminating ‘best practices’ for potential replication.
High-stakes testing impacts learners by driving
decisions that have important consequences for each student’s future:
promotion, retention, graduation, diploma awards, and possible postsecondary
placements. State and local education agencies should ensure that any
examination used for high-stakes consequences for individual students actually
measures what it is intended to measure for all students. The State must
therefore ensure that all students have equal access to the core curriculum,
regardless of the location of the school district or school, and that the core
curriculum is accurately reflected in the test content. When tests are used in
making educational decisions for individual students, they should accurately
measure the student’s abilities, knowledge, skills, or needs in ways that
do not discriminate or violate federal law on the basis of the student’s
race, national origin, gender, or disability.
The following principles,
which embody research recommendations and ‘best practices’ developed
by the National Research Council, CRESST, and the National Academy of Sciences
Board on Testing and Assessment should apply to any testing that has
consequences for individual students, institutions, or systems:
There is a fine line between using multiple test and assessment measures to make sound educational decisions for students and over-testing learners to meet accountability requirements. |
If tests are claimed to measure content and performance standards, analyses should document the relationship between the items and specific standards or sets of standards. To the extent possible, language assessments should be used to measure academic performance against standards, and English assessments to measure growth in English proficiency.
The validity of measures that have been administered as part of an accountability system should be evaluated and documented for the various purposes of the system.
Evidence of test validity for students with different language backgrounds should be made available publicly.
Speakers of languages other than English should have appropriate assessments based on language and English proficiency.
Standards set for passing or passing at different levels of proficiency should be made clear. In particular, the justification for different ‘cut scores’ should be made on the basis of validity evidence.
Commentary: Students are required to take a
multitude of assessments, many of which cover the same subject matter, thereby
making more sense to combine assessments than to duplicate items. Especially
with the passage of the new federal Leave No Child Behind (the reauthorized
Elementary and Secondary Education Act), the State risks turning schools in to
assessment centers. The State must be careful to develop assessments and tests
that are robust and not duplicative. To support innovative non-duplicative
results, the State should encourage establishment of a cross-segmental forum for
sharing effective practices and ensuring assessment and testing
alignment.
A combination of measures should be used to assess students,
including:
Coordinated information systems would provide students easy access to their own academic records. |
Contents | Summary | Background | I. Delivery |
II. Organization | III. Assessment | IV. Certification | V. Planning |
VI. Adult Ed. | Conclusion | Presenters | Members |