Transcript text

Revisiting
Performance
Assessment 2.0
st
in 21 Century
Accountability
Eva L. Baker
National Center for Research on Evaluation, Standards, & Student Testing
Graduate School of Education & Information Studies
University of California, Los Angeles
Korea Institute for Curriculum and Evaluation
Seoul, Korea
27 October 2010
KICE 1
© Regents of the University of California
Assessment Revisited: Update
on U.S. Educational Policy 2.0
• What have been the major issues in U.S. plans
for reform, assessments, and accountability?
• What is Race to the Top?
• Where are we now? Performance assessment?
• What practical changes will need to occur?
• Comments and concerns
KICE 2
© Regents of the University of California
Educational Reform Update
• Continuation and extension of President
Bush’s policies
• Educational performance (NAEP) is flat with
some small gap closing; international standing
poor
• Belief is that large bureaucracies (districts)
and teacher unions cause low growth
• Adopting of policies to centralize some
aspects of education and decentralize others
• Competition among States for resources
KICE 3
© Regents of the University of California
Race to the Top Provisions
• Adopting standards and assessments that prepare
students to succeed in college and the workplace
(Common Core of Standards 40+ states)
• Building data systems that measure student growth
and success, and inform teachers and principals
how to improve instruction
• Recruiting, developing, rewarding, and retaining
effective teachers and principals, especially where
they are needed most
• Value Added Modeling (VAM)
• Turning around their lowest-performing schools
KICE 4
© Regents of the University of California
Where Are We Now? Race to
the Top
• Competition among states for discretionary
education money
• Competition heated because of poor economy and
layoffs from schools (40 states have applied)
• Need to get State legislative approval to conform to
federal requirements—charter schools; individual
and teacher databases; different union relationships
• Note: “good” schools and districts have high test
scores, despite evidence of direct practice, cheating,
and reduction in curriculum offerings
• Teachers are mostly blamed for failure, and unions
are demonized, charter schools are the answer
KICE 5
© Regents of the University of California
Additional Efforts
• Standards focus on college- and careerreadiness in reading and math (science and
history)
• Create pipelines and incentives to put the most
effective teachers in high-need schools
• Alternative pathways to teacher and principal
certification
• Charter school legislation (State-lifting caps)
• STEM
KICE 6
© Regents of the University of California
Race to the Top—Round 1
• 10 billion available
• Tennessee and Delaware won round one
1. Tennessee
$500,000,000
2. Delaware
$100,000,000
KICE 7
© Regents of the University of California
Race to the Top Awards—
Round 2
1.
Massachusetts
$250,000,000
2.
New York
$700,000,000
3.
Hawaii
$75,000,000
4.
Florida
$700,000,000
5.
Rhode Island
$75,000,000
6.
District of Columbia
$75,000,000
7.
Maryland
$250,000,000
8.
Georgia
$400,000,000
9.
North Carolina
$400,000,000
10. Ohio
$400,000,000
KICE 8
© Regents of the University of California
Common Core State Standards
(CCSS)
• Mathematics and Literacy, for every grade
• Literacy includes “informative reading and
writing,” topics include biological sciences
and history
• Standards developed at the initiative of the
State education officer and governor
• 35+ states participate, may have additional
standards
KICE 9
© Regents of the University of California
Quality of Common Core State
Standards
• Generally regarded as excellent
• Fewer, Clearer, Higher = Many, Clearer, Higher
• Good illustrative material
• Sequences unverified
• Use of ontologies by CRESST to clarify
meaning and internal relationships funded by
the Gates Foundation
• Developing formative and curriculum-based
assessments and evaluating existing options
KICE 10
© Regents of the University of California
Common Core Standards
Progression
KICE 11
© Regents of the University of California
Formal Representation
1X5
KICE 12
© Regents of the University of California
Assessment Funding for RTT
• To avoid redundant procurement, States
banded together in three different consortia to
bid for resources to design and specify
assessments
• Assessments to include formative, for
classroom use, interim for prediction within
grade, accountability
• Reporting by growth and value-added
modeling
• Enables performance assessment as part of
the package
KICE 13
© Regents of the University of California
Two Consortia Won
• Smarter Balance, with a bottom-up teacher
approach that promises the use of computer
adaptive testing (45) states ($160M) dollars
• PARCC supported by Achieve, Inc., a
Washington think-tank (26 states; $185.87M)
• Period of performance – 4 years, collaboration
with universities, commercial groups
• Both groups at this point seem relatively
conservative in their views
KICE 14
© Regents of the University of California
Financing Assessment
• Total awards for assessment in the $350M
area
• Testing companies are anxious, needing to
win the RFPs from the States even if they
don’t agree with their plans
• Maintenance would go to publishers with
States providing oversight
• Who owns standards? Assessments?
KICE 15
© Regents of the University of California
Smarter Balanced Consortia
• New generation assessment system—deep disciplinary
understanding and higher-order thinking skills demanded
by knowledge-based economies
• Relevant to ongoing improvements in instruction and
learning
• Useful for all members of the educational enterprise:
students, parents, teachers, school administrators,
members of the public, and policymakers
• Formative and summative assessments, organized around
Common Core Standards, that support high-quality
learning and the demands of accountability
• Feasible, innovative assessment and fiscally sustainable
• RFPs to private companies
KICE 16
© Regents of the University of California
Smarter Balance
• Assessments are integrated into a coherent
learning system of standards, curriculum,
assessment, instruction, and teacher
development
• Teachers and other instructional experts
develop formative and summative
assessments of standards—supports are
provided to enable thoughtful teaching
• Teachers are integrally involved in the design,
development, and scoring of assessment items and
tasks
• Teachers’ roles include the construction and review
of items/tasks, the definition of scoring guides,
selection of student work exemplars, and scoring
KICE 17
© Regents of the University of California
21st Century Skills
• Assessments include evidence of actual student
performance on challenging tasks that evaluate
standards of 21st Century learning
• They emphasize deep knowledge of core concepts
within and across the disciplines, problem solving,
analysis, synthesis, and critical thinking
• Technology is designed to support assessment and
learning systems
• by delivering the assessments; enabling adaptive
technologies to better measure student abilities across
the full spectrum of student performance
• evaluate growth in learning
• supporting online simulation tasks that test higher-order
abilities
KICE 18
© Regents of the University of California
Performance Assessment
• Open ended
• Multiple steps
• Synthesis of standards
• More than one content area?
• Team or individual
• Judged by raters or markers
KICE 19
© Regents of the University of California
English Language Arts
KICE 20
© Regents of the University of California
Technology-Driven
• Semi-automated design with templates
• Just-in-time steps
• Help possible
• Simulations of processes
• Complex problem solving and decision
making
• Written products, demos, research
KICE 21
© Regents of the University of California
Technical Issues—Again
• Few “items”—old psychometrics
• Comparability of tasks, quantitative and
qualitative features
• Confounded “difficulty”
• Vertical scaling
• Reporting for accountability, instructional
feedback
• Cost
• Renewal
KICE 22
© Regents of the University of California
Technical Needs
• Robust and rapid online scoring
• Diagnostic and accountability accuracy
• Instructional sensitivity
• Extensive or inappropriate help
• Memorable tasks need to be replenished
• Cost (back to reusable designs)
• Game frameworks
KICE 23
© Regents of the University of California
Additional Concerns
• Economics suggest that total amount available
from Federal government will be reduced
• Likely outcome is reduced innovation
• What kinds of validity studies should be
conducted?
• CRESST scholars have proposed studies of
accuracy of classification, utility of findings for
improvement, instructional sensitivity,
accommodations, and other fairness
approaches—a struggle since there will be few
revision resources and no full trial of the system
KICE 24
© Regents of the University of California
Research and Development
Needs
• Validity approaches
• Modeling
• Sampling
• Verification of sequences of CCS and assessments
• Utility to teachers of different levels of training and
experience
• Openness of feedback
• Continuous modification and longitudinal data
• Value-added issues
KICE 25
© Regents of the University of California
Speaking Truth to Power
• Not easy unless you are friends
• Policymakers in U.S. have odd mix of
research-based ideas and preferences—now
almost an overwhelming wave in one direction
• Desire conflicting goals without recognizing
technical conflicts
• What should researchers do?
KICE 26
© Regents of the University of California
http://www.cse.ucla.edu
Eva L. Baker
voice
fax
email
KICE 27
310.206.1530
310.267.0152
[email protected]
© Regents of the University of California