ERIC Identifier: ED338705
Publication Date: 1991-12-00
Author: Baker, Eva L. - Linn, Robert L.
Clearinghouse on Tests Measurement and Evaluation Washington DC.
The Center for Research on Evaluation, Standards, and Student
Testing (CRESST). ERIC/TM Digest.
How do we judge the quality of education? How do we determine what our
students know and what they can do? What information should teachers,
administrators, and policymakers have to make improvements in our educational
system? How can we improve the kinds of measures we use so that assessment
contributes to the well-being of individuals and organizations?
The mission of the new, OERI-supported Center for Research on Evaluation,
Standards, and Student Testing (CRESST) focuses on these essential questions,
seeking to advance understanding of educational quality through research and
development on the design, implementation, analysis, and use of assessment
After describing CRESST's R&D goals and underlying principles, this
Digest outlines three major programs of the Center.
the infrastructure for improved assessment
improved learning based assessments
to develop and improve assessments in practice
RESEARCH AND DEVELOPMENT GOALS
Recognizing the shortcomings
of traditional testing methods and realizing the need to achieve national
educational goals, CRESST's research programs are directed at five major goals:
Provide leadership to improve assessment policy and practice at the national,
state, and local levels.
objective includes: national agenda setting, establishing mechanisms for the
exchange of state-of-the-art assessment research and products, reinforcing a
national infrastructure supporting assessment change, scrutiny of existing
practices, and communicating with the broad array of policy and practitioner
constituencies interested in improved educational practice. Program One is
devoted to reaching this goal.
Improve the quality, sensitivity, and fairness of student performance
objective, conducted principally in Program Two, includes synthesis of
state-of-the-art alternatives, development of new prototype assessments
measuring national educational goals, and strengthening the underlying design
and validation theory of new and existing assessment alternatives.
Improve the validity of models and indicators for judging the quality of
on this objective, conducted in Programs Two and Three, includes synthesis of
existing alternatives, development of new models linking the results of
measures, and equitably assessing the progress and quality of schools.
Improve understanding of assessment development, implementation, and effects as
they occur in school practice.
objective, conducted principally in Program Three, includes: analyses of
alternative models for implementing new assessment approaches at the state,
local, and classroom levels; clarification of potential impediments and
facilitators in implementing new approaches; and scrutiny of actual effects of
new and existing alternatives.
Improve understanding of assessment policy and its contribution to educational
on this objective, conducted principally in Programs One and Three, includes:
analysis of the interplay between policy goals and assessment design, studies of
alternative policy configurations, and formulation of principles that optimize
CRITERIA FOR IMPROVED EDUCATIONAL ASSESSMENT
proposed work's core is an expanded set of criteria for judging the validity of
educational assessment. These criteria, to be refined and modified through the
course of the CRESST award, focus attention on the consequences and character of
assessments as well as on more traditional technical and practical issues.
Beyond traditional concerns with reliability and validity for specific purposes,
the criteria include attention to:
The criteria serve as a focus for CRESST research and as new standards which
others can use to evaluate their assessment alternatives. They also underscore
our interest in improving the quality of current practice.
OVERVIEW OF CRESST RESEARCH PROGRAMS
@Program One: Building the Infrastructure for Improved Assessment
Program One seeks to assure that the R&D efforts of CRESST and of the
many other players currently working on assessment reform will have optimal
impact on educational policy and practice. To accomplish this goal, the program
will: strengthen the network of those currently engaged in assessment
initiatives, promote collaboration between R&D projects and those seeking to
change assessment, and initiate strategies to promote change in assessment
policies and practices. In close contact and collaboration with major players in
the assessment arena, the program will also serve an important national
agenda-setting function, helping to target precious R&D resources on areas
to critical need. @Program Two: Designs for Learning-Based Assessments:
Prototypes and @Models
Program Two will create new prototypes for assessing student performance and
new models for analyzing and validating assessment results. Requiring major
rethinking of both traditional forms of testing and traditional psychometric
theory, the prototypes will be built on new developments in cognitive and
curriculum theory and on renewed understanding of the types of capabilities our
nation and its citizens need for future productivity. @Program Three:
Collaborative Development and Improvement of @Assessments in Practice
Program Three also addresses the creation of alternative assessments. Here,
however, the developmental work is conducted collaboratively with state and/or
local constituencies, subject to their time and resource constraints and
targeted to their goals and objectives. Beyond technical concerns with the
nature of innovative student assessments, work in this program also seeks to
understand the implementation and impact of new alternatives in policy and
practice. What are the effects of various assessments and their combinations on
students, teachers, administrators, parents, and policymakers? How can
assessments best be configured to meet multiple user needs? How can they best
support learning and instructional improvement? Focusing particularly on serving
low-income and disadvantaged students, the program will produce solutions to
common technical problems, develop strategies for facilitating the
implementation of improved assessment methods, and make recommendations for
The CRESST team is composed of researchers from
the following institutions: UCLA Center for the Study of Evaluation, the lead
institution; University of Colorado; RAND Corporation; University of Chicago;
University of California, Santa Barbara; University of Pittsburgh. In addition
to the practitioner communities identified above, we have cooperative
relationships with professional organizations, organizations of business and
industry, the military, and individual consultants.