ERIC Identifier: ED381984
Publication Date: 1995-06-00
Author: Fuchs, Lynn S.
Source: ERIC Clearinghouse on
Disabilities and Gifted Education Reston VA.
Connecting Performance Assessment to Instruction: A Comparison
of Behavioral Assessment, Mastery Learning, Curriculum-Based Measurement, and
Performance Assessment. ERIC Digest E530.
A major impetus for the performance assessment movement has been the need to
reconnect large-scale and classroom assessment to learning so that assessment
affects learning positively, enhancing instruction.
IN WHAT WAYS CAN ASSESSMENT ENHANCE INSTRUCTION?
teachers are better informed of the learning progress and difficulties of their
students, they can make better decisions about what a student needs to learn
next and how to teach that material in a manner that will maximize the student's
learning. Teachers make three types of decisions using assessment results:
1. Instructional placement decisions--what the student knows and where he or
she should be in the instructional sequence--i.e., what to teach next.
2. Formative evaluation decisions--information to monitor a student's
learning while an instructional program is underway--how quickly progress is
being made, whether the instructional program is effective, and whether a change
in instructional program is needed to promote the student's learning.
3. Diagnostic decisions--which specific difficulties account for the
student's inadequate progress so the teacher can remediate learning progress and
design more effective instructional plans.
WHAT CRITERIA SHOULD ASSESSMENTS MEET IF THEY ARE TO INFORM INSTRUCTIONAL DECISIONS?
These assessments should meet seven
Measure important learning outcomes.
Address all three purposes of assessment.
Provide clear descriptions of student performance that can be linked to
Be compatible with a variety of instructional models.
Be easily administered, scored, and interpreted by teachers.
Communicate the goals of learning to teachers and students.
Generate accurate, meaningful information (i.e., be reliable and valid).
HOW DOES PERFORMANCE ASSESSMENT COMPARE TO OTHER METHODS OF LINKING ASSESSMENT TO INSTRUCTION?
Other methods of linking
assessment to instruction include behavioral assessment, mastery learning, and
*Behavioral assessment. Behavioral assessment relies on direct observation
and recording of target behaviors, using repeated observations in the setting
where the behavior occurs. Environmental factors (i.e., the situations in which
the behaviors occur) and their effect on the behaviors are examined.
For example, if a teacher wanted to instruct a student in grocery shopping,
she would first analyze the tasks associated with grocery shopping, put them in
order, and design behavioral objectives that measure each task. Tasks might
include creating a shopping list, finding the items in the store, and finding
the price of each item. The teacher would then collect data on each task to
identify those in which the student needed instruction. The teacher would begin
instruction at the point in the task sequence where the student was unable to
correctly complete the task. Once the student could correctly complete a task,
the teacher would move on to the next step, moving through the sequence until
all of the tasks were mastered.
Behavioral assessment meets some but not all of the criteria for assessments
listed above. It can inform the teacher about the student's placement in the
instructional sequence and can help the teacher reach formative evaluation and
diagnostic decisions. It communicates clearly what the essential learning
content is, and it is feasible to administer, score, and interpret. In addition,
its repeated measurements support the reliability of assessments. However,
behavioral assessment tends to focus on discrete tasks that do not necessarily
add up to important outcomes. It is limited to observable behaviors, and its
small units of instruction can be difficult for students to piece together and
apply to real-world outcomes. Additionally, the assessment system dictates a
behavioral approach to instruction, which can limit the teacher's instructional
*Mastery learning. In mastery learning, a curriculum is broken down into a
set of subskills, which are then ordered in a hierarchy of instructional
objectives. For each step in the instructional hierarchy, a criterion-referenced
test is designed, and a performance criterion indicating mastery of the subskill
is specified. The teacher starts at the lowest step in the hierarchy, pretests,
teaches the objective, and posttests on the material. If the student does not
demonstrate mastery, the teacher uses corrective strategies until mastery is
achieved. The teacher then advances the student to the next, more difficult step
in the hierarchy.
Like behavioral assessment, mastery learning provides information for
instructional placement, formative evaluation, and diagnostic decisions. It
communicates clearly to teachers and students about what is important to teach
and learn. However, mastery learning suffers from the same limitation as
behavioral assessment: it focuses on discrete behaviors in both assessment and
instruction. Because little emphasis has been placed on its reliability or
validity, users do not know what exactly is being assessed, how to interpret the
resulting information, and how to use the measures effectively. Moreover, the
measurement system dictates a specific approach to instruction, leaving the
teacher few instructional choices. The focus of measurement changes each time a
student achieves mastery of a step in the curriculum, and the steps may be of
unequal difficulty, so progress cannot be judged over time. Finally, because
different students need to be measured simultaneously on different steps of the
curriculum, mastery learning systems can become unmanageable for teachers.
*Curriculum-based measurement (CBM). The focus of CBM is long-term. The
teacher establishes a broad outcome for the student such as competently
performing mathematics at the third-grade level at the end of the school year.
Then the teacher uses CBM methods to measure student proficiency: he or she
creates a pool of equivalent assessments, each of which samples the key problem
types from the third grade curriculum. Each week, the student completes one or
two assessments. Because each assessment is of equal difficulty and incorporates
all of the important problem types to be learned over the year, the CBM data
base produces a total score graphed over time to show progress over the year.
Analysis of the student's performance on separate skills embedded in the
assessment can also be conducted for diagnostic problem-solving to improve the
CBM satisfies six of the criteria for assessments. It addresses the three
purposes for assessment, and it incorporates standardized measurement
techniques, providing reliability and validity. It offers detailed information
on a student's performance on specific skills and can be used to determine how
to improve an instructional program. Its measurement framework is not tied to
any particular model of instruction, so a broad range of instructional options
can be used. A teacher can use widely varying methods with the same child to see
which method is most beneficial. Students know how they are evaluated and can
set personal learning goals. In addition, the assessment demands are manageable
in classroom settings, and to make them even more easily manageable, computer
programs have been developed to administer assessments and manage the data.
However, CBM has two drawbacks with respect to the criteria for assessments.
The system requires longer time periods to reveal growth, and the connection
between assessment results and instructional decisions is not as clear as with
behavioral assessment or mastery learning. Controversy also exists about the
importance of the learning outcomes associated with CBM. That is, it relies on
pencil and paper tasks in math and spelling and one-dimensional assessments in
reading, while current discussions about outcomes stress the utility of
multidimensional measures that can cut across curriculum areas.
*Performance assessment. Three key features of performance assessment are:
(1) students construct, rather than select, responses; (2)assessment formats
allow teachers to observe student behavior on tasks reflecting real-world
requirements; and (3) scoring reveals patterns in students' learning and
An example of a performance assessment task is provided below:
group of five families on your block is going to have a garage sale in which
clothes, toys, and books will be sold. Your family has 12 items to sell and will
need 18 square feet to display these items; the Hamletts have 13 items and need
20 square feet; the Phillips, 7 items and 10 square feet; the Garcias, 15 items
and 15 square feet; the Nguyens, 10 items and 30 square feet. Rental tables
measure 6 feet by 2.5 feet and cost $6.00 a day. The garage where the sale will
be held is 20 feet by 30 feet. Newspaper advertising costs $11 for the first 10
words and $1.50 for each additional word.
How many tables will you need? Explain how you got this number.
Draw a diagram showing how the tables can be arranged in the garage to allow the
customers to move about with at least 4 feet between tables.
Write an ad for your sale that includes enough information.
How much money do you have to earn from your sale for the families to break
The students are aware of the scoring system and the criteria used to
determine the scores. Their responses will be classified as exemplary,
competent, minimal, inadequate, or no attempt based on a rubric that specifies
the characteristics of responses in each of these categories. This problem
offers one version of what a teacher's use of performance assessment might look
like. In practice, many varieties of performance assessment are used. This
problem measures massed mathematical concepts that include addition,
multiplication, decimals, data analysis, perimeters, area, spatial sense,
graphic representation, money, and communication about mathematics. Students
take about 50 minutes for the assessment, and it can be completed individually
or in small groups. The problem is anchored in a real-life, age-appropriate
situation and represents real applications of mathematics.
HOW WELL DOES PERFORMANCE ASSESSMENT SATISFY THE SEVEN CRITERIA FOR ASSESSMENT?
Today, performance assessment is relatively new,
undeveloped, and yet to be studied systematically. Many practitioners are
experimenting with its use and contributing to its development and refinement.
Yet they are often in the undesirable position of interpreting vague design
features and operationalizing those features into specific assessments on their
own. These assessments take a variety of forms, some of which are closer than
others in approximating the conceptual and theoretical underpinnings of
1. Measure important learning outcomes. The extent to which performance
assessment measures important student outcomes depends on the specific
assessment problem or task. Performance assessment tasks should reflect
important, real-world performances that are tied to desired student outcomes
that are relevant to the workplace and everyday life. They should connect
meaningfully with specific instructional methods that can be realistically
managed in school settings.
2. Address all three purposes of assessment. It is unclear how performance
assessment can be used to formulate instructional placement or formative
evaluation decisions. Ideally, alternate forms of the problem could include the
same concepts administered over time in order to yield information about
individual students' progress. Although performance assessment offers the
promise of addressing all three assessment purposes, specific methods for doing
so have yet to be developed.
3. Provide clear descriptions of student performance that can be linked to
instructional actions. When performance assessment tasks address a variety of
concepts in age-appropriate, real-world situations, teachers can form a picture
of student performance across skills and identify the student's problem-solving
strategies. However, this depends on the teacher's skill in identifying student
competencies, gleaning information about students' strategic behavior, and
relating these observations to specific instructional techniques. Consultation
methods or computerized strategies for generating profiles of student competence
4. Be compatible with a variety of instructional models. Theoretically,
performance assessment could be used with a variety of instructional approaches.
Teachers should experiment with a variety of instructional methods as they
implement performance assessment, especially with students who have serious
5. Be easily administered, scored, and interpreted by teachers. Performance
assessment can require large amounts of teacher time to design and administer
assessments and to scrutinize student performances. It is easy to see how this
type of assessment could generate so many different plans for intervention
strategies for different students that teachers in a classroom situation with 20
or 30 students would be unable to manage. Performance assessment developers need
to solve the problem of how to implement plans based on performance assessments
within the constraints of classroom life.
6. Communicate the goals of learning to teachers and students. When it is
clearly apparent that an assessment is aligned with instructional goals,
teachers should be able to use that assessment to direct their instruction, and
students should be able to use it to establish personal learning goals. This
depends, however, on the extent to which the scoring rubric used is clear,
concrete, and visible.
7. Generate accurate, meaningful information (i.e., be reliable and valid).
Performance assessment represents a vision that can shape the future direction
of classroom-based assessment, but it requires much additional scrutiny and
development before it can fulfill its promise. REFSmDerived from Fuchs, L.
(1994). Connecting Performance Assessment to Instruction. Reston, VA: The
Council for Exceptional Children. (Product #P5058).