Benchmark Assessment for

Improved Learning
A N AAC C P OL IC Y BR IE F

Joan L. Herman, Ellen Osmundson, & Ronald Dietel

For a more detailed report, please refer to our Full Report available at aacompcenter.org

AACC: Assessment and Accountability Comprehensive Center: A WestEd and CRESST partnership.
aacompcenter.org

Copyright © 2010 The Regents of the University of California

The work reported herein was supported by WestEd, grant number 4956 s05-093, as administered by the U.S.
Department of Education. The findings and opinions expressed herein are those of the author(s) and do not
necessarily reflect the positionsor policies of AACC, WestEd, or the U.S. Department of Education.

To cite from this report, please use the following as your APA reference:
Herman, J. L., Osmundson, E., & Dietel, R. (2010). Benchmark assessments for improved learning (AACC Policy Brief).
Los Angeles, CA: University of California.

The authors thank the following for reviewing this policy brief and providing feedback and recommendations:
Margaret Heritage, (CRESST); and for editorial and design support: Judy K. Lee and Amy Otteson (CRESST).
 Benchmark Assessments for Improved Learning
An AACC Policy Brief

Joan L. Herman, Ellen Osmundson, & Ronald Dietel

INTRODUCTION

The No Child Left Behind Act of 2001 (NCLB, 2002) has produced an explosion of interest in the use of
assessment to measure and improve student learning. Initially focused on annual state tests, educators
quickly learned that results came too little and too late to identify students who were falling behind. At
the same time, evidence from the other end of the assessment spectrum was clear: teachers’ ongoing use
of assessment to guide and inform instruction—classroom formative assessment—can lead to statisti-
cally significant gains in student learning (Black & Wiliam, 1998).

Between state and formative assessment is benchmark assessment1, defined as follows:

Benchmark assessments are assessments administered periodically throughout the school year, at speci-
fied times during a curriculum sequence, to evaluate students’ knowledge and skills relative to an explicit
set of longer-term learning goals. The design and choice of benchmark assessments is driven by the pur-
pose, intended users, and uses of the instruments. Benchmark assessment can inform policy, instructional
planning, and decision-making at the classroom, school and/or district levels.

In the following sections, we describe the role of benchmark assessment in a balanced system of assess-
ment, establish purposes and criteria for selecting or developing benchmark assessments, and consider
organizational capacities needed to support sound use.

A BALANCED ASSESSMENT SYSTEM

Benchmark assessment operates best when it is seen as one component of a balanced assessment system
explicitly designed to provide the ongoing data needed to serve district, school, and classroom improve-
ment needs. The National Research Council (NRC) defines a quality assessment system as one that is (a)
coherent, (b) comprehensive, and (c) continuous (NRC, 2001).

Components of a coherent system are aligned with the same significant, agreed-upon goals for student
learning, (i.e. Important learning standards). A comprehensive system addresses the full range of knowl-
edge and skills expected by standards while providing district, school and teachers with data to meet their
decision-making needs. A system that is continuous provides ongoing data throughout the year.

1 We consider the terms interim assessment, quarterly assessment, progress monitoring, medium-cycle, and medium-scale
assessment interchangeable with benchmark assessment.

Benchmark Assessments for Improved Learning 1

 FIGURE 1.
Quality Assessment System: Multiple formative classroom assessment feeding into each benchmark
assessment and multiple benchmark assessment feeding into annual assessment.

Classroom Benchmark Annual

Assessment Assessment Assessment

Where do benchmark assessments fit in a balanced assessment system? While annual state assessments
provide a general indicator of how students are doing relative to annual learning standards, and while
formative assessment is embedded in ongoing classroom instruction to inform immediate teaching and
learning goals, benchmark assessments occupy a middle position strategically located and administered
outside daily classroom use but inside the school and/or district curriculum. Often uniform in tim-
ing and content across classrooms and schools, benchmark assessment results can be aggregated at the
classroom, grade, school, and district levels to school and district decision-makers, as well as to teachers.
This interim indication of how well students are learning can fuel action, where needed, and accelerate
progress toward annual goals.

Figure 1 highlights our conceptualization of the interrelationships between these three types of assess-
ments—classroom, benchmark, and annual—in a balanced system. The learning targets assessed by
frequent classroom-formative assessment contribute to the long-term targets addressed by periodic
benchmark assessments. Benchmark data flows into the annual assessment, which in turn transfers into
subsequent years of teaching, learning, and assessment.

Key Questions to Consider When Selecting Benchmark Assessments

As educational leaders consider the addition of benchmark assessments to an already assessment-heavy
calendar, it is important to establish clear understandings of the nature and purpose of these assessments.
We suggest that policymakers answer the following questions prior to adopting or developing bench-
mark assessments for their school or district:

1. What purposes do you expect benchmark assessments to serve?

2. What criteria should you use to select or create benchmark assessments?
3. What organizational capacity is needed to successfully support a benchmark assessment
program?

2 Joan L. Herman, Ellen Osmundson, & Ronald Dietel :: AACC Report

 PURPOSES OF BENCHMARK ASSESSMENTS

Benchmark assessments often serve four interrelated but distinct purposes: (a) communicate expecta-
tions for learning, (b) plan curriculum and instruction, (c) monitor and evaluate instructional and/or
program effectiveness, and (d) predict future performance. We briefly discuss and illustrate examples of
each purpose, highlighting what, how, and by whom the results could be used. Note that the four pur-
poses are not mutually exclusive—many benchmark assessments address more than one purpose.

Communicate Expectations

Benchmark assessments communicate a strong message to students, teachers, and parents about what
knowledge and which skills are important to learn. Teachers, who want their students to perform well
on important assessments, tend to focus classroom curriculum and instruction on what will be assessed
and to mimic assessment formats (see, for example, Herman, 2009). This last quality, how learning is
measured, provides additional rationale for not limiting benchmark assessments to traditional multiple-
choice formats, which too often emphasize low-level knowledge. Constructed response items (i.e., essays,
extended multi-part questions, portfolios, or even experiments) as can innovative multiple-choice items,
not only provide an important window into students’ thinking and understanding, but also can commu-
nicate expectations that complex thinking and problem-solving should be a regular part of curriculum
and instruction.

Plan Curriculum and Instruction

Benchmark assessments can serve curriculum and instructional planning purposes by providing educa-
tors information needed to adjust curriculum and instruction to meet student learning needs. To do so,
benchmark assessments must be aligned with content standards and major learning goals and provide
reliable information on students’ strengths and weaknesses relative those goals.

Monitor and Evaluate Learning

Benchmark assessments can also be used for monitoring and evaluation purposes, by providing informa-
tion to teachers, schools, or districts about how well programs, curriculum, or other resources are help-
ing students achieve learning goals. Benchmark assessments can help administrators or educators make
mid-course corrections if data reveal patterns of insufficient performance, and may highlight areas where
a curriculum should be refined or supplemented. Special care, however, must be taken in using bench-
mark assessment to monitor or evaluate student progress. Most benchmark assessments are designed
to measure what students have learned during the previous period of instruction and do not provide an
indicator of progress. Simply comparing students’ scores from one time point to the other does not tell
you whether student performance is improving, unless the tests are specially designed to do so.

Benchmark Assessments for Improved Learning 3

 Predict Future Performance

Benchmark assessment can provide data to predict whether students, classes, schools and districts are
on course to meet specific year-end goals—or commonly, be classified as proficient on the end-of-year
state test. Results that predict end-of-year performance can be disaggregated at the individual student,
subgroup, classroom, and school levels to identify who needs help and to provide it.

Addressing Multiple Purposes

Given the scarcity of time and resources in educational settings, it should come as no surprise that many
organizations attempt to use one assessment for multiple purposes. However, the National Research
Council warns: “…the more purposes a single assessment aims to serve, the more each purpose is com-
promised. (NRC, p. 53, 2001).

CRITERIA FOR BENCHMARK ASSESSMENTS

A plethora of benchmark assessments are currently available to educators, ranging from glossy, high-tech
versions designed by testing companies to locally developed, teacher-driven assessments. Below we de-
scribe important criteria and principles that schools, districts, and states should consider when selecting
or developing benchmark assessments.

Validity
Validity is the overarching concept that defines quality in educational measurement. Simply put, valid-
ity asks the extent to which an assessment actually measures what it is intended to measure and pro-
vides sound information supporting the purpose(s) for which it is used. The dual definition means that
benchmark assessments themselves are not valid or invalid, rather that validity resides in the evidence
underlying an assessment’s specific use. An assessment whose scores have a high degree of validity for
one purpose may have little validity for another. For example, a benchmark reading assessment may be
valid for identifying students likely to fall short of proficiency on a state test but may have little validity
for diagnosing the specific causes of students’ reading difficulties.

The evaluation of quality in any benchmark assessment system starts with a clear description of the
purpose(s) an assessment is intended to serve and serious consideration of a range of interrelated issues
bearing on how well a given assessment or assessment system serves that purpose(s).

4 Joan L. Herman, Ellen Osmundson, & Ronald Dietel :: AACC Report

 Alignment
Alignment describes how well what is assessed—the content and the type of learning—matches both
what schools are trying to teach and the assessment purposes. The over-riding questions are the extent
to which the assessments reflect:

• What is most important for students to know and be able to do in a specific content area?
• What is the depth and breadth of district and school learning goals?
• What is the sequence of the local curriculum?
What to look for:
Many assessment companies report that their assessments are aligned with specific state standards.
But educators should turn to the assessment framework or specifications and even the item pools
provided by assessment providers to examine the match between the assessments and their curriculum
and purposes.

Alignment to Curriculum.

An independent benchmark assessment analysis should compare school or district goal specifications
with the:

1. Framework used to develop or classify available items: Is it consistent with local conception
of curriculum and what’s important for students to learn?
2. Distribution and range of assessment item content by grade level: do they match those of
the local curriculum or domain specification?
3. Distribution and range of types of learning addressed by grade level: do they cover the full
range including complex thinking and problem solving applications?
4. Specific distribution of items on each assessment by content and cognitive demand: do they
reasonably represent the learning goals for the instructional period being assessed?

Alignment for Intended Purposes.

Alignment with learning goals is fundamental, but benchmark assessments must also be designed to
serve their intended purpose(s). If benchmark assessments are intended to serve instructional planning
purposes, they must provide diagnostic feedback on student strengths and weaknesses and help to iden-
tify the source of student difficulties. Good diagnosis involves mapping assessment items to a theory of
how each learning toward particular objectives or standards develops and of the major obstacles or mis-
conceptions that occur. Distractors for multiple-choice items, for example, can be designed to represent
common errors. Look for the developers’ theory of diagnosis and whether it is consistent with teacher
and curriculum perspectives.

Benchmark Assessments for Improved Learning 5

 For benchmark assessment that are intended to serve predictive purposes, the benchmark assessment
scores should be highly correlated with proficiency levels on the end-of-year test. Known as predictive
validity, correlations in the range of 0.7 and above between benchmark assessment results and the state
assessments provide a reasonable amount of certainty that students who perform well on the benchmark
assessment will also perform well on the state assessment (Williams, 2009). However, if teachers suc-
cessfully use benchmark results to help low performing students improve, then the strong relationship
between benchmark and end-of-year state test scores may decline.

RELIABILITY

Reliability is an indication of how consistently an assessment measures its intended target and the extent
to which scores are relatively free of error. Low reliability means that scores should not be trusted for
decision-making.

Measurement experts have a variety of ways of looking at reliability. For example, they look for con-
sistency of scores across different times or occasions (i.e., whether a student took a test early in the day
or later, now or next week). Measurement experts also look for reliability in scoring: results should be
consistent regardless of who scores the test or when. Reliability is a necessary but not sufficient criterion
of test validity. For example, an assessment may be highly reliable, but might not measure the “right”
knowledge and skills. Alternately an assessment may provide a highly reliable total score, but not provide
reliable diagnostic information.

What to look for:

Test publishers typically provide reliability indices for their benchmark assessments with other technical
information about item difficulty and discrimination. It is essential to review this technical information
before purchasing or using benchmark assessments or item banks. For schools and districts developing
their own benchmark assessments, specific statistical guidelines should be used to evaluate the reliability
of assessment items prior to their widespread use (Brown & Coughlin, 2007).

Fairness and Bias

Fairness and bias comprise the next critical feature of quality benchmark assessments.

A fair test is accessible and enables all students to show what they know; it does not advantage some
students over others. Bias emerges when features of the assessment itself impede students’ ability to
demonstrate their knowledge or skill. Technically, bias is present when students from different subgroups
(e.g., race, ethnicity, language, culture, gender, disability) with the same level of knowledge and skills
perform differently on an assessment.

6 Joan L. Herman, Ellen Osmundson, & Ronald Dietel :: AACC Report

There are two primary forms of test bias: offensiveness, and unfair penalization (Popham, 1995).
Offensiveness becomes an issue when the content of an assessment offends, upsets, or distresses particu-
lar subgroups, thereby negatively impacting performance. Assessment items that present unfavorable
stereotypes of different cultures, genders, or other subgroups could adversely affect these subgroups’
performance.

Unfair penalization occurs when the content-irrelevant aspects of an assessment make the test more
challenging for some students than for others because of differences in language, culture, locale, or
socioeconomic status.

What to look for:

Technical information regarding fairness and bias should be provided by benchmark assessment devel-
opers and should include demographics of the sample as well as scores and other technical evidence for
various subgroups. Additionally, benchmark assessments should be examined prior to their use to ensure
that the particular items will not be offensive to students in that organization. Guidelines for developing
benchmark assessments that are free from bias reflect the same steps outlined above:

1. Developers should be sensitive to the demographic characteristics of the students;

2. Documentation describing the steps taken to minimize bias in the assessment items should
be provided;
3. Organizations should examine how well an item functions for specific subgroups; and
4. If particular subgroups perform differently, validity of those items for the subgroups should
be investigated.

High Utility
A final criterion important to consider when selecting or developing benchmark assessments is utility.
The overarching question schools, districts and states should ask to determine a benchmark assessment’s
utility is: How useful will this assessment be in helping us to accomplish our intended purposes?

What to look for:

Specific utility questions to ask include what are the costs of purchasing or creating benchmark assess-
ments? Who scores the assessments and what are the associated costs? What training may be necessary
and how will it fit into existing schedules? Are there opportunity costs to be considered for students,
teachers and administrators? How will the results be reported? Who analyzes and reports the scores and
to which groups? How will the results fit with other assessments, both formative and end-of-year state
tests? Who will use the results? How will we evaluate the quality of the assessments and their use?

Benchmark Assessments for Improved Learning 7

 Balance
No benchmark assessment program will perfectly match the preceding criteria nor meet all of a school or
district’s assessment needs. There will be tradeoffs, with each benchmark assessment embodying differ-
ent strengths and weaknesses. One benchmark assessment for example, may be well aligned with district
learning goals and state standards, but be too difficult or costly to administer. Another benchmark as-
sessment may be more feasible, but fail to provide good diagnostic information or lack data on fairness
or accommodations for special populations. However, the preceding criteria, supporting the validity of a
quality benchmark system, will help you find the right balance for your district or school on fairness or
accommodations for special populations. The preceding criteria, which support the validity of a quality
assessment system, can help you find the right balance for your district or school.

B U I L D I N G O R G A N I Z AT I O N A L C A PA C I T Y F O R B E N C H M A R K A S S E S S M E N T

In the process of selecting or developing benchmark assessments, districts and schools need to carefully
consider the infrastructure and systems needed for the benchmark assessment process to run smoothly
and efficiently so that educators can make good use of assessment results. Decisions about how, when,
and by whom the assessments will be administered, scored, analyzed, and used will influence the kinds
of resources and support school personnel need. Below we describe four conditions necessary to sustain
effective use of benchmark assessments adapted from current research (Goertz, Olah, & Riggan, in press)
and practice. Undergirding the plan is a culture conducive to data use, including high expectations, trust,
and valuing data.

1. Begin with a written plan.

Investing in benchmark assessments is a costly, time-intensive undertaking. A written district or school
accountability plan, including benchmark assessments, can help your school or district reach goals and
save time. At minimum we suggest that your plan includes the purposes of the benchmark assessments,
individual responsibilities, reporting, data use, professional development, resources, and evaluation.

2. Identify systems for analyzing and reporting data.

Whether an organization chooses to purchase a software program to make benchmark assessment data
available to teachers, principals, and districts or to develop their own tool, it is important that data are
quickly and easily available to all stakeholders. Clear rules should be established and communicated
widely to ensure users have access to the data they need without compromising confidentiality or the
data system.

8 Joan L. Herman, Ellen Osmundson, & Ronald Dietel :: AACC Report

3. Provide professional development.
Schools and districts can do much to encourage the use of data from the benchmark assessments by pro-
viding high-quality, ongoing professional development. Teachers, schools, and district personnel require
assistance in building their technical skills to access, organize, and interpret benchmark assessment data.
Professional development should include content and pedagogical skills that help teachers differentiate
instruction and revise instructional strategies and approaches based on data.

4. Allocate time.
Districts and schools should build time into their calendars to make effective use of benchmark data.
Data users, including assessment and content experts, need time to adequately analyze the data in ways
that are both meaningful to their context and robust in terms of the analyses. Teachers need time for in-
structional planning to address weaknesses identified by assessment results. Time is also needed to enact
instructional plans. There is little value in pinpointing gaps in student understanding if the district pac-
ing guide mandates that teachers forge ahead to the next topic, regardless of student performance and
needs

CONCLUSION

Good benchmark assessments can be an important addition to a comprehensive assessment system.

Benchmark assessments should be well aligned with curriculum and provide a continuous, comprehen-
sive stream of information to plan and guide instruction. Validity, sufficient resources, and professional
development are three other key components.

For a more detailed report, please refer to our Full Report available at aacompcenter.org

Benchmark Assessments for Improved Learning 9

 REFERENCES

Black, P., & Wiliam, D. (1998). Assessment and classroom learning. Educational Assessment: Principles, Policy
and Practice. 5(1), 7–74.

Brown, R. S., & Coughlin, E. (2007). The predictive validity of selected benchmark assessments used in the Mid-
Atlantic Region. Retrieved from: http://www.mhkids.com/media/articles/pdfs/resources/PredictiveValid-
ity.pdf

Goertz, M. E., Olah, L. N., & Riggan, M. (in press). Using formative assessments: The role of policy supports.
Madison, WI: Consortium for Policy Research in Education.

Herman, J. L. (2009). Moving to the next generation of standards for science: Building on recent practices
(CRESST Report 762). Los Angeles: University of California, National Center for Research on Evaluation,
Standards, and Student Testing (CRESST).

National Research Council. (2001). Knowing what students know: The science and design of educational as-
sessment. Washington, DC: National Academy of Sciences.

No Child Left Behind Act of 2001, Pub. L No. 107–110, 115 Stat. 1425 (2002).

Popham, W. J. (1995). Classroom assessment: What teachers need to know. Boston, MA: Allyn and Bacon.

Williams, L. L. (2009). Benchmark testing and success on the Texas Assessment of Knowledge and Skills: A cor-
relational analysis (Doctorate dissertation, Publication No. AAT 3353754). Phoenix, AZ: University of
Phoenix. Retrieved from: http://gradworks.umi.com/33/53/3353754.html

10 Joan L. Herman, Ellen Osmundson, & Ronald Dietel :: AACC Report